Climate Change and Variability Trends in Pakistan and Its Environmental Effects

CLIMATE CHANGE AND VARIABILITY
TRENDS IN PAKISTAN AND ITS

ENVIRONMENTAL EFFECTS
Thesis submitted to Department of Environmental

Sciences University of Peshawar in partial fulfillment of
the requirements for the degree of Doctor of Philosophy
By
Salma
Department of Environmental Sciences

University of Peshawar
Session: 2011
CONTENTS
List of Contents ------------------- i

List of Tables---------------------- v
List of Figures--------------------- vii
List of Abbreviation----------- ix
Acknowledgment------------------ x
Abstract ----------------------- ---- xi
LIST OF CONTENTS
S. No Title Page No
Chapter # 1 Introduction`` 1
1.1 Climate of Pakistan 14
1.2 Objectives 17
1.3 Justification 17
Chapter#2 Literature Review 18
2.1. Studies related to the determination of Temperature trend 18
2.2 Studies related to the determination of Rainfall trend 21
2.3 Collective studies related to the determination of 22

Temperature, Rainfall and River Discharge trends
Chapter #3 Research Methodology 26
3.1 Data Acquisition 26
3.2 Climate Zones of the study area 27
3.2.1 Zone 1 27
3.2.2. Zone 2 27
3.2.3. Zone 3a 27
3.2.4 Zone 3b 27
3.2.5. Zone 4 28
3.3. Limitation of the data 29
3.4 Data Analysis 30
i
3.4.1. Mean 30
3.4.2. Outliers 30
3.4.3. 5% Trimmed Mean 30
3.4.4. Median 30
3.4.5. Standard Deviation 31
3.4.6. Standard Error 31
3.4.7. 95% confidence interval (CI) for the mean 31
3.4.8. Skewness 31
3.4.9. ANOVA test 32
3.5. Trend Visualization 34
3.5.1. Mapping 34
3.5.2. Graphical Presentations 35
3.6. Trend Forecasting 35
Chapter # 4 Results and Discussion 37

4.1. Temperature 37
4.1.1. Introduction 38
4.1.2. Data collection and analysis 38
4.1.3. Results of temperature data 39
4.1.3.1. Descriptive statistics and ANOVA result of Mean Minimum 39

and Maximum Temperature
4.1.3.2. Zonal Analysis 42
4.1.3.2.1. Zone1 42
4.1.3.2.2. Zone2 44
4.1.3.2.3. Zone 3a 45
4.1.3.2.4. Zone 3b 47
4.1.3.2.5. Zone 4 48
4.1.3.3. Single station analysis 49
ii
4.1.3.4. Whole study area analysis 53
4.1.3.5. Entire study period analysis 57
4.1.4. Discussion 59
4.2. Rainfall 63
4.2.2. Data collection and analysis 63
4.2.3. Results of rainfall data 64
4.2.3.1. Descriptive statistics and ANOVA results 64
4.2.3.2. Zonal Analysis 66
4.2.3.2.1. Zone1 66
4.2.3.2.2. Zone2 68
4.2.3.2.3 Zone 3a 69
4.2.3.2.4. Zone 3b 71
4.2.3.2.5. Zone 4 72
4.2.3.3. Single station analysis 73
4.2.3.4. Whole study area analysis 75
4.2.3.5. Entire study period analysis 77
4.3. River flows 79
4.3.2. Data Collection and Analysis 80
4.3.3. Results of ensuing rivers 80
4.3.3.1. Descriptive statistics and ANOVA results 80
4.3.3.2. Single Gauge station analysis 81
4.4. Hydro-Meteorological characteristics 82
4.4.1 River basin analysis 83
iii
4.4.1.1. Astor River 83
4.4.1.2. Chitral River 85
4.4.1.3. Gilgit River 87

4.4.1.4. Indus River at Chilas Station 89
4.4.1.5. Indus River at Skardu Station 91
4.4.1.6. Swat River 93
4.4.2 Discussion 95
4.5. Trend Forecasting 97

4.6. Environmental effects of climate change in Pakistan 102
Chapter # 5 Conclusion 105
References 108
iv
LIST OF TABLES
S. No Title Page No
3.1. Ensuing rivers and their geographical information 28
4.1. Descriptive statistics and ANOVA results of minimum and maximum 40

temperatures for all climate zones.
4.2. Descriptive statistics and ANOVA results (three decades) for the whole 41
country
4.3. Descriptive statistics and ANOVA result for zone 1 43
4.4 Duncan test for homogenous subset of Alpha for zone1 44
4.5. Descriptive statistics and ANOVA result for zone2 45
4.6. Duncan test for homogenous subset of Alpha for zone 2 45
4.7. Descriptive statistics and ANOVA result for zone 3a 46
4.8 Duncan test for homogenous subset of Alpha for zone 3a 46
4.9. Descriptive statistics and ANOVA result for zone 3b 47
4.10 Duncan test for homogenous subset of Alpha for zone 3b. 48
4.11 Descriptive statistics and ANOVA result for zone 4. 49
4.12 Duncan test for homogenous subset of Alpha for zone4. 49
4.13 Descriptive statistics and ANOVA result for climate zones. 65
4.14 Descriptive statistics and ANOVA result for the whole 66
4.15 Descriptive statistics and ANOVA result for zone 1 67
4.16 Dunnett T3 test for multiple comparison 67
4.17 Descriptive statistics and ANOVA result for zone2 68
4.18 Dunnett T3 for multiple comparison 69
4.19 Descriptive statistics and ANOVA result for zone3a 70
4.20 Dunnett T3 for multiple comparisons. 70
4.21 Descriptive statistics and ANOVA result for zone 3b 71
4.22 Dunnett T3 for multiple comparisons. 71
v
4.23 Descriptive statistics and ANOVA result for zone4 72
4.24 Dunnett T3 test for multiple comparisons 73
4.25 Descriptive statistics and ANOVA result for ensuing rivers 81
4.26 Single gauge station data analysis for trend determination 82

4.27 Water yield per annum in catchment‘s basin 82
4.28 Hydro-meteorological characteristics of Astor river basin 84
4.29 Hydro-meteorological characteristics of Chitral river basin 86
4.30 Hydro-meteorological characteristics of Gilgit river basin 88
4.31 Hydro-meteorological characteristics of Indus river/ Chilas basin 90
4.32 Hydro-meteorological characteristics of Indus river/ Skardu basin 92
4.33 Hydro-meteorological characteristics of Swat river 94
4.34 Trend forecasting of the climatic variables through different climatic 98
models
vi
LIST OF FIGURES
S.No Title Page No

1.1 Climate change due to natural processes and anthropogenic 2
actions and its environmental effects
3.1. (A): Basic information of the selected climatic zones, stations 26
and ensuing rivers along with their longitudes and latitudes.
4.1 Mean annual temperatures in all climate zones of the study area. 41
(A): Mean annual minimum temperature (B): Mean annual
maximum temperature
4.2 Mean annual minimum ,maximum and mean temperatures in 42
the whole study area
4.3 Map of Pakistan showing significant increase, decrease and no 50
change in the minimum temperature. (A): For period 1976-1990;
(B): For period 1991-2005.
change in the maximum temperature. (A): For period 1976-1990;
(B): For period 1991-2005.
4.5 Whole study area analysis showing significant increase, decrease 54
and no change in the minimum temperature (A): For period
1976-1990; (B): For period 1991-2005.
4.6 Whole study area analysis showing significant increase, decrease 56
and no change in the maximum temperature (A): For period
1976-1990; (B): For period 1991-2005.
4.7 Mean annual temperatures of the entire period 1976-2005. (A): 59
Mean annual minimum temperature (B): Mean annual maximum
temperature. (C): Mean annual temperature
4.8. Mean annual rainfall in all climate zones of the study area. 65
4.9 Mean annual rainfall in the whole study area. 66

change in the mean annual rainfall. (A): For period 1976-1990;
(B): For period 1991-2005..
4.11. Whole study area analysis showing significant increase, 76
decrease and no change in the mean annual rainfall (A): For
period 1976-1990; (B): For period 1991-2005.
4.12. Mean annual rainfall trend for the entire study period from 1976- 77
2005
4.13 Single river discharge on decadal basis 81
4.14 Flow duration curve for Astor river from period 1976-2000 84
vii
4.15 Hydrograph of mean annual discharge of Astor river 85
4.16 Flow duration curve for Chitral river from period 1976-2000 87
4.17 Hydrograph of mean annual discharge of Chitral river 87
4.18 Flow duration curve for Gilgit river from period 1976-2000 89
4.19 Hydrograph of mean annual discharge of Gilgit river 89
4.20 Flow duration curve for Indus river/Chilas from period 1976- 91
2000
4.21 Hydrograph of mean annual discharge of Indus river/Chilas 91
4.22 Flow duration curve for Indus river/Skardu from period 1976- 93
2000
4.23 Hydrograph of mean annual discharge of Indus river/ Skardu 93
4.24 Flow duration curve for Swat river from period 1976-2000 95
4.25 Hydrograph of mean annual discharge of Swat river 95
4.26 Mean minimum temperature trend from 1976-2035. (B): Mean 99-101
maximum temperature trend from 1976-2035. (C): Mean annual
mean temperature trend from 1976-2035. (D): Mean annual
rainfall trend from 1976-2035 with included drought period. (E):
Mean annual rainfall trend from 1976-2035 with excluded
drought period
viii
LIST OF ABBREVIATION
ANOVA Analysis of Variance

ANP Asia Pacific Network
Arc GIS Arc Geographical Information System
ARIMA Autoregressive Integrated Moving Average
CDPC Climate Data Processing Centre
CH4 Methane
CNG Compressed Natural Gas
CO2, Carbon Dioxide
CRCD Climatologically Record from Climate Data Centre
DTR Diurnal temperature change
EPA Environmental Protection Agency
FAO Food and Agriculture Organization
GHG Green House Gases
HCDN Hydro Climatic Data Network
HMSS Harmonic Mean Sample Size
IPCC The Intergovernmental Panel on Climate Change
LCL Lower Critical Limit
N2O Nitrogen oxides
NASA National aeronautics and Space Administration
NEQS National Environmental Quality Standards
NOAA and National Oceanic and Atmospheric Administration
NSIP National Stream flow Information Program
NWFP North West Frontier Province
O3 Ozone
PMD Pakistan Meteorological Department
SPSS Special Package for Social Sciences
SRES Special Report on Emission Scenarios
UCL Upper Critical Limit
UIB Upper Indus Basin
UNFCCC United Nations Framework Convention on Climate Change
UNDP United Nations Development Programme
UNEP United Nations Environment Programme
USGS United States Geological Survey
WAPDA Water and Power Development Authority
WHO World Health Organization
ix
ACKNOWLEDGMENT
First of all, I would like to thank my supervisors Dr. Shafiq-ur-Rehman Professor,

Department of Environmental Sciences University of Peshawar and Mr. Mushtaq Ali
Shah Director in Regional Meteorological Centre Peshawar, under whose inspiring,
affectionate, continuous and easily accessible guidance, this work has been done.
My sincere thanks go to Prof. Dr. Noor Jehan Chairperson, Department of
Environmental Sciences, for providing a very good working environment and facilities in
the Department of Environmental Sciences, University of Peshawar.
I am cordially thankful to Mr. Alam-Zaib Deputy Director in Regional
Meteorological Centre Peshawar and Mr. Shehzad Data Manager for their cooperation
during data management. I also express heartiest thanks to Prof. Dr. Sareer Badshah
Chairman Department of Statistics University of Islamia College for his guidance in
statistical work, helpful discussion and attitude. Finally, I appreciate the cooperation of
Dr. Ayesha S. Abdullah, Dean and Assistance Professor in the Peshawar Medical
College, Dr. Alia Naz (Assistant Professor), Hafsa Habib (Clinical Psychologist), Dr.
Shah Zaman (Cardiologist), Tariq Aziz (Data Manager), Jaffar Wazir (IT Assistance)
and Muhammad Javed for their moral and refreshing support.
Last but not the least it would be incomplete without paying regards to my father
who prayed for my success and bright future and provided financial support throughout
my educational career. I love my parents
x
ABSTRACT
The issue of climate change has emerged very strongly during the last two decades on
global scale in view of its projected implications on the environment of vulnerable states.
Steadily rising temperature and its impacts on the cryosphere, rainfall and river flows are
evident in many regions around the world. Changes in rainfall directly affect hydrology,
water resources, agriculture, ecosystems, healthcare and disaster management sectors.
Seen in the context of Pakistan, a country with very large population, agriculture based
economy and high vulnerability index to natural disasters, it was important to ascertain
climate variability and determine trends of temperature, rainfall and river flows on spatial
and temporal basis. The study was conducted across the country to assess variations in
mean minimum and maximum temperatures and rainfall in different climate zones of
Pakistan. For this purpose datasets comprising 30 years of weather parameters from 1976
to 2005 were acquired from 30 meteorological observatories from different parts of the
country based on the Kööpen classification system. In order to study the impact of
change in temperature and rainfall on river flows, discharge record for the past 25 years
of six relevant rivers was examined.
The results of temperature analysis have indicated that the climatic zone located in high
mountains in north, northwest and west have not been affected much as compared to the
plain and coastal areas where a positive trend has been recognized. An increase of about
0.11°C/decade in mean temperature including 0.1°C/decade in minimum temperature and
0.12°C/decade in maximum temperature has been determined for Pakistan as a whole. On
the one hand rainfall data has revealed an overall decline of -1.18mm/decade but on the
other hand river discharges have increased in Indus and Swat rivers.
Use of statistical models for forecasting future trends indicate an overall increase in the
mean minimum & maximum temperature of the country till 2030 while rainfall shows a
declining trend. Although the average increase of 0.11°C/decade is less than the global
mean, the potential implications are far reaching. It is concluded that climate change
phenomenon may be given key consideration in planning, designing, and implementing
developmental schemes, food security, and disaster risk management sectors to adapt and
mitigate undesirable scenarios.
xi
Chapter # 1
INTRODUCTION
The earth‘s climate has warmed and cooled for millions of years, long before we
appeared on the scene and can be observed from air surface and ocean temperatures
(IPCC, 2007). There is no doubt that the climate is growing warmer currently; indications
of that change are all around us. Rising of sea levels around the world, frequent and
uneven precipitation distribution, flash flooding, prolonged droughts, glacial melting
(Alam, 2009), tropical cyclones, hurricanes, severe dust storms, dry and cold spells etc.
All these adverse consequences are the direct or indirect results of changing climate and
have altered water availability and quality, impact on plantation and land use pattern and
many others environmental impacts as shown in the figure 1.1.
Climate is the pattern of weather parameters averaged over a long term (Mishers,
2004). It means that climate change refers to long term fluctuations in maximum and
minimum temperature, precipitation in its various forms, surface wind, humidity, cloud
type and amount, solar radiation and other elements in earth‘s climate system. Natural
processes such as variation in solar irradiance, change in earth‘s axial tilt, changes in the
ocean circulation and volcanic activity are some of the natural causes responsible for that
(Kondratyev, 1998; Rodo, 2003). Changes in cloud cover, precipitation, soil moisture,
and atmospheric circulation are the main factors responsible for much of the rising trend
differential in all over the globe and cause frequent heat waves, droughts, extreme
precipitation events and related impacts such as wildfires, heat stress, vegetation changes
and sea-level rise (Dai et al., 1999; Przybylak, 2000; Braganza et al., 2004). The climate
system can also be influenced by changes in the concentrations of various gases in the
atmosphere, which affect the earth balance of radiation. These gases are called green
house gases (GHG), including carbon dioxide (CO2), methane (CH4), ozone (O3),
nitrogen oxides (N2O) etc. The observed increase in temperature is mainly due to
combustion of fossil fuels (oil, natural gas, and petrol) for industry and transport, rapid
industrialization, fast growing population, deforestation for agriculture, unplanned
urbanization and infrastructure developments.
1
Figure1.1. Climate change due to natural processes and anthropogenic actions and
its environmental effects
Variation in solar irradiance

Natural climate Change in earth‘s axial tilt
variability Volcanic activity
Ocean circulation
Up growing Population
Anthropogenically Rapid Industrialization
induced climate Increase use of fossil fuel for
variability Transport and Industry
Deforestation for Agriculture
Unplanned Urbanization
Increase number of Energy Resources
Climate Change
and its
Environmental  Global warming
 Sea level rise
Impacts
 Frequent and uneven precipitation pattern
 Melting of glaciers and sea ice
 Increase in frequency and intensity of
Extreme weather events like:
 Frequent flooding
 Prolonged droughts
 Soil erosion and land sliding
 Shortage in freshwater availability
 Altered water quantity and quality
 Contaminations in soil and water bodies
 Threats to natural ecosystem
 Shifting of plants and animals on elevation
 Decrease in crop yields
 Shifts in the growing season
 Loss of biodiversity
 Increase health risks (infections,
malnutrition, mental condition)
 Settlement patterns
2
In IPCC (2001) report it is mentioned that due to the human activities, the earth's
green house effect has been intensifying since the later part of the 20th century and global
climate will remain warmer in the next decades of the 21st century (Karl et al., 2006;
Trenberth et al, 2007). The increased use of fossil fuels has raised the level of CO2 in the
atmosphere, and the destruction of natural vegetation has prevented the environment from
restoring the balance. Level of other GHG including CH4, NO2 and O3 has also been
arising and the net result has been a gradual global warming (IPCC, 2001). Karl and
Trenberth (1999) in their paper described that human induced actions are the dominant
factor in recent global warming and if climate mitigation measures are not adopted at the
right time, the situations will be worse in near future. For this purpose, the international
response to mitigate climate change has been organized under United Nations Framework
Convention on Climate Change (UN/FCCC), and in 1997 the first protocol with
quantitative GHG emission reduction targets was signed (UN/FCCC, 1998). The
Intergovernmental Panel on Climate Change (IPCC) scientifically studied and assessed
all aspects of global climate system and its interaction with atmosphere, land surfaces,
oceans, glaciers, sea ice and ecosystem. Findings of these IPCC assessments have been
arranged in authoritative reports in 1990, 1995, 2001 and 2007, and these reports are the
primary source of the scientific information.
IPCC in its 1st report mentioned that emission of GHG are increasing and not
remaining stable due to increased energy consumption as a result of population growth
and industrialization especially in the developing countries. Further significant increases
in GHG levels are expected, particularly as developing countries become more
industrialized. Based on present warming trends, World Energy Outlook (2010), predicts
a 53% increase in global energy demand by 2030, with 70% of that coming from
developing countries. Fossil fuels are considered to be responsible for this increase, and
the subsequent result will be to rise the temperature. Recent researches find that these
temperature trends are actually much higher than previously thought. From the period of
pre-industrial to the present, the concentration of GHG in the atmosphere is 280-430 ppm
and increasing by approximately 2.5 ppm per year. It is estimated that, if the world‘s
GHG emissions remain at a level of 550 ppm at a minimum, then there is a 63% chance
3
of exceeding the dangerous limit of 2oC temperature increase till 2100 (The Stern Review
on the Economics of Climate Change, British report, and 2006).
The studies of Mann and Jones (2003) have shown that the mean surface
temperature over the past two millennia detecting that late 20th century warmth is
unprecedented and attributed to the anthropogenic forcing of climate (Thorne et al.,
2003). Since in the late 19th century, average global surface temperature has been
recorded with an increase of approximately 0.6°C, with 95% confidence intervals of 0.4
and 0.8°C and future change is expected in the order of 1.4 °C – 5.8 °C over the 21st
century (Folland et al., 2000). This increase has been noted in two periods, from 1910-
1945 and 1976 to the present (Jones et al., 2001). The warming rate from 1976,
(0.17°C/decade), has been slightly larger than the rate of warming during the 1910-1945
(IPCC, 2007). However, changes in climate are already being observed—the last 60 years
were the warmest in the last 1000 years (Wassmann and Dobermann, 2007). The most
recent warming period is faster over land as compared to the oceans warming (Christy et
al., 2000).
In addition, the 1990 was the warmest decade (Palutikof, 2001) and 2005 was the
warmest year followed by, 2010, 2007 2003, 2002, 1998, 1997 (IPCC 2007), since the
start of the global mean temperature record is from 1856 (Jones and Moberg, 2003).The
data of the global temperature from December 2004 through November 2005 show that
2005 has been recorded as the hottest year in comparison with 1998. The year 2005 show
a clear global warming trend of the last several decades by equaling the record warmth of
1998 without the occurrence of El Nino, which in 1998 added extra heat from the ocean
to the earth's surface (Hansen et al., 2006; WMO, 2005). Similarly, recent researches
from Global Climatologically Record from Climate Data Centre and National Oceanic
and Atmospheric Administration (NOAA) and NASA showed that 2010 equal to 2005 as
the earth warmest year on record over the lat 131 years (instrumented monitoring stations
date back to 1880). According to NOAA the since 1876 show an increase in mean
November and December temperatures by 2oC and 1.5oC respectively and in October
rainfall. Mean humidity in December has also been increasing since 1950.
Similarly, observations from all over the globe detect that an increase in global
temperature is considerably, causing seawater to expand and sea level to rise. Rising of
4
sea levels can be observed during the 20th century around the northern and southern
Pacific Oceans, Indian Ocean and Caribbean Sea but the rise would not be geographically
uniform like warming of the land surface (Cruz et al., 2007; Un-Ohrlls, 2009). In 1961
and 2003 global sea level rose at an average rate of 1.8 mm per year but the rate was
more fast at about 3.1 mm per year over the period of 1993 to 2003 (IPCC, 2007, and
WMO, 2010). World Health Organization (WHO) (2000) mentioned in its report that
about 13 million to 94 million people living near the coast are at increase risks of
flooding if sea level rise up to 40 cm. The sea-level rise could cause flash flooding,
increase the salinity of water, effect low lying areas like drown wetlands and coastal
marshes and erode beaches. Coastal areas of China, India, Bangladesh, Africa and the
small island are especially susceptible to increasing salinity of their water resources
(UNEP, 2007).
Furthermore, Arctic summer ice is also melting at an alarming rate than ever
before due to the global warming as mentioned in several studies. A study conducted by
Stroeve et al (2007) found that four out of the five lowest years of sea ice coverage have
occurred since 2002. Another similar work on the decline of Arctic summer sea ice melt
is done by Overpeck et al (2005). They conclude that the Arctic could be completely free
of summer sea ice well before the end of this century or within 1000 years. Records of
past ice-sheet melting indicate that the rate of future melting and the related sea-level rise
could be faster than last century. This recent study found by Otto-Bliesner et al (2006)
that during the last interglacial period approximately 130,000 to 127,000 years ago, sea
level ranged from 4 meters to more than 6 meters higher than today. Climate models
project that the high northern latitudes will remain warmer than they were during the last
interglacial period. Temperatures in the late 21st century would be warm enough to melt
at least large portions of Greenland and West Antarctica. Millions of people would be
vulnerable to flooding and displacement from the resulting sea level rise, and the
economic loss associated with coastal inundation. But it's not just the ice system that is
changing. Changes in the permafrost or the permanently frozen ground also reflect a
warming trend in the Arctic. Recent study of the National Center for Atmospheric
Research's Community Climate System Model (2006) projects that under GHG
scenarios; there will be up to 80 % decline in permafrost at the end of the 21st century. If
5
permafrost melting occurs at a greater scale, it might be result in the rapid release of
CH4 gas, a potent global warming pollutant.
Additionally, melting or advancing of glaciers would also affect the land use
patterns around the glacier. For instance, agriculture lands or herding near the glaciers,
alter the habitat of animals and plants that have become adapted to surviving closer to the
glacier (FAO, 2004). Iqbal (2008) from Pakistan reports that large scale melting of
mountain glaciers and polar ice caps have been occurred with fastest rate, particularly in
the Arctic region. Melting ice and thermal expansion of oceans are the key factors for sea
level rise. In addition, rising sea levels will also lead to salt water contamination of
groundwater supplies, threatening the quality and quantity of freshwater supply. Since
1990, the IPCC has documented such changes as an evidence of global warming.
Meanwhile, work carried out in Kalimanjaro by Kaser, (2004) concludes that
increased air temperature leads to govern the glacial melting in a direct manner. The
effect of global warming on the mountainous areas is most visible from the shrinkage of
glaciers and reduction in reduced snow cover duration (Barry, 2002). There is some
disagreement among the scientists that all glaciers of Himalaya-Karakorum-Hindukush
region are retreating. According to the World Glacier Monitoring Service (2002), most
of the world‘s mountain glaciers have been shrinking for the last 30 years, including the
greater Himalayan as well (Hasnain, 1999; Mastny 2000; Sherestha, 2004).
According to US global change research, the warmer temperatures are likely to

increase evaporation from land surfaces and causing severe rain event. While increasing
severe rain events are responsible to the increase surface runoff. Increases in
precipitation, of 5-1% per decade in the mid/high latitudes of the Northern Hemisphere,
have also been observed since the 20th century (IPCC, 2001). Analysis of heavy rainfall
events indicates a probability of over 90% that a 2-4 % increase in frequency has
occurred during the past 50 years in the Northern Hemisphere (IPCC, 2001). Similarly,
climatologists analyzed temporal and spatial variability in extreme precipitation events
like both floods and droughts have increased. In Southern Africa, Latin America, the
Mediterranean, and some parts of Southern Asia, temperature rise may also induce a
prolonged drought which along with others anthropogenic factors causing deforestation,
6
weathering and soil erosion etc (UNEP, 2007). More intense and longer droughts have
been observed since the 1970s, particularly in tropical and subtropical (IPCC, 2001). One
effect of the rising temperatures that are expected over the next century is that the
atmosphere‘s capacity to hold moisture will go up. For every 1º C increase in
temperature, the water-holding capacity of the atmosphere rises 7%. Increased moisture
in the atmosphere will lead to more intense precipitation events – even when the annual
total amount of precipitation is slightly reduced. Changes in precipitation patterns are
already observable. Over the past century, Eastern North America has gotten wetter,
while southern Africa and the Mediterranean have become drier. In the 21st century, the
Northeastern U.S. is expected to receive more precipitation, while the Southwest is
expected to become even drier (Noah, 2007). Similarly, unpredictable rainfall has been
noted in Northern Europe, some parts of North, South and Central Asia. Cruz et al (2007)
in his studies revealed that most of the regions of Asia like Eastern and the Southern parts
will receive summer as well as intense winter precipitation with a greater probability in
the intense precipitation frequency. In the northern Indian Ocean annual rainfall is likely
to increase in winter season (December, January, and February), and around the Maldives
in summer season (June, July, and August), while decreases occur in Mauritius in the
same months of summer (Mimura et al., 2007).
Furthermore warmth environment has also influenced the forest resources and
alters biomes production and its quality. Firstly, compared with forest and agricultural
lands, temperature variation has greater affect on albedo, soil quality and humidity
variations (Bonan, 2002). Similarly, short and long term climate change is also leading to
change crop management practices and techniques etc.
As a consequence of climate change, warming may also have considerable
impacts on plants carbon nitrogen ratio, biomass production, yields, root morphology,
shoots morphology, soil nutrients uptake etc. High CO2 levels in the atmosphere can
affect plant responses to different limiting factors including water, light, nutrient
availability. For instance, a mild change may result in vegetation stress, rapid plant loss
and desertification in certain circumstances (Rashid, 2008).
Studies have shown that in North Asia an increase of 1°C of average temperature
could increase the duration of the wild fire season by 30 times (Cruz et a., 2007), which
7
could have various adverse consequences on key forest ecosystem functions including
structure and composition of soil, outbreaks of pest, loss of biodiversity, species habitat
quality, and prevalence of diseases. Wildfires in the western United States have been
recorded increasingly over the past few decades, but the extent of these changes had
never been analyzed until recently. Westerling and Bryant (2007) found that there has
been recorded a four time increase in the number of western wildfires and 6.5 times as
much area burned in 2003 as compared to 1987. But the wildfire of 1987 has also been
recorded more than 1970. The length of the wildfire season was observed 78 days. The
average burn duration for large wildfire has observed from 7.5 to 37.1 days over the same
time period. They concluded that wildfire in United States is primarily due to changes in
climate, specifically increased spring and summer temperatures and earlier spring
snowmelt.
In non-temperate climates, climatic warming may enhanced the ability of present

day forests to withstand these changes. According to Siddiqui (2001) a change of 3oC can
lead to an elevation shift of forests of about 500 m, and even though species have the
ability to adapt warming since the last Ice Age 10,000 years ago, the predicted change
(2.5oC for a doubling of CO2) is very high and will exceed their rate of ―migration‖ to
keep up. A rate of change of 0.0100C is a threshold value for many species survival
(Siddiqui, 2001). Cruz et al (2007) write that in Asia, approximately 50% of the region‘s
total biodiversity is at threat and becoming almost extinct due to change in temperature.
He estimated that in China 105 to 1,522 plant species and 5 to 77 vertebrates and about
133 to 2,835 plants and 10 to 213 vertebrates in Indo-Burma would be extinct at the end
of the century.
Further, the main concern in the discussion of climate change is crop yield. At
mid to high latitudes, crop yields may increase for low levels of change in temperature,
but will decline at higher levels of temperature change. Bruce et al (1998) in their studies
on impacts of global climate change on agriculture yields describes that these impacts can
be categorized into direct and indirect impacts. Direct impacts include increase in
temperature which cause shortening or elongation of growing season, change in duration
and intensity of precipitation, water and soil quality and extreme weather events while
8
indirect impacts are sea level rise, rate of soil erosion, and ultra violet radiations. Seshu
and Cady (1984) have estimated a yield decrease of 0.71 t/ha with an increase in
temperature from (18oC) to (19oC), a decrease of 0.41 t/ha with an increase of
temperature from 22oC to 23oC and 0.04t/ha from 27oC to 28oC. There are several
reasons for this variation in response to a 1oC changes in temperature at different
temperature regimes. Among these are the effect of temperature on different stages of
plant growth particularly, sudden sterility of the plant caused by higher temperature
(Yoshida, 1981). Beside the adverse effects of climate change, some of them are likely to
be beneficial. In middle to high latitudes and at high altitudes there may be an increase in
agricultural productivity, depending on crop type, growing season, changes in
temperature regimes and the seasonality of precipitation (Parry, 1990; Watson et al.,
1997). In these areas, temperature increase can enhance crop growth by allowing earlier
planting of crops in the spring, faster maturation and earlier harvesting (Rosenzweig and
Hillel, 1995).
Changing climate will also have significant impacts on the availability of water,
as well as the quality and quantity of water. Increasing trend in temperature and its
association with precipitation can induce stress in hydrological cycle, resulting in dryer
dry seasons and wetter rainy seasons, and subsequently enhanced the risks of more
extreme flash floods and drought. Melting glaciers will increase flood risk during the
rainy season, and reduce the dry-season water supplies to the one-sixth of the World‘s
population (NEF, 2001).
Many of the world‘s countries are already under struggle for the availability of
fresh water supply for drinking, household use, agriculture and industry. Irrigation
demand, industrial pollution and water borne sewerage will lead to put the pressure on
existing water resources and all these will be significantly intensified by climate change.
Due to reduce rainfall and increasing temperatures will further lead to reduce the
availability of water supply including smaller flows in springs, rivers and under ground
levels.
9
The quality of existing water supplies will become a further concern in some
regions of the globe. Water acquires most of its geochemical and biochemical substance
during its cycle from clouds to rivers, through the biosphere, soils and geological layers.
Changes in the amounts or patterns of precipitation will change the route/ residence time
of water in the watershed, thereby affecting its quality. As a result, regardless of quantity,
water could become unsuitable as a resource if newly-acquired qualities make it unfit for
the required use (IPCC, 2007). For example, in areas with relatively high water tables, or
under intensive irrigation, increased evaporation due to higher temperatures will raise the
concentration of dissolved salts. Further, increased flooding could raise water tables to
the point where agrochemicals/ industrial wastes from soil leach into the groundwater
supply. Likewise, higher ocean levels will lead to salt water intrusion in coastal
groundwater supplies, threatening the quality and quantity of freshwater access to large
percentages of populations.
The largest impacts of global warming mainly come from the incidence of the
extreme weather events, such as heavy rainfall, severe storms, heat waves and droughts.
In most parts of the Europe and US, extreme cold waves and record snowfalls were
observed during winter season of 2009/2010 in Europe. In 2008 severe prolonged
drought hit the several states of the South Africa such as Argentina, Uruguay, Paraguay
and southern Brazil causing severe damage to agriculture, livestock and water resources.
For most parts of the South Africa it was observed as one of the driest years on record.
Heavy rainfall in 2010 in China contributed to floods and landslides along with
devastating mud/rockslide in the northwest of the country. Similarly extreme
precipitation events in West Africa have been observed in 2010 year, with the worst
flooding in 50 years (WMO, 2010). But by their very nature, extreme events are hard to
study and predict precisely because they are relatively rare and can be influenced by local
factors, such as terrain and snow pack. Unfortunately, the increase in heavy rains would
not necessarily make more water available for hydropower or irrigation because natural
and engineered reservoirs are greatly affected by severe storms. Moreover, both the
Northwest and the Gulf Coast would also see an increase in the number of dry days
(Diffenbaugh et al, 2005). It is estimated in the Special Report on Emission Scenarios
(SRES, 2000), 120 million to 1.2 billion will experience increased water stress by 2020s
10
and by the 2050s the number will range from 185 to 981 million people. New evidence
on recent trends, particularly, on the increasing tendency in the intensity and frequency of
extreme weather events are more evident in Asia in the 21st century. For instance, a
decline in summer precipitation was observed over the central parts of the arid and semi-
arid Asia and leading to the expansion of deserts and periodic severe water stress
conditions. Increased rainfall intensity, particularly, during the summer monsoon, could
increase flood prone areas in temperate and tropical Asia. In South-East Asia, extreme
weather events associated with El-Nino were reported to be more frequent and intense in
the past 20 years.
Significantly longer heat wave duration has been observed in many countries of
Asia, as indicated by pronounced warming trends and several cases of severe heat waves
(Mukhopadhyay, 1998; Kawahara and Yamazaki, 1999; Zhai et al., 1999; Lal, 2003; Zhai
and Pan, 2003; Ryoo et al., 2004; Batima et al., 2005a; Cruz et al., 2006; Tran et al.,
2005). Generally, the frequency of occurrence of more intense rainfall events in many
parts of Asia has increased, causing severe floods, landslides, and debris and mud flows,
while the number of rainy days and total annual amount of precipitation has decreased
(Zhai et al., 1999; Khan et al., 2000; Shrestha et al., 2000; Izrael and Anokhin, 2001;
Mirza, 2002; Kajiwara et al., 2003; Lal, 2003; Min et al., 2003; Ruosteenoja et al., 2003;
Zhai and Pan, 2003; Gruza and Rankova, 2004; Zhai, 2004). However, there are reports
that the frequency of extreme rainfall in some countries has shown a decreasing tendency
(Manton et al., 2001; Kanai et al., 2004).
Increasing frequency and intensity of droughts in many parts of Asia are

attributed mainly due to rise in temperature, particularly during the summer and drier
months of the year (Webster et al., 1998; Duong, 2000; PAGASA, 2001; Lal, 2002, 2003;
Batima, 2003; Gruza and Rankova, 2004; Natsagdorj et al., 2005). The recent studies
conducted by Fan and Li (2005) indicate that the frequency and intensity of tropical
cyclones originating from the Pacific Ocean have been increased over the last few
decades. While, cyclones originating from the Bay of Bengal and Arabian Sea have been
noted to decrease since 1970 but the intensity has increased (Lal, 2001). In both cases, the
damage caused by intense cyclones has risen significantly in the affected countries,
11
particularly India, China, Philippines, Japan, Vietnam and Cambodia, Iran, Bangladesh
and Tibetan Plateau (PAGASA, 2001; ABI, 2005; GCOS, 2005a, b). About 80% of
Bangladesh is prone to floods and every year at least one-third of the country‘s territory
is affected (Hoque 1995). Over the last 30 years South Asia has experienced more than
65,000 deaths and about a billion people have been affected by floods and landslides.
This accounts for around 33% of all flood events in Asia. The India and Bangladesh
floods of 1998 caused over 2,600 deaths, displaced 25 million people, and caused an
estimated US$ 3.4 billion of damage. In 2007, floods in India, Nepal, and Bangladesh
caused more than 3,400 deaths, affected 30 million people, and caused an estimated US$
5 billion in economic damage. The impact of floods can only rise as the population in
floodplains grows and the value of infrastructure increases (Mandira et al., 2008).
Occurrence of extreme events like outbreaks of flood, prolonged droughts,

decreasing water availability, intense rainfall, tropical cyclones and unexpected cold and
heat waves have been considered the widespread of diarrhea and other infectious diseases
like cholera, Hepatitis, dengue fever etc, in South Asian countries like India, China and
Bangladesh (Hajat et al., 2003; Ebi et al., 2006; Takahashi et al., 2007; Knowlton et al,
2007; Shah, 2008). According to a study by Adams (1997), there are annually about 1200
deaths due to cold and hot weather. The direct and indirect impact of climate change are
mainly responsible for the transmission of vector born infectious diseases, alteration of
water quality and change in the productivity of agro-ecosystems (Haines et al., 2000).
Further, many researches such as (WMO, 2010; Cavallo and Ilan Noy, 2010;
Peduzzi, 2005; Munich, 2002; International Federation of the Red Cross and Red
Crescent Societies, 1999; Erbach and Gaudet ,1998) have been conducted on the
occurrence and intensities of the extreme weather events. According to those reports,
temperature rising and frequent rainfall are the main factors of economic damage in the
world. Mozaharul et al (2007) in his report, suggests that increase in the intensities and
frequencies of these extreme events will further lead to hunger, susceptibility to diseases,
economic losses, human survival and well being of current as well as future generations.
According to Munich report (2002) extreme climatic events took heaviest toll on human
life and high damage costs, resulting in more than 725,000 lives and economic costs of
12
US$ 700 billion worldwide during the second half of the 20th century. The same has been
mentioned by Mirza, (2003) in his report about the economic damage occurring in the
developing countries due to extreme weather events. Around US$ 35 billion are spent
annually during the past decade in the developing countries (Mirza, 2003). Many
scientists from different parts of the world have consensuses that extreme weather events
are due to global warming and it is assumed that occurrence and intensities of extreme
events will increase with the increase in the global temperature as in IPCC report (Shah,
2008). The dramatic increase in extreme events is likely to have significant economic
consequences on the order of hundreds of billions of dollars annually, as estimated by
Ackerman et al, (2008).
Like other countries in the world, recently, abnormal weather behavior caused
extreme events in Pakistan about 40% of the people of Pakistan are highly prone to
frequent multiple disasters with variations in rainfall patterns, storms, floods and drought
(Hussain et al, 2010, Oxfam Report on Climate Change, 2011). According to data
published by Pakistan Meteorological Department:
 In 1992, worst flood in Jhelum River.
 In 1998 to 2001 severe droughts occurred in the southern and central parts of the
country.
 Severe cyclonic storm in 1999, hit the coastal areas of Pakistan and India.
 At least one million people were affected by seasonal monsoon rains in southern
Pakistan in 2003. Heavy rains caused 162 deaths in the Sindh province.
 Heavy rains in the south and snow in the north as well as Kashmir region
triggered flooding and avalanches, killing at least 486 people during second week
of February 2005.
 Heavy rainfall of 1300mm, far above than normal figure of 900mm recorded in
Balakot in the monsoon of 2006.
 Similarly temperature up to 43oC has been recorded for more than a week in
Chitral and other cities of northern areas in June 2005. While minimum
temperature dropped as low as -12oC in Parachinar, Kalam, Malamjaba and Dir
for consecutive ten days in the 1st week of Jan 2007 whereas their normal is -1oC
to -2oC.
13
 Country has faced super flood in 29th July 2010, after heavy monsoon rainfall hit
the Khyber Pakhtunkhwa, Sindh, Punjab and many parts of Balochistan in over
eighty years. In this worst flood, an estimated 2000 people were dead and has
damaged or destroyed over 700,000 homes (Encyclopedia, 2011). In Sindh
Province approximately 26 million people got infected from this super killing
flood. The Pakistan Meteorological Department reported that over 200 mm of rain
fell over a 24-hour period in Khyber Pakhtunkhwa and Punjab. A record-breaking
274 mm rain fell in Peshawar during 24 hours; the previous record was 187 mm
of rain in April 2009.
 Sever dust storm phenomenon occurred frequently causing more than 100Km/h
wind speed in many cities of Pakistan during recent decades.
 Current temperature of January 2012 is recorded -13/-15 oC
1.1. Climate of Pakistan

Climatically, Pakistan‘s climate is varied because of the difference in elevation
from one end of the country to another. Pakistan is situated in the extreme northwestern
part of the Indian subcontinent, between latitude 24oN to 37oN and longitude of 61oE and
75oE. The plains of Pakistan, drained by the river Indus and its tributaries, are surrounded
by many mountain ranges in the north, northwest and the west (Rodo, 2003). The
country has Arabian Sea in the south bordering India in the east and Iran in the south east,
Afghanistan on the north-west and china in the north. High mountains ranges comprise
Himalaya and Karakorum ranges with a small part of the Hindukush range are present in
the north of the country. Pakistan is the only place where these three great mountain
ranges meet. Besides the northern mountains, there are western highlands separated by
Kabul river from the mountainous north and consist of series of dry and lower hills. Some
parts of Balochistan and Sindh provinces constitute deserts (Hussain et al., 2005). The
country occupies a total geographical area of 803943 km2 (Pant and Rupa Kumar, 1997).
Climate variability in Pakistan is due to its unique geography and geology. During
the summer months from April to September the mountainous north is pleasant and
temperate, but the Indus Valley swelters in temperatures of 40ºC or more. In late summer
14
the southern region experiences monsoons, particularly along the coast. In winter the
low-lying areas cool down appreciably to average temperatures of between 10ºC to 25ºC,
while the northern mountains ice up with the air well below freezing (Rodo, 2003). The
dominant component of the climate variations was spatial shifts in the rainfall patterns,
associated with fluctuations in the general circulation of the atmosphere in the Indian
subcontinent (Rodo, 2003), like the monsoon winds that bring rainfall in summer. The
Western Depression originating from the Mediterranean region and entering Pakistan
from the west that brings rainfall in winter. These mid-latitudinal cyclones make a long
land journey and are thus robbed most of the moisture laden air by the time they reach
Pakistan. Other climate controlling factors are the sub-tropical location of the country
that keeps the temperature high particularly in summer and the oceanic influence of the
Arabian Sea that keeps down the temperature at the coast. In the west and north higher
altitudes keep the temperature down throughout the year. During the summer in the
south, a temperature inversion layer is created which does not allow the moisture-laden
air to rise and condensation to take place
The precipitation received by the country can be divided into two main seasons,
winter and monsoon precipitation. The monsoon rainfall enter into Pakistan from east and
north east during the month of July to September. During this duration a good amount of
rainfall is received in the north and north eastern catchments areas of the country. Winter
precipitation (December to March) are mainly received from western disturbances
entering from Iran and Afghanistan. The weather systems entering from Afghanistan are
called the primary western disturbances and cover only the north and north western parts
of the country, whereas those approaching from the Iran are secondary and cover a large
area of the country including Balochistan, Punjab, KPK, Kashmir and northern areas and
sometimes Sindh province (Kazi et al, 1951; FAO, 1987; Khan, 1993 and 2002; Kureshy,
1998; Luo and Lin, 1999). A large amount of snowfall is received in the northern areas,
upper KPK, Kashmir and northern Balochistan and is the main source of water supply for
water reservoirs of the country in dry season. This water received from the snow melt,
and from the seasonal rains plays an important role in the agricultural and socioeconomic
activities of the country. The agricultural of Pakistan is mainly climate dependant and
every area has its own crops and fruits according to its climate. The country‘s most
15
important crops and fruits are grown in winter season in different areas according to its
climate conditions. If there is any abnormality in the usual climate condition the nation
suffers it for the whole year and there is also a huge loss to the economy (Shah, 2008).
However, due to the heterogeneity in the topography and climatic conditions in
Pakistan, the rise in the surface temperature, rainfall and others climate parameters will
not be same in all its regions. In terms of meteorological variable, atmospheric
temperature is observed in every synoptic hour but daily maximum temperature,
minimum temperature, mean temperature its impact on rainfall and discharge of the
ensuing rivers play an important role in assessing the climate and hydrology of a certain
area. Therefore, these variables are considered the main indicators for the determination
of trends in climatic variables. Determination of recent trends over the last three decades,
after comparing with the global temperature trends, indicate that the consequences of
climate change have already occurred in Pakistan.
Although Pakistan has a reasonably good network of observatories, but there are
very few studies on the climate variability. Rasul and Chaudhry (2006) conducted
research on global warming and expected snowline shift along northern mountains of
Pakistan. They found that rising trend in temperature is not due to natural processes but
mainly because of excessive anthropogenic activities. Growing population and
accelerated deforestation which they believed to be the main cause of atmospheric
heating among others. Snow and ice extent is shrinking must faster than ever now, as a
whole, on the Himalayan, Karakorum and Hidukash ranges. Similarly Chaudhry and
Rasual (2007) in their study describe the frequency and intensity of extreme precipitation
events along the foothills of southern slop of Himalayan region. They found that due to
these extreme events, number of landslides and lightening increasing and cause many
causalities and damage to existing infrastructure.
Another attempt has been made by Bhutiyani et al (2007). Their results showed
that the northwest Himalayan region has warmed significantly during the last century at a
rate which is higher than the global average. The rise in temperature in the northwest
Himalayan has been primarily due to rapid increase in both maximum as well as
minimum temperatures. But maximum temperature increasing more rapidly than
minimum temperature. Recently, Shah (2008) attempted to construct an all Pakistan
16
minimum temperature trend, by taking 30 stations, including the hilly regions of the
country parallel to Himalayan, Hindukash and Koh-e-Suleiman ranges. He obtained a
value of 0.4 oC for all Pakistan mean minimum temperature in winter season for two
periods i.e. 1976-1990 to 1991-2005 and found per decade increase 0.133 oC. Therefore
the present study was conducted to achieve the following objectives
1.2. Objectives
To determine the change and variability in minimum and maximum temperatures
and rainfall of the study area.
To evaluate their effects on water discharge in the ensuing rivers.
To compare the results of the study area with the global trends.
To determine the future trends of the said variability and change.
1.3. Justification
Now- a-days climate change has become an important additional stress on
ecosystem like other environmental problems such as population growth, increasing
resource demands, unsustainable development and pollution. All these problems in many
cases may be greater or equal than those of climate change. The impacts of climate
change or GHG on environment are multiple and most significant. Firstly, compared with
forest and agricultural lands, climate has greater affect on albedo, soil quality and
humidity variations (Bonan, 2002). Similarly, short and long term climate change is also
leading to change air humidity, cloud density, precipitation intensity and duration, water
quality and crop management practices and techniques etc. Study of climate change
involves changes in trends of minimum and maximum temperatures, precipitation rate
and water discharge in the rivers, comparing them to global or regional causes.
According to my knowledge, no published data and research is available on climate
change in Pakistan. Research on climate change and variability trends will give us a
valuable insight into how, when and why the study area‘s climate has changed in the past,
how it is currently changing and what will be in the future. Interpreting the past and
present data and its possible predicting models for future climate change and its effects
on the environment.
17
Chapter # 2
LITERATURE REVIEW
The rising trend in temperatures, associated with the changes in precipitation

pattern, enhance the hydrological cycle because as much of the heat at the land surface is
evaporated. Temperature change is one of the major factor, affecting precipitation
distribution, may be in the form of snow, glacier melting, and rainfall which are
considered the major contributors to the ground water source. Many studies have been
reported on the effects of climatic change and variability in a variety of disciplines. Most
of these studies have focused on the increased temperature trends and its impact on
precipitation pattern during the 20th century. This increase trend in temperature has been
evident over a range of precipitation pattern, particularly in the form of heavy and
extreme hydrological events like floods in most parts of the world (Karl and Knight,
1998; Groisman et al, 2004). Increase intensities in precipitation distribution, and
prolonged droughts are likely to be the most important issues in determining the impacts
of climate change (Allen and Ingram 2002). Reviews of relevant researches including
detection of temperature trend, rainfall trend and river discharge ratio are as follows:
2.1. Studies related to the determination of Temperature trend
 Rupa, Kumar and Pant (1994, 1997) found 0.4°C/100 years statistically
significant increase in the surface temperature data of India. According to their
research, the increasing tendency of mean temperature is mainly due to the
increasing trend in maximum temperature, as the minimum temperature trend is
found less during the current century. However, it has also been reported that
there do exist some smaller sub-regions with statistically significant increasing
and decreasing trends.
 Easterling et al (1997) conducted research on monthly and daily dataset of 5400
meteorological stations all around the world by obtaining the data from respective
national metrological center. They found an increase of 0.5oC in the mean
temperatures 0.050 oC in maximum temperature and 0.018° oC in minimum
temperature respectively for the globe during the last century. They found in their
18
research increase in the minimum temperature at a rate faster than the maximum
temperature.
 From Turkey, the study of Mikdat (1997) indicates an increase of 1.1oC during
the analysis period of 1939-1989. The study of the temperature trend is carried out
from 18 stations from all over the Turkey. Statistically no significant trend is
indicated because of downward trend from 1955-1989. Mostly increasing trend
occurred in the spring and winter minimum rather than the maximum temperature
 Shrestha et al (1999) analyzed the temperature data of Nepal during the period
1977-1994. From the study they found 0.06 ºC increase in the mean temperature
per year, mostly in high altitude regions such as Middle Mountain and the High
Himalaya. Significantly downward trends were observed in Terai and Siwalik
regions. Further winter was found warm as compared to other seasons.
 In another study Shrestha et al (2005) found a rising trend of about 0.04 oC in
mean temperature for all over Nepal during the time period of 1975-2005.
 Similarly, Bonsal et al (2001) in his study described that spatial and temporal
characteristics of climatic variables can easily be understand and explain by using
long period time series of reliable and homogenous set of temperature data.
 Houghton et al (2001), in his study revealed that warming in daily minimum
temperature is stronger than that for maximum temperatures. On the basis of these
identifiable changes, they conclude that there would also be changes in extreme
temperature events in the recent future, such as the frequency of days with
extremely low or extremely high temperatures. Their studies have revealed that
there is a significant decrease in days with extremely low daily temperatures but
no significant decease has been observed in days with extremely high
temperatures.
 Yan et al, (2002), investigate a comprehensive change in extreme temperature
over Europe and China. A decrease in the cold extremes and a warming trend in
the minimum temperature over the china during the last few decades were
observed by them.
 In the context of temperature variability, Ahemad et al (2003) found a rise 0.5oC
to 1oC has been observed in the northern arid and western dry mountains and
19
coastal areas of Pakistan. Both minimum and maximum temperatures have been
dropped in the north western mountains while an increased has been noted in the
western parts of the country.
 Zhai and Pan (2003) computed data of daily surface air temperature data for about
200 stations during 1951–1999 in China. Studies were carried out on the basis of
trends assessment and changes in frequencies of some extreme temperature
events. For China as a whole, maximum temperature over 35oC were considered
hot days found with decreasing trends and minimum temperature below 0 oC were
considered frost days recorded with significant upward trends. Meanwhile, rising
trends were observed in the frequencies of warm days and warm nights and
declining trends were found in the frequencies of cool days and cool nights in
China.
 In the study conducted by Russel et al (2004) found an increase of 0.296 °C per
decade in mean temperature for the globe and 0.287°C and 0.296°C per decade
for maximum and minimum temperatures respectively during the time period
from 1950-2004. For their study they selected from all over the globe
approximately 4280 stations for maximum temperature, 4284 for minimum, and
4157 for DTR stations
 Similarly Farooq and Khan (2004) analysis of past climatic data depicts that there
is found temporal as well as spatial change in the rate of temperature variation.
This change will result a greater impact on all aspects of life.
 Sheikh et al (2005) in his final report submitted to Asia Pacific Network (APN)
for Global Change research found an increasing trend in the annual mean
minimum temperature in monsoonal dominated areas of the Pakistan.
 According to Kumar et al (2003) and Yadav et al (2004) western Himalayan
region is showing a different response to global warming they experienced in their
data that there is an increase in diurnal temperature change (DTR) a cooling of
mean temperature in some seasons, possibly result of local forcing factors.
 Nadia and Islam (2006), found in their study for Pakistan that trend in temperature
indices reflects an increase in both maximum and minimum temperature. There is
significant change in minimum temperature than maximum temperature.
20
Percentile shows that the daily minimum temperature will become warmer as
compared to the increase of daily maximum temperature in summer where in
winter the change in maximum temperature is high than threshold value.
2.2. Studies related to the determination of Rainfall trend
 Chaudhary (1994) attempted to construct an all- Pakistan summer monsoon
rainfall series, by taking the area- weighted average of 38 stations, excluding the
hilly regions of the country parallel to the Himalayan mountain range, and
covering about 88% of the total area of the country. Chaudhary obtained a value
of 13.3 cm for all Pakistan mean summer monsoon rainfall, which accounts for
about 58.5% of the annual rainfall, based on the data during 1901-1990.
 Singh and Sontakke (1996) used 34 stations for precipitation and 15 stations for
temperature. The earliest records available are from 1856 for precipitation and
1876 for temperature. They computed the monthly percentage departures from the
long term mean at the available stations and then average all the available months
and stations to obtain the all- Pakistan mean annual rainfall anomaly series for the
period 1856-1993. They obtained a low –frequency variability of all Pakistan
mean annual rainfall, in terms of 31 year moving averages and decadal means of
both mean rainfall and variability.
 Brunetti et al (2001) has aimed to investigate the precipitation pattern (intensity
and extreme events) for about seven stations located in the northeastern Italy for
the period 1920–1998. In their study they found a negative trend in the number of
wet days associated with an increase in the number of heavy rainfall events to
total precipitation
 Liu et al (2005) analyzed heavy precipitation events in China over the studied
period (1960-2000). They found only 2% increase in the total precipitation and 5
% increase in total extreme precipitation events at 95% confidence interval.
 Similarly Zhai et al (2005) assessed trends in total precipitation and frequency of
daily precipitation extremes over china by taking into account the daily
precipitation dataset of 740 stations for the period 1951-2000. Their results
showed no significant trend in total precipitation for China as a whole, but its
intensity and frequency has been noted with significant increased. Meanwhile
21
annual total precipitation has been found significantly decreased over southern
northeast, north China, and over the Sichuan Basin except western and the
southeastern coast.
 In India rainfall trend analysis was carried out by Parthasarathy and Rajeevan
(2006) to examine the trends in Monthly, seasonal and annual rainfall time series
for about 36 meteorological sub-divisions using a large network of about 1476
raingauge stations. Their results showed that rainfall trend is decreasing in June,
July and September time series to annual rainfall while contribution of August
rainfall is increasing in few subdivisions
 Zhang et al (2007) detected the anthropogenic impact on global precipitation.
They showed a latitudinal redistribution of mean global precipitation, indicating
increased precipitation at high latitudes while decrease at lower latitudes and
found significant impacts on ecosystems, agriculture and human health in regions
that are sensitive to changes in precipitation.
 In Pakistan, the IPCC reports (2007) 10–15% decrease in precipitation in the
coastal belt and hyper-arid plains. In Northern Pakistan, there has been an
increase in summer and winter precipitation over the last 40 year.
 Shahid (2010) analyzed the spatial pattern of annual and seasonal rainfall trends
in Bangladesh by selecting 17 stations during the time period 1958-2007. He also
assessed the possible change in wet and dry events in Bangladesh by using
standardized precipitation index method. His result showed a significant increase
in the mean annual and pre-monsoon rainfall and the number of wet months of
Bangladesh. While found decreased in the dry months in the most parts of the
country.
2.3. Collective studies related to the determination of Temperature, Rainfall and
River Discharge trends
 Lettenmaier et al (1994) examine trends in monthly stream flow, as well as
temperature and precipitation, for 1948–1988, finding large increases in cold-
season stream flow in the Northeast and Midwest of the US over that period.
 Gan (1998) found in their results warming trend in more weather stations than are
the drying trends by selecting the 37 weather stations, along with 50 stream flow
22
basins and 13 sets of evapotranspiration. Their data detected that Canadian
Prairies have become warmer and somewhat drier in the last 4–5 decades.
 Arnell (1999a) studied the effect of climate change on water resources for
specific geographic regions in Europe. Arnell (1999b) also investigated the
changes in water resource stress as a result of predicted climate change and
population growth with its socio-economic problems.
 Olsen et al (1999) describe that in the developing countries, variation in climate
change and river discharge can cause devastating floods in both urban and rural
areas, or droughts that create potentially catastrophic food shortages and
frequently serious human health.
 Lins and Slack (1999) focus on streams with multi-decade records that extend
through 1993 and find that many streams have shown significantly larger low and
medium flows, while high flows do not increase as much.
 Furthermore, Vörösmarty et al (2000) argue that changes in mean climate in the
next 25 years will increase the demand for water supply in a greater degree with
increasing population and economic development.
 Bart Nijssen et al (2001) stated that global increases in temperatures has directly
affect the hydrology of the land surface through changes in the accumulation and
melting of snow, as well as in evapotranspiration by plants. Climatic changes are
also predicted due to the result of changes in precipitation amounts, intensities
and patterns.
 Manton et al (2001) have analyzed the trends in extreme daily temperature and
rainfall in Southeast Asia and South Pacific by taking 91 stations from 15
countries including Australia, New Zealand, French Polynesia, New Caledonia,
Indonesia, Japan, Malaysia, Myanmar, New Guinea, Fiji, Samoa, Thailand,
Vietnam, Philippine and Solomon Islands during the period 1961 to 1998.
Significant increases were detected in the annual number of hot days and warm
nights, with significant decreases in the annual number of cool days and cold
nights. For precipitation pattern, the number of rain days has been found
significantly decreased throughout Southeast Asia and the western and central
South Pacific, but found increased in the north of French Polynesia, in Fiji, and at
23
some stations in Australia. Further the total proportion of annual rainfall from
extreme events has been found increased at a majority of stations.
 McCabe and Wolock (2002), focusing on the period 1941–1999, find the same
pattern of increasing minimum and median annual flow, but point out that the
increase took place abruptly around 1970, coincident with an increase in recorded
precipitation.
 Kahya and Kalayci (2004) documented the trend analysis of stream flow in
Turkey. They selected 31 years period of monthly stream flow for 26 basins all
over the country. Basins located in western of the country show significant trend
at (p< 0.05) whereas eastern basins show no trend. Moreover some basins in the
southern Turkey exhibit rising trend.
 Boer and Faqih (2004), conducted research on mean annual temperature and
rainfall data from all over Indonesia. They found in their studies an increasing
trend of approximately 0.3oC and declining trend in rainfall pattern approximately
2% to 3%.
 Small et al (2006), have analyzed trends in stream flow and total annual
precipitation over the drainage of 218 basins across the eastern half of the United
States during the period of 1948-1997 using significance level of (p< 0.05). They
found increasing trend of precipitation in fall season with primarily increase rate
of stream flow and no change has been recorded in flow rate of stream and
precipitation at the basin scale during the other seasons.
 Archer and Fowler (2006) have analyzed Karakoram and Hindu Kush Mountains
of the Upper Indus Basin (UIB) for seasonal and annual trends over the period
1961–2000. Strong contrasts are found between the data of winter and summer
temperatures and between maximum and minimum temperatures. In the data
analysis winter mean and maximum temperature show significant increases while
mean and minimum summer temperatures show downward trend. Decreases of
approximately 20% in summer runoff are recorded in the rivers of Hunza and
Shyok. This is because of 1°C observed downfall in mean summer temperature
since 1961.
24
 Dash et al (2007) in their study describe that climate change is mainly due to
over exploitation of the natural resources such as deforestation, change in land
use pattern, shortage of fresh water sources, destruction of the ecosystems,
environmental pollutions etc. All these factors cause changes on regional sectors
in the short time scale and on the whole global climate in the long run.
 Rose (2007) analyzed a 68 years trend of rainfall and runoff during the period
1938-2005 in a five-state region in the southeastern United States such as
Georgia, South Carolina, North Carolina, Virginia, and Maryland. The average
annual precipitation for this study area was found 1201 mm and the average
runoff was calculated 439 mm. There was observed no significant consistent trend
with respect to runoff, rainfall, and time in the study area. Declining trend in the
rainfall and runoff were found in the recent decade (1996-2005) than the previous
one.
 Krakauer and Fung (2008) analyzed a subset of approximately 1000 United
States geological survey stream gauge stations to calculate the annual stream flow
per unit area since 1920 on a uniform grid over the coterminous (Being of equal
extent or scope or duration) United States. They found no regular increase in
stream flow with time scale except the record of late 1960 in which the
precipitation trend is recorded higher. In their study both precipitation and stream
flow show non significant declining trends in the early 1990, but the overall
analysis supports decreasing trend in stream flow and in precipitation amount as
result of rising temperature and GHG.
 Luce and Holden (2009) have conducted research on distribution of annual stream
flow from 43 Hydro Climatic Data Network (HCDN) gages in the Pacific
Northwest of the United States between 1948 and 2006 using techniques of least
squares and quantile regression. At 25th percentile annual flow record 72%of the
stations showed significant (p < 0.10) declines, with half of the stations exceeding
a 29% decline and a maximum decline of 47% between 1948 and 2006. Very
fewer stations showed statistically significant declines in either median or mean
annual flow, and only five had a significant increase change in the 75th percentile
results trend from the substantially drier period.
25
Chapter # 3
RESEARCH METHODOLOGY
3.1. Data Acquisition

The dataset used in this study consisted of daily minimum and maximum
temperatures and rainfall records obtained from the Climate Data Processing Centre
(CDPC) of Pakistan Meteorological Department (PMD). The dataset spreads over a
period of 30 years (1976-2005) covering the whole country i.e. 30 stations from extreme
north to south and east to west along with some missing periods. The criterion for
selection of these stations is that they cover almost whole of Pakistan. The period of the
study has been chosen as long as possible in function of data availability. Also in this
study , the data set of those stations has been used which were not changed/ displaced in
last 30 years, because some of the meteorological stations were re-established/ relocated
and the change of site may give difference in the data value. The basic geographical
information is presented by Fig (3.1) and their detail is shown in Table (3.1) in Annexure.
Figure 3.1 Basic information of the selected climatic zones, stations and ensuing rivers along with
their longitudes and latitudes.
The climatic variables such as temperature and rainfall is greatly affected by

altitude, location in relation to the coast, rivers, desert, mountainous area and other such
26
geographical features. Therefore, for analysis, it is difficult to consider the whole country
as one region because different stations have different geography and microclimates. For
this purpose the country was divided into five zones, considering their climatic conditions
as defined by the survey of Pakistan based on Kööpen classification of microclimate.
These zones were named zone 1, 2, 3a, 3b, 4, along with their latitude and longitude and
selected meteorological Observety (Fig. 3.1).
3.2. Climate Zones of the Study Area
3.2.1. Zone 1
Zone one comprises those stations having cold climate and high mountains,
situated in the north of Pakistan. These stations are Chitral, Gilgit, Muzaffarabad, Said-u-
Sharif, Skardu, Astor, Dir, Chilas Parachinar and Kakul. These are mostly hill stations
located between 34 N to 38 N in the Himalaya, Hindukash and Koh-e- Sufaid mountain
ranges. The Indus tributaries pass in the north are Chitral river, Astor river, Gilgit river,
Indus river at Chitral, Swat river, Panjgora river, Shingar and Shayok rivers. Ensuing
river at Parachinar is Kurram River and Neelam River at Muzaffarabad station.
3.2.2. Zone 2
This zone has mild cold climate and Sub Mountains, located between 31N to 34
N. The stations are Sialkot, D.I.Khan, Islamabad, Peshawar, Cherat and Lahore. In this
zone Kabul River is present at Peshawar station Ravi River at Lahore and Chenab River
is near to Sialkot. .
3.2.3. Zone 3a
Climate is cold in winters and hot in summers. Most of them are mountainous
stations with high elevations from mean sea level and cover an area between 27N to 32N
and 64 E to 70 E. Stations included in this zone are Quetta, Zhob, Kalat and Khuzdar.
The river present in this zone is Gomal River.
3.2.4. Zone 3b
This is the hottest and dry zone of the country where highest maximum
temperatures are recorded in stations of Sibbi and Jacobabad. The area is almost plain
with some area included in Thar Desert. Stations included are Sibbi, Jacobabad,
Bahawalpure, Khanpur, Multan and Rohri. No river passes near to these stations.
27
3.2.5. Zone 4
Zone 4 is a big zone having many stations and coastal cities, near to Arabian Sea.
The coastal part comprises only a small part of this region and climate above coastal parts
in Balochistan as well as in Sindh province is mostly arid to hyper arid. The selected
stations from this zone are Hyderabad, Karachi, Nawabshah and Jewani.
Spatial and temporal changes in temperature trend and rainfall pattern mainly
cause changes in the flow rate of river discharge. A change can be occurred abruptly (step
change) or gradually (trend) and sometimes may be converted into complex form.
Climate change is often recognized as a progressive trend towards rising. Studies of
temperature, precipitation change are typically complicated by factors such as missing
values, seasonal and other short-term fluctuations or climate variability and by lack of
homogeneity of the data e.g. due to changes in instrument and observation techniques and
location change of the station. The time series used have been chosen according to the
PMD data quality check and in function of the length of the period of records. All time
series used are continuous from 1976 to 2005, as the PMD has estimated the missing
values.
For the present study, river discharge time series are available for northern areas
rivers for six gauge-stations with record length of twenty five years from 1976-2000. The
river discharge data is acquired from Water and Power Development Authority (WAPDA) of
Pakistan as shown in Table (3.1) along with their longitude and latitude and other basic
geographical information.
Table 3.1 Ensuing rivers and their geographical information
River Station Basin Catchment’s area Altitude Latitude Longitude Province Installed
(km2) (m)
Astor Doyan Indus 4040 1550 35o 31' 74o 44’ AJ&K 1974
Chitral Kelash Indus 11396 2810 35o 50’ 71o 48’ KPK
Chitral
Gilgit Alam Indus 25900 2195 35o 45’ 74 o 37’ AJ&K 1966
Bridge
Indus/ Paratap Indus 142709 1100 35o 44’ 74o 37’ AJ&K 1962
Chilas Bridge
Indus/ Kachura Indus 112665 2341 35o 27’ 75 o 25’ AJ&K 1970
Skardu
Swat Kalam Kabul 2020 1921 35o30’ 72o 35‘ KPK 1961
Source: Pakistan Water and Power Development Authority, 1990
28
3.3. Limitation of the data
The reliability and consistency in data quality control is justified in the following
way:
1. This data set of climate variables covers the period of 1976-2005 for the 30 stations,
from all over Pakistan except the dataset of Astor station which has randomly missing
seven years data in minimum temperature record. In the study area there are fourteen
ensuing rivers but the data set available for river flow consists of six rivers of north
areas and covers 25 years record from 1976-2000.
2. The record length of the available data is not sufficient as compare to the other
similar studies conducted for the same purpose. But according to World
Meteorological Organization the ―climate normal is considered for a minimum period
of 30 years‖. However they provided a good basis for assessing variation in climate
variables.
3. Due to complex orography and different microclimate conditions of the country, the
corresponding data is divided into five zones as defined by Kööpen for classification
of different climates.
4. Another reason to divide the country into zones is to find out the homogeneity in the
means for the trend analysis for each zone and then for the country.
5. There was no specific change in the position of any selected observing station during
the considered time period (1976-2005).
6. The reliability of the real time data mainly depends on the measuring device and the
observer. Any change in the data record and equipment up- gradation and failure,
such changes are accounted by PMD and WAPDA.
7. Although meteorological and hydrological data is quality controlled by PMD and
WAPDA bodies respectively. But for the present study, the quality and consistency
of the data is rechecked for missing values, outliers and occurrence of non-realistic
extreme values such as 30th February etc.
8. The reliability of the data is also checked through descriptive statistics, significance
tests and visualization of the data using SPSS and MS Excel.
29
3.4. Data Analysis
The first step of the analysis includes the understanding of the climate variables
not only for time series of each individual station, but also for the zonal or regional
averages. For this purpose, descriptive statistics has been used to determined the
deviation of the 30 years mean annul data from the average values. Statistical indicators
such as standard deviation, standard error, medians, and skewness were calculated for all
five zones and individual stations as well. The Descriptive procedure displays summary
statistics for several variables in a single table and calculates standardized values.
Statistical procedures involved in this study are:
3.4.1. Mean: It is also known as arithmetic average and has been calculated by using the
following formula
Where is mean of climate variable, xi is each independent variable and n is the total
number of variables or in other words it can be defined as, the arithmetic mean is the
sum of all observations divided by the number of observations n. It is the most widely
used measure of central tendency. One problem with using the mean is that it does not
reveal the typical outcome. If there is one outcome that is very far from the rest of the
data, then the mean will be strongly affected by this outcome. Such an outcome is called
outlier.
3.4.2. Outliers: An outlier is an observation whose value is distant from the values of
the majority of observations. It is sometimes more technically defined as ―a value whose
distance from the nearest quartile is greater than 1.5 times the inter-quartile range‖.
Outliers pull the mean in their direction, and have always been carefully examined.
3.4.3. 5% Trimmed Mean: This mean is used for getting rid of the outliers from the data
or remove the 5% extreme values on the top and 5% extreme values on the bottom of the
data. Trimmed mean mainly concerns with outliers of the data; however the median is
used more often since more people understand it.
3.4.4. Median: An alternative measure is the median. The median is the middle value. If
there are even numbers of events, then medians is the average of the two middles. The
30
median is better for describing the typical value. There is an alternative value that also is
resistant to outliers. When there are n observations and these are sorted from smaller to
larger, and then the median is equal to the value with order number (n+1)/2. The median
is equal to the 50th percentile. If the distribution of the data is Normal, then the median is
equal to the arithmetic mean. The median is not sensitive to extreme values or outliers,
and therefore it may be a better measure of central tendency than the arithmetic mean.
3.4.5. Standard Deviation: A measure of how dispersed are the data from the mean
value or in other words variability of the population from which the sample was drawn.
3.4.6. Standard Error: the standard error of the sample mean is calculated by dividing
the standard deviation by the square root of the sample size.
Standard error depends on the sample size increase, as the extent of chance variation is
reduced.
3.4.7. 95% confidence interval (CI) for the mean: This is a range of values which
contains the population mean with a 95% probability or the 95% CI for the mean
represents a statistical uncertainty of the arithmetic mean.
3.4.8. Skewness: A measure of the data's symmetry (Kenney and Keeping, 1962) or in
simple words the situation of the data in terms of its departure from the mean value. If
the data distribution is ―Normal‖ skewness will be equal to zero. A distribution with a
significant positive skewness value has a long right tail and mass of the distribution is
concentrated on the left of the figure. If the left tail is longer, the mass distribution is
concentrated on the right of the figure. This distribution is said to be left skewed. In
other words there are one or more extreme large values.
31
Negatively skewed distribution Normal distribution Positively skewed distribution
or Skewed to the left Symmetrical or Skewed to the right
Skewness <0 Skewness = 0 Skewness > 0
3.4.9. ANOVA test

After the interpretation of descriptive statistics, the reliability and homogeneity of
the data is determined by applying the ANOVA test. For analysis work, the entire period
1976-2005 were divided into three decades extending from 1976 to 1985 (1st decade),
1986 to1995 (2nd decade) and 1996 to 2005(3rd decade), and fifteen years of two groups
from 1976-1990 and 1991-2005 to recognize the trend of variation of each climatic
parameter and their effect on the flow rate of river discharge. The ANOVA test in the
means provides linearity tests and association measures to help in the understanding of
structure and strength of the relationship between the zones and their means. The
ANOVA procedure generates a one-way analysis of variance for a dependent variable by
an independent variable. ANOVA is used to test the hypothesis that several means are
equal. This test is an extension of the two-sample t test. In addition to determining that
differences exist among the means and which means differs from others. For this
purpose, there are two types of tests for comparing means. The first one is contrast tests
and second one is post hoc tests. Contrasts are tests set up before running the experiment
and post hoc tests are run after the experiment has been conducted.
Post hoc tests are designed for situations in which a significant F- test value with
a factor that consists of three or more means and additional exploration of the differences
among means are obtained. They provide specific information on which means are
significantly different from each other. Post hoc tests further contain two types of
analysis i.e. a number of tests for assuming equal variances for example Duncan Multiple
Range test and second tests involved assuming unequal variances like Dunnett T3 test.
Duncan‘s Multiple Range test provides different critical difference values for particular
comparisons of means depending on how adjacent the means are (Kirk, 1982: Winer et
al., 1991). The homogeneity of the data is determined through Levens test, mostly used
32
for the equality of group variances. This test is not dependent on the assumption of
normality (Aguilar, 2003). For example, for assessment of trend in the mean of a series:
The null hypothesis = H0 assumed that there is no change in the mean of a series and
The alternative hypothesis = H1 assumed that the mean is either increasing or decreasing
over time.
It is the probability that a test detects trend when none is present. Thus a 5%
significance level would be considered as strong evidence against the null hypothesis. In
climatic variables the chance of error or assumption (alternate hypothesis) of differences
in group means of two variables would be analyzed by keeping into account the
probability value up to 85%. After determining the fact that the data is homogeneous, for
the dataset of temperature and river discharge analysis, Duncan test is applied. The main
reason for the application of this test is that if in analysis of variance, the null hypothesis
is rejected, i.e., at least two of the group means differ significantly, and give no clear
results about the means of groups that which group means differ and how many group
means differ from each other. For this purpose the Duncan test is designed to answer
these questions. Similarly the Dennett T3 also gives the same results but it is most
appropriate to use when the variances are unequal. Therefore, for analysis of rainfall data
Dunnett t3 is used.
The following procedure is used for the calculation of ANOVA test (Dundas,
2009).
In the first step, for each dependent variable mean is calculated separately for each
station, zone and for country, like:
In the second step grand mean is calculated from all of the means of dependent variables:
33
Third step is carried out by calculating the sample variance for each variable:
Like 3rd step, in the forth step sample variance is calculated for each dependent variable:
Then in next step grand mean is calculated from all of the sample variances:
In the last step F Statistic is calculated:
SPSS automatically calculates the absolute difference between each observation

and the group mean by performing ANOVA test on those differences. It means that
ANOVA test gives a measure of the difference between all means for all combinations of
means.
In this study, sample size (n) is kept 10 in decadal analysis and 15 in two group
analyses. The critical significance level of the test is of at least p= 0.05 with 95% confidence
interval. The principle feature of these methods is that to simply filter the mean annual
temperature and rainfall records and their influence on the increase/decrease of the river
flow rate. In this sense, the use of ANOVA test makes it easy to compare the means of
different periods and to assess the significance of the changes for individual station, zone
and country.
3.5. Trend Visualization
Furthermore, through visualization process, the climate variability and change is
determined with the help of mapping and graphical presentation.
34
3.5.1. Mapping
As a consequence of data visualization, ArcGIS software is used for ten and
fifteen years trend analysis. The main application in ArcGIS is ArcMap, which is used for
all mapping and editing tasks as well as for map-based query and analysis. A map is the
most common view for users to work with geographic information. It's the primary
application in any GIS to work with geographic information. ArcMap represents
geographic information as a collection of layers and other elements in a map view.
Common map elements include the data frame containing map layers for a given extent
plus a scale bar, north arrow, title and a symbol legend.
3.5.2. Graphical Presentations
Graphical presentations have been carried out by plotting the hydrology data into
time series and flow duration curve for the entire study period i.e. 19760- 2000. In this
study, the four annual seasons were used according to their meteorological definition,
whereby summer encompasses the months of June, July and August; fall is September,
October and November; winter is December, January and February; and spring is March,
April and May. The definitions of the seasons used in this study are according to Nayava
(1980) who used this division of seasons for Sri Lanka‗s climate description. The main
seasons experienced on the Indian sub-continent are the Southwest Monsoon (June–
September), post-monsoon season (October–November), winter (December– February)
and the pre-monsoon season (March–May).The basic aim of this analysis is to find out
the particular season in which climate is going towards a warmer trend and to detect the
period (month) where discharge rate of the river is more than the ordinary level.
3.6. Trend forecasting
In the end of the data interpretation, forecasting procedure is used on 30 years
data of minimum, maximum, mean temperatures and rainfall records, to get a trend for
future. For river flow record forecasting procedure is not applied because of short data
record and minimal number of observing gauge stations. According to the National
Stream flow Information Program (NSIP) under the United States Geological Survey
(USGS), river discharge records of at least 30 years, and greater than 50 years, are
essential for assessing trends because climate naturally varies from year-to-year and
35
decade-to-decade. Such short-term natural variations can falsely indicate a long-term
trend.
Expert Modeler in SPSS automatically select the best fit model for the dependent
variables along with model fit descriptive statistics table. The table measures the
statistics for selected models are stationary R-square (R2), root mean square error
(RMSE), mean absolute error (MAE), mean absolute percentage error (MAPE),
maximum absolute error (MaxAE), maximum absolute percentage error (MaxAPE),
normalized Bayesian information criterion (BIC). Residuals: autocorrelation function,
partial autocorrelation function etc.
The application of adequate series of statistical tests on annual minimum,
maximum temperatures, rainfall and river discharge data are to investigate the temporal
as well as spatial changes in hydro-meteorological parameters, either in the form of a
regular trend or in the form of a abrupt change.
36
Chapter # 4
RESULTS AND DISCUSSION
In order to analyze the data and determine trends in temperature, rainfall changes and
their effects on the ensuing river flow, means were computed for the entire period (1976-
2005) and by breaking datasets into 10 years and 15 years, for each station and for the
country as a whole. The reliability of data and homogeneity of means were statistically
tested by applying different statistical tests, using SPSS as mentioned in Chapter 3.
Interpretation of each variable i.e., minimum and maximum temperatures, rainfall and
river flow was performed separately and compared with global results cited in Chapter 2.
In order to determine trends in temperature and rainfall and their influence on the
river discharge rate datasets of all variables were divided into following combinations:
1. 10 Years Trend: Zonal analysis was carried out on decadal scale i.e. 1st decade
(1976-1985), 2nd decade (1986-1995) and 3rd decade (1996-2005) to work out
their mean values and statistical significance by ANOVA test.
2. 15 Years Trend: Single station and whole study area analysis were also
performed on two groups of 15-years period i.e. 1976-1990 and 1991-2005. The
data was visualized through maps, using Arc GIS software.
3. 30 Years Trend: Entire Study period was analyzed by plotting the 30 years data
into real time series graphs
4. Hydro-Meteorological Characteristics: Influence of temperature and rainfall
on ensuing rivers has been determined through hydro-meteorological
characteristic of river basin analysis. Data presentation has been carried out
through graphical presentation into Ms. Excel sheets.
5. Trend Forecasting: Entire study period was analyzed into trend forecasting
models by using SPSS.
4.1. TEMPERATURE
4.1.1. Introduction
Climate change, as one of the most important global environmental challenges

facing humanity, could cause far-reaching environmental effects. The observed increase
37
in temperature has great implications on agriculture, ecosystem, loss of biodiversity, fresh
water supply and increased health risks (Alam, 2009). According to Folland et al (1999)
climate warming due to anthropogenically induced actions is evident more prominently
from minimum temperature as compared to the maximum temperatures in most parts of
the worlds. Some researchers have found in their studies that an increase in daily
minimum temperatures at a rate nearly twice than that of maximum temperatures since
approximately 1950 (IPCC, 2001). Not all regions show greater warming at night, but it
is the dominant signal in many regions. Therefore, all these signals call for the urgency of
monitoring minimum and maximum temperature records (Jones, 2001). Increase in
daytime temperatures much above normal however was not so evident. The magnitude of
change in temperature varies from region to region for instance in North America the east
coast is projected to warm by 2ºC, while in Alaska and North Canada it is estimated to be
10ºC. In addition, summer temperatures in the American Southwest are found to be
significantly rising as compared to North American temperature averages (IPCC, 2001).
In most parts of the world there have been differential changes in daily maximum and
minimum temperature, resulting in both a narrowing of diurnal temperature range (DTR)
and an increase in mean temperature (Karl et al., 1993; Easterling et al., 1997; Jones et
al., 1999). Most regions of Europe, US and Australia have shown increase in the daily
maximum and decrease in the intra-seasonal daily temperature variability except fewer
days with extremely low minimum temperature. No significant increase has been
recorded in days with extreme high maximum temperature (Karl et al., 1995; Collins et
al., 2000).
4.1.2. Data collection and analysis

As mentioned above data comprising daily minimum and maximum temperatures
records from 30 meteorological stations, covering the whole of Pakistan, from 1976 to
2005 has been analyzed for this study. The data were obtained from the Climate Data
Processing Centre (CDPC) of Pakistan Meteorological Department. The stations included
in this study were selected on the basis of their latitudinal position, elevation from sea
level, length of record, completeness and reliability of data so that a synoptic view of the
entire country could be obtained. For determination of trends, averages were computed
38
for 10, 15 and 30 year periods for each zone and the whole the country. The reliability of
data and homogeneity of means were statistically tested by applying ANOVA. In this
sense, the use of different statistical tests makes it easy to compare the means of different
periods and to assess the significance of the changes for each zone and country.
4.1.3. Results
4.1.3.1. Descriptive statistics and ANOVA result of Mean Minimum and Maximum
Temperatures
Monthly and annual means of minimum and maximum temperatures, 5% trimmed
mean, 95% confidence interval of the mean, medians, standard errors and percentile were
carried out by using descriptive statistics for real time data.
Table 4.1 contains the descriptive statistics of minimum and maximum
temperatures of different climate zones of Pakistan. The number of valid cases indicate
the length of data (30 years from 1976-2005). The results indicate that zone1 shows the
lowest value of mean Min and Max temperatures i.e. 9.26 and 22.89oC respectively with
median of 8.84oC and 23.31oC. Zone4 in minimum temperature and zone3b in maximum
temperature shows the highest value of mean i.e. 20.35 and 33.70 oC and median (20.86
and 33.65oC). Similarly, the highest value of standard error is observed for zone3a
(0.35oC) and lowest value for zone3b (0.09oC). Besides, zone2 (0.32oC) has higher value
of standard error in maximum temperature but zone 3b (0.08 oC) shows lower value of
standard error. Percentile at 75th shows larger value for zone4 and zone3b and at 25th
percentile show lower value for minimum temperature. For maximum temperature
percentile shows higher value for zone3b and lower value for zone1. In the data set of
minimum temperature skewness is found negative for zone1, zone2, zone3b and zone4
respectively and only zone3b shows positive value. Similarly for maximum temperature
in zone3a and zone3b, the skewness has been recorded with positive sign while for rest of
the zones it is negative. The ANOVA results for minimum temperature (F= 608.601, p=
0.00) and for maximum temperature (F= 453.944, p= 0.00) are found most significant at
p-value (0.05) level within the zones and between the zones. But it is significantly
different from each other in the homogenous subset of Alpha at 0.05 in the Duncan test,
as shown in the Table (4.1 in Annexure) and Fig. (4.1 A & B).
39
Table. 4.1 Descriptive statistics and ANOVA results of minimum and maximum
temperatures for all climate zones
Minimum temp (F= 608.601, p = 0.00)
Zone 1 Zone 2 Zone 3a Zone 3b Zone 4
Stations 10 6 4 6 4
30 Years Valid/Missing Values 293/7 180/0 120/0 180/0 120/0
Mean All value used 9.26 15.83 10.27 18.92 20.35
5%trimmed mean 9.32 15.84 10.26 18.96 20.37
Median 8.84 15.94 9.66 19.09 20.86
Std. Error 0.20 0.20 0.35 0.09 0.15
95% CI Lower Bound 8.87 15.54 9.58 18.74 20.05
Upper Bound 9.66 16.11 10.97 19.11 20.65
Percentile 25 7.07 14.37 7.55 17.93 18.72
50 8.84 15.94 9.66 19.09 20.86
75 12.29 17.17 13.69 20.03 21.37
Skewness -0.12 -0.20 0.02 -0.33 -0.49
Maximum temp (F= 453.944, p = 0.00)
Stations 10 6 4 6 4
30 Years Valid/Missing values 300/0 180/0 120/0 120/0 120/0
5% trimmed mean 23.04 29.36 25.49 33.70 32.86
Median 23.31 29.78 25.57 33.65 32.87
Std. Error 0.20 0.32 0.24 0.08 0.20
95% CI Lower Bound 22.50 28.57 25.04 33.55 32.46
Upper Bound 23.29 29.85 25.99 33.86 33.25
Percentile 25 21.11 28.52 23.31 32.92 31.00
50 23.31 29.78 25.57 33.65 32.87
75 25.65 31.70 27.69 34.47 34.63
Skewness -0.69 -0.84 0.12 0.06 -0.04
(A)
30
25
20
oC
15
10
0
1976-1985 1986-1995 1996-2005
40
(B)
40
35
30
25
oC
20
15
10
0
1976-1985 1986-1995 1996-2005
Figure: 4.1. Mean annual temperatures in all climate zones of the study area. (A): Mean annual
Minimum temperature; (B) Mean annual Maximum temperature. The error bars indicate the
standard deviation and all zones show significant increase in temperature at p=0.05 level.
Similarly the overall trend for the whole study area (country) is found most
significant at (p<0.05) level between the decades and within decades in mean maximum
and minimum temperatures as shown in the Table (4.2). The trend of minimum
temperature for the whole study area over the three decades is found non-significant
(p>0.1). However, in the Duncan test (Fig. 4.2 & Table. 4.2 in Annexure), mean
minimum temperature of the whole study area is found slightly significant, having
probability (87%).
Table. 4.2 Descriptive statistics and ANOVA results (three decades) for the whole
study area
Whole Study Area 95% CI for Mean
N F-value Sig. Mean Std. Error Lower Upper
Min temp 30 1.47 0.25 14.16 0.08 14.01 14.32
Max temp 30 4.48 0.02 27.99 0.10 27.80 28.19
Mean temp 30 4.06 0.03 21.08 0.07 20.93 21.23
41
35 21.01 o C
30 20.87 o C 21.34 o C
o
27.89 C
25
27.75 o C 28.35 o C
14.12 o C
oC
20
14.03 o C 14.33 o C
15
10
5
Min Temp Mean Temp Max Temp
1976-1985 1986-1995 1996-2005
Figure 4.2. Mean annual Minimum, Maximum and Mean temperatures in the whole study area. The
error bars indicate the standard deviation and all temperatures show significant increase at p= 0.05
level.
4.1.3.2. Zonal Analysis

To examine temperature variations among different zones, the mean monthly data
of minimum and maximum temperatures for all individual stations were averaged to
obtain the zonal annual mean and whole study area mean. All data were compiled into 10
and 15 and in the last for trend forecasting models.
4.1.3.2.1. Zone1
Zone1 comprise ten stations, mostly covering the north parts of the country. These
stations are located between 34oN to 38oN in the Himalayan, Hindukash and Koh-e-
Sufaid mountain ranges. The Analysis of Variance shows mixed results. In the Table
(4.3) Kakul (F = 4.58, p = 0.02) and Parachinar (F= 16.42, p= 0.00) show the most
significant results at alpha of 0.05 and the remaining stations are found non-significant in
mean minimum temperature. On the other hand for maximum temperature Astor, Chilas,
Chitral, Gilgit, Kakul, Muzaffarabad, Parachinar, and Said-u-Sharif are found with most
significant results (p< 0.05). In the Duncan test (Table 4.4) for minimum temperature,
Astor station is found significant having probability significance greater than rejecting
the null hypothesis. Similarly, differences between the decadal means indicate increase in
the 1st decade, decrease in the 2nd and again increase in 3rd decade in all stations except
42
Chital, Gilgit, Parachinar and Skardu stations. The overall ANOVA result for zone1 (F =
0.37, p = 69) is not found significant between the decades and within the decades for
minimum temperature which means null hypothesis is accepted and alternate is rejected.
Standard error computed for the zone in minimum temperature is 0.20 oC. Similarly for
maximum temperature in the Duncan test, Dir (0.07 oC) and Skardu (0.08 oC) stations are
not found with significant trend in the first and third decades. But their probability is
greater than rejecting the null hypothesis. The mean of these stations in the 3rd decade is
quite different from the previous two decades with increasing trend. The ANOVA result
for maximum temperature (F= 0.98, p= 0.38) is found non-significant among the decades
and within the decades.
Table. 4.3 Descriptive statistics and ANOVA result for zone1
Minimum temp (Zone1 F= 0.37, p= 0.69) 95% C I for Mean
Station N F-value Sig. Mean Std. Error Lower Upper
Astor 23 1.66 0.22 3.53 0.24 3.02 4.03
Chilas 30 0.33 0.72 14.54 0.09 14.36 14.72
Chitral 30 1.01 0.38 8.71 0.12 8.47 8.95
Dir 30 0.76 0.48 7.96 0.14 7.67 8.26
Gilgit 30 1.25 0.30 7.72 0.08 7.55 7.89
Kakul 30 4.58 0.02 10.50 0.15 10.20 10.80
M-abad 30 2.65 0.09 13.54 0.09 13.35 13.73
P-Chinar 30 16.42 0.00 7.50 0.49 6.50 8.50
Skardu 30 0.12 0.89 5.15 0.14 4.85 5.44
S-Sharif 30 0.30 0.75 12.15 0.14 11.85 12.45
Maximum temp (Zone 1 F= 0.98, p= 0.38) 95% C I for Mean
Astor 30 3.54 0.04 15.60 0.15 15.30 15.91
Chilas 30 6.68 0.00 26.35 0.14 26.06 26.64
Chitral 30 2.63 0.09 23.54 0.16 23.22 23.86
Dir 30 2.08 0.15 22.87 0.14 22.58 23.16
Gilgit 30 5.01 0.01 24.36 0.13 24.09 24.63
Kakul 30 3.28 0.05 22.61 0.12 22.37 22.84
M-abad 30 3.89 0.03 27.36 0.18 27.00 27.73
P-Chinar 30 4.31 0.02 21.39 0.14 21.11 21.67
Skardu 30 1.84 0.18 18.91 0.18 18.54 19.27
S-Sharif 30 3.34 0.05 25.94 0.16 25.62 26.26
43
Table. 4.4 Duncan test for homogenous subset of Alpha for zone1
Minimum temp (Zone 1) Subset for alpha = 0.05
Station HMSS 1st Decade 2 nd
Decade rd
3 Decade Sig (1) Sig (2)
Astor 7.45 3.90 2.83 3.63 0.1
Chilas 10 14.64 14.46 14.52 0.46
Chitral 10 8.93 8.69 8.52 0.19
Dir 10 8.05 7.72 8.13 0.28
Gilgit 10 7.61 7.90 7.65 0.18
Kakul 10 10.46 10.04 11.01 0.20 0.10
M-abad 10 13.46 13.33 13.82 0.56 0.11
P-Chinar 10 9.47 8.19 4.85 1.00 0.14
Skardu 10 5.25 5.12 5.07 0.66
S-Sharif 10 12.31 12.06 12.08 0.52
Maximum temp (Zone1) Subset for alpha = 0.05
Station HMSS 1st Decade 2nd Decade 3rd Decade Sig (1) Sig (2)
Astor 10 15.71 15.11 15.99 0.09 0.42
Chilas 10 26.82 25.75 26.47 1.00 0.26
Chitral 10 23.34 23.26 24.02 0.06
Dir 10 22.55 22.83 23.23 0.07
Gilgit 10 24.25 23.97 24.86 0.34 1.00
Kakul 10 22.28 22.60 22.95 0.24 0.19
M-abad 10 26.78 27.43 27.88 0.12 0.26
P-Chinar 10 20.90 21.51 21.76 0.05 0.42
Skardu 10 18.47 18.97 19.28 0.08
S-Sharif 10 25.69 25.66 26.47 0.93 1.00
4.1.3.2.2. Zone2
Zone2 contain six stations as shown in the Table (4.5), situated in the plain sub
mountain areas between 31oN and 34oN. Three stations of Zone2 show increasing trend
in the mean minimum temperature. Islamabad (F= 6.01, p= 0.01), Lahore (F=18.69, p=
0.00) and Peshawar (F=18.09, p= 0.00) indicating most significant results at 0.05 while
Cherat and D.I.K is found non-significant. But in Duncan test the probability of D.I.K
station (Sig= 0.14) between the first two decades is greater (85%) for rejecting the null
hypothesis. The overall minimum temperature for zone2 (F= 2.21, p= 0.1) is observed
almost significant with standard error 0.2oC. The significant trend in zone2 for maximum
temperature (F= 0.09, p= 0.91) and for all stations in the zone such as Cherat, DIK,
Islamabad, Lahore, Peshawar and Sialkot are found almost negligible (p>0.05) as
depicted by the Table 4.5. Similarly Duncan test for the mean of decadal differentiation
also shows no significance in the sub homogenous set of Alpha 0.05 in the Table 4.6.
44
Minimum temp (Zone2 F= 2.21, p= 0.1) 95% C I for Mean

Cherat 30 0.48 0.62 12.75 0.12 12.51 12.99
DIK 30 1.35 0.28 16.87 0.16 16.55 17.19
Isl-bad 30 6.07 0.01 14.45 0.11 14.22 14.68
Lahore 30 18.69 0.00 18.49 0.14 18.20 18.78
Peshawar 30 18.09 0.00 16.03 0.14 15.74 16.32
Sialkot 30 2.48 0.10 16.36 0.11 16.14 16.58
Maximum temp (Zone2 F= 0.09, p= 0.91) 95% C I for Mean
Cherat 30 0.88 0.43 20.75 0.14 20.45 21.04
DIK 30 0.98 0.39 31.62 0.13 31.35 31.88
Isl-bad 30 1.44 0.25 28.74 0.14 28.45 29.04
Lahore 30 0.71 0.50 34.89 0.14 34.61 35.17
Peshawar 30 0.48 0.63 29.80 0.12 29.54 30.05
Sialkot 30 0.32 0.73 29.46 0.12 29.21 29.71
Table. 4.6 Duncan test for homogenous subset of Alpha for zone 2
Minimum temp (Zone2 ) Subset for alpha = 0.05
Decade 3 rd
Decade Sig (1) Sig (2) Sig (3)
Cherat 10 12.61 12.75 12.89 0.36
DIK 10 16.53 17.14 16.95 0.14
Isl-bad 10 14.06 14.42 14.88 0.06
Lahore 10 17.85 18.37 19.26 1.00 1.00 1.00
Peshawar 10 15.55 15.70 16.83 0.52 1.00
Sialkot 10 16.43 16.06 16.59 0.14 0.53
Maximum temp (Zone2) Subset for alpha = 0.05
Station HMSS 1st Decade 2nd Decade 3rd Decade Sig (1)
Cherat 10 20.61 20.62 21.02 0.28
DIK 10 31.54 31.45 31.87 0.22
Isl-bad 10 28.62 28.53 29.08 0.15
Lahore 10 34.69 35.10 34.88 0.27
Peshawar 10 29.66 29.76 29.96 0.38
Sialkot 10 29.38 29.39 29.60 0.51
4.1.3.2.3. Zone3a
This zone includes four stations, located in the southwestern part of the country. It
covers an area between 27oN to 32oN and 64oE to 70oE. All four stations are mountainous
with high elevation from mean sea level. Decadal analysis of thirty years time series data
for zone3a (F= 0.06, p= 0.94) shows no significant results as shown in Table 4.7.
45
Similarly Duncan test (Table 4.8) also shows no significant probability for individual
stations in decadal means of the homogenous subset of data. All decadal means are
significantly different from each other and also from the means of the other stations and
show no significant results. Kalat station shows no change in the mean of the first two
decades. The means are almost same but in the 3rd decade the value of the mean is larger
than the 2nd decade for minimum temperature analysis. Quetta and Khuzdar stations
indicate the most significant results in terms of maximum temperature except Zhob
station (p>0.05). The significance values computed for the data of Kalat station are
higher than the critical tabulated value but the probability of the station is quite high
(~87%: Table 4.7). Duncan test for maximum temperature (Table 4.8) also shows
significant probability (0.08 oC) for Kalat in the 1st and 3rd decade. The overall analysis of
variance for zone3a is observed with non-significant result (F= 1.30, p= 0.28).
Table. 4.7 Descriptive statistics and ANOVA result for zone 3a
Minimum temp (Zone 3a F= 0.06, p= 0.94) 95% C I for Mean
Kalat 30 0.33 0.72 5.41 0.23 4.95 5.88
Khuzdar 30 0.01 1.00 14.83 0.29 14.24 15.43
Quetta 30 0.57 0.57 8.41 0.12 8.16 8.66
Zhob 30 0.45 0.64 12.44 0.25 11.94 12.94
Maximum temp (Zone 3a F= 1.30, p=0.28) 95% C I for Mean
Kalat 30 2.30 0.12 22.19 0.15 21.88 22.50
Khuzdar 30 4.70 0.02 28.99 0.18 28.61 29.37
Quetta 30 6.45 0.01 24.88 0.17 24.53 25.23
Zhob 30 0.77 0.47 25.99 0.18 25.62 26.35
Table. 4.8 Duncan test for homogenous subset of Alpha for zone 3a
Minimum temp Zone3a Subset for alpha = 0.05
Station HMSS 1st Decade 2nd Decade 3rd Decade Sig (1)
Kalat 10 5.28 5.28 5.68 0.51
Khuzdar 10 14.82 14.81 14.87 0.94
Quetta 10 8.22 8.48 8.52 0.37
Zhob 10 12.47 12.14 12.72 0.38
Maximum temp Zone3a Subset for alpha = 0.05

Kalat 10 21.99 21.96 22.63 0.08
Khuzdar 10 28.43 28.89 29.65 0.26 0.07
Quetta 10 24.34 24.72 25.58 0.30 1
Zhob 10 25.74 25.93 26.28 0.26
46
4.1.3.2.4. Zone 3b
This zone is the hottest zone of the country where highest maximum temperature
is recorded for Sibbi and Jacobabad. It consists of six stations and all of them indicate
significant increase in temperature. The ANOVA result (Table 4.9) for zone 3b (F= 4.17,
p= 0.02) is found most significant at 0.05 level for minimum temperature. Duncan test
shows a significant increase in the mean of 3rd decade i.e. 1996-2005 for all stations in
this zone for minimum temperature (Table 4.10). All of the decadal means are found
significantly different from each other but not revealed significant results. Furthermore
for maximum temperature, Zone 3b shows most significant result (F= 3.38, p= 0.04).
Two stations, Bahawalpure (0.08oC) and Rohri (0.08oC) are found with significantly
higher value than p=0.05 level as shown in the Table 4.9. Similarly Multan (0.11 oC) and
Sibbi (0.13 oC) exhibit higher significance level with 88% and 86% probability,
respectively, for the decadal comparison of means in the Duncan test (Table 4.10). It
means that alternate hypothesis can be rejected for all these four stations, because of their
greater probability value.
Table. 4.9 Descriptive statistics and ANOVA result for zone 3b
Minimum temp (Zone 3b F=4.17, p= 0.02) 95% C I for Mean
B-Pur 30 2.83 0.08 18.22 0.13 17.95 18.50
Jaco-bad 30 3.31 0.05 20.11 0.08 19.93 20.28
Khanpur 30 10.20 0.00 17.23 0.14 16.95 17.52
Multan 30 4.57 0.02 18.22 0.09 18.04 18.41
Rohri 30 3.66 0.04 19.78 0.11 19.55 20.00
Sibbi 30 5.80 0.01 19.98 0.11 19.76 20.20
Maximum temp (Zone3bF= 3.38, p= 0.04) 95% C I for Mean
B-Pur 30 2.85 0.08 32.88 0.13 32.61 33.16
Jaco-bad 30 0.83 0.45 34.21 0.15 33.90 34.53
Khanpur 30 0.30 0.74 33.46 0.13 33.19 33.73
Multan 30 1.57 0.23 32.65 0.12 32.39 32.90
Rohri 30 2.81 0.08 34.18 0.11 33.95 34.42
Sibbi 30 1.46 0.25 34.82 0.14 34.54 35.10
47
Table. 4.10 Duncan test for homogenous subset of Alpha for zone 3b
Minimum temp Zone 3b Subset for alpha = 0.05
Decade 3 rd
Decade Sig (1) Sig (2)
B-Pur 10 17.85 18.23 18.59 0.23 0.26
Jaco-bad 10 20.04 19.90 20.38 0.47 0.09
Khanpur 10 16.76 17.04 17.90 0.29 1.00
Multan 10 17.93 18.21 18.53 0.16 0.13
Rohri 10 19.94 19.39 20.00 1.00 0.82
Sibbi 10 19.93 19.62 20.38 0.17 0.06
Maximum temp Zone 3b Subset for alpha = 0.05
B-Pur 10 32.72 32.63 33.30 0.78 0.07
Jaco-bad 10 33.95 34.25 34.44 0.24
Khanpur 10 33.42 33.35 33.60 0.49
Multan 10 32.61 32.40 32.93 0.11
Rohri 10 34.16 33.89 34.51 0.31 0.19
Sibbi 10 34.60 34.73 35.14 0.13
4.1.3.2.5. Zone 4
Zone 4 contains four stations, out of which only two stations Hyderabad and
Karachi shows significant trend in minimum temperature. The overall trend of this zone
(F= 3.80, p= 0.03) shows increase among three decades (Table 4.11). In the Duncan test
the decadal means indicate significantly increasing trend except for Jewani station, which
show downward trend in the 3rd decade (1996-2005) as shown in the Table 4.12. In terms
of maximum temperature, three out of four stations reveal significant trend except Jewani
(Table 4.11) but in homogenous subset of alpha Jewani shows significant change (p=85
%) in the first two decades (Table 4.12). The mean of the 3rd decade is slightly lower than
the previous decadal mean. However the significant p –value for zone is found slightly
greater (p= 84%).
48
Table. 4.11 Descriptive statistics and ANOVA result for zone 4
Minimum temp (Zone 4 F= 3.80, p= 0.03) 95% C I for Mean
H-abad 30 3.86 0.03 21.54 0.14 21.25 21.82
Jewani 30 0.91 0.41 20.92 0.12 20.67 21.18
Karachi 30 13.58 0.00 20.73 0.34 20.03 21.42
N-Shah 30 2.02 0.15 18.21 0.11 17.99 18.42
Maximum temp (Zone 4 F= 1.88, p= 0.16) 95% C I for Mean
H-abad 30 5.08 0.01 33.84 0.14 33.56 34.12
Jewani 30 1.64 0.21 30.10 0.14 29.82 30.38
Karachi 30 16.83 0.00 31.94 0.13 31.67 32.21
N-Shah 30 6.70 0.00 35.55 0.14 35.26 35.83
Table. 4.12 Duncan test for homogenous subset of Alpha for zone4.
Minimum temp Zone4 Subset for alpha = 0.05
Station HMSS 1st Decade 2nd Decade 3rd Decade Sig (1) Sig (2) Sig (3)
H-abad 10 21.29 21.28 22.04 0.98 1.00
Jewani 10 21.07 21.02 20.69 0.25
Karachi 10 19.12 20.77 22.29 1.00 1.00 1.00
N-Shah 10 18.01 18.13 18.49 0.08
Maximum temp Zone4 Subset for alpha = 0.05
Station HMSS 1st Decade 2nd Decade rd
3 Decade Sig (1) Sig (2)
H-abad 10 33.31 33.97 34.24 1.00 0.37
Jewani 10 29.75 30.28 30.27 0.14
Karachi 10 31.20 32.26 32.37 1.00 0.63
N-Shah 10 35.10 35.41 36.13 0.29 1.00
4.1.3.3. Single Station Analysis

Although single station analysis of temperature on the basis of 15-years period
suggests differences in the two time periods (1976-1990 to 1991-2005), but the overall
trends for minimum temperature of all seasons were found increasing. Out of thirty
stations, six stations were observed with great increase in minimum temperature during
the 1991-2005 period as shown in Fig.4.3 (B) compared to 1976-1990 period as shown in
Fig. 4.3 (A) and their significant values are shown in Table 4.3 (Annexure).
49
(A)
(B)
Figure 4.3: Map of Pakistan showing significant increase, decrease and no change in the minimum
temperature. (A): For period 1976-1990; (B): For period 1991-2005.
50
In the above figure six out of thirty stations show an increase of significance.
These stations are Islamabad (0.55oC), Peshawar (0.67oC), Karachi (2.09oC), Lahore
(1.04oC), Multan (0.40oC) and Khanpur (0.63 oC). Except Khanpur, all of these stations
are big and congested cities of the country; located in different zones i.e. zone2, zone3b
and zone4, having large populations, big industrial estates, high environmental pollution
and less cultivated areas. Khanpur (0.63oC), Sibbi (0.36oC), Jacobabad (0.32oC) and
Kalat (0.76oC) are relatively small cities of Pakistan with less urbanization and
population effects.
It was also observed that Astor, Chitral, Saidu Sharif, Skardu, Kakul and Khuzdar
indicate a significantly less drop of temperature during the 2nd 15-year period (1990-
2005) but their significant values are larger than alpha (p>0.05) therefore, no change in
the trend is considered. Two stations namely Parachinar and Jewani have been observed
with significantly decreasing trend i.e. -4.09 oC and -0.46 oC respectively, while rest of
the stations including Astor, Skardu, Gilgit, Chitral, Saidu Sharif, Dir, Chilas, Kakul,
Cherat, Sialkot, Quetta, Rohri, Hyderabad, Nawabshah, Khuzdar, Muzaffarabad, Zhob,
D.I. Khan and Bahawalpure have shown no significant change. The intensity of the colors
on the map (Fig. 4.3) from high to low shows the range of minimum temperature from
3.40 - 21.77 oC. Parachinar and Jewani are situated in different zones (zone1 and zone4)
and completely different in terms of their latitude, longitudes and atmospheric conditions.
Jewani is a coastal area situated at comparatively low altitude in the southwest bordering
the Arabian Sea while Parachinar is a town situated in the Kurram Agency (Federally
Administrated Tribal Area) in northwestern part of the country in Koh-e-Sufaid mountain
range along Pak-Afghan border. The sudden drop of temperature in both stations
indicates their abnormal behavior. Hence, further research is needed to investigate other
meteorological variables in detail.
Similarly for the mean maximum temperature, Fig.4.4 (A) depicts a greater rising
trend for Karachi (0.82oC), Quetta (0.67 oC) and Nawabshah (0.57 oC) while Parachinar,
Kalat, Hyderabad, Khuzdar and Zhob show variation from 0.5 oC to 0.7 oC in the second
15-years period i.e. 1991-2005 as compared to 1976-1990 as shown in Fig.4.4 (B). The
intensity of the map color (Fig.4.4 A) from high to low shows range from 35.25 oC -15.47
51
o
C. Those stations which show significant increase are located in different parts of the
country and included in different zones mainly from zone3a, zone4 and zone1. Table 4.4
(Annexure) lists all four stations of zone3a (Quetta, Zhob, Khuzdar and Kalat) located in
southwestern part of the country show maximum increase in temperature and display
significant trend. Quetta is the capital city of Balochistan province and considered as one
of the atmospherically polluted city while Kalat, Khuzdar and Zhob are mountainous
towns of the province with less population and environmental pollution. Karachi and
Hyderabad (zone4) are big urban centers with heavy industry and transport vehicles. The
remaining stations are observed with variation in temperature from 0.1 to 0.5 oC and come
under the category of non significant change because of higher p-value. Only Chilas
station shows downfall in temperature by about -0.5 oC during the first half of the study
period (1976-1990). The temperature range from 35.82 - 15.76 oC in Fig.4.4 (B) shows
the intensity of the maximum temperature from high to low.
(A)
52
(B)
Figure 4.4: Map of Pakistan showing significant increase, decrease and no change in the maximum
temperature. (A): For period 1976-1990; (B): For period 1991-2005.
4.1.3.4. Whole Study Area Analysis

Analysis of the minimum temperature spatially and temporally across the country
has been summarized, by dividing the study area into distinct climatic zones and study
period into two 15-years intervals (Table 4.5 in Annexure & Fig. 4.5). A general
progressive increase in mean minimum temperature with inter-zonal variations can be
easily recognized across the country, except zone1, from the first half of study period
(1976-1990) to the second half (1991-2005). Figures 4.5 (A & B) show zonal means of
minimum temperature ranging from 9.43 - 20.04 oC during the first half to 9.10 - 20.65
o
C during the second half respectively. Zone 4 clearly exhibits the highest increase of
0.61 oC among other zones.
53
(A)
(B)
Figure 4.5: Whole study area analysis showing significant increase, decrease and no change in
minimum temperature. (A): For period 1976-1990; (B): For period 1991-2005.
54
The overall change observed for the whole study area is 0.29 oC from the first half
to second half of the study period and works out to 0.10 oC rise per decade in the
minimum temperature. The variations observed in zonal analysis are most significant
(zone1 p= 0.00, zone2 p= 0.00, zone3b p= 0.047 and zone4 p= 0.048) for all zones except
zone3a (p= 0.635> 0.05), which is non-significant. Although the variations are different
but the overall trend is increasing. The overall change for the study area in two time
periods of fifteen years is not found significant (F= 0.408, p= 0.523). From the above
analysis it can be inferred that plain areas of the country show a clear warming trend
during the period 1991-2005 as compared to 1976-1990. It is also observed that
temperature variations in northern parts of the country from Himalayan and Hindukush
mountain ranges (zone1) are found negligible. Similarly in the south and southwest
(zone3a) of the country variations are negligible but the southeast (zone3b) shows an
increasing trend. Both these zones lie in the same latitudinal window but zone3a lies in
mountainous region having moderate to cold climate, whereas zone3b lies near the Thar
Desert, having hot climate.
In similar manner, the maximum daily temperatures of the study area have also
registered a temporal increase from the first half to the second half of the study period
(Table 4.5 in Annexure). Figure 4.6 (A & B) shows that zones 3a and 4 exhibit more
increase (0.62oC & 0.56oC) as compared to other zones of the study area. The variations
observed in zonal analysis are also found most significant i.e. p= 0.012 and p= 0.011 for
zone3a and zone 4 respectively
55
(A)
(B)
Figure. 4.6: Whole study area analysis showing significant increase, decrease and no change in
maximum temperature. (A): For period 1976-1990; (B): For period 1991-2005.
56
The overall increase in maximum temperature for the whole study area from first
half to the second half works out to 0.36oC and 0.12oC per decade. The result of ANOVA
test is found non-significant (F= 0.762, p= 0.383).
From the above analysis it is concluded that in the decadal analysis zone3b is
found statistically most significant in comparison to other zones. It was observed that
stations included in zone3a show a greater warming trend. Warming trend seems to be
more prominent in southwestern parts of the country although other parts of the country
also show a gradual increase except the central plains and high mountain areas in the
north.
4.1.3.5. Entire Study Period Analysis
The time series of 30 years annual mean minimum and maximum temperatures
clearly showed that the overall change in temperature is not found uniform during the
study period. Figure 4.7 (A & B) is the combine result of minimum and maximum
temperatures showing the regular increase in the last decade with some significant
decreasing fluctuations in 1976 to 1990. Overall results in Fig. 4.7 (C) show that mean
temperature increased with fluctuations but have the significant trends of rise in
temperature during the whole study period but this change in two decade is not more
significant than the last decade. The time series plot also shows the highest mean annual
Min and Max temperatures for the year 2003 and 2004 (14.8 oC & 21.7 oC) respectively
whereas the lowest mean annual Min and Max temperatures of 13.4 oC and 20.4 oC were
recorded in 1982-1983 and 1997 respectively. The trend line generated by the software
through the scattered data point also indicates a clear rising trend for temperatures from
1976-2005.
57
(A)
Time series of Mean Annual Min Temp for the whole study area 1976-2005
15.0
14.8
14.6
14.4
14.2
oC
14.0
13.8
13.6
13.4
13.2
13.0
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
Time series
(B)
Time series of Mean Annual Max Temp of the whole study area 1976-2005
29.0
28.8
28.6
28.4
28.2
oC
28.0
27.8
27.6
27.4
27.2
27.0
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
Time series
58
(C)
Time series of Mean Annual Temp of the whole study area 1976-2005
21.8
21.6
21.4
21.2
21.0
oC
20.8
20.6
20.4
20.2
20.0
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
Figure. 4.7: Mean annual temperatures of the entire period 1976-2005. (A): Mean annual Min temp;
(B): Mean annual Max temp; (C): Mean annual Mean temp.
4.1.4. Discussion
The study has revealed a mixture of results. The descriptive analysis shows that
zone1 has the lowest means (9.26oC & 22.89oC) and medians (8.84oC & 23.31oC) both in
minimum and maximum temperatures respectively while zone4 shows the highest mean
values (20.35 & 20.86 oC) followed by zone3b (mean, 33.70 oC and medians, 33.65 oC).
In the zonal analysis of the country, other statistics involved in the descriptive procedure
are quite different from each other and show significant means and medians.
Furthermore, for determination of temporal and spatial variation in temperatures
of the study area, analysis of variance (ANOVA) test was applied to two combinations of
the dataset. First ANOVA test was applied to the three decadal datasets and secondly to
the two 15-years periods. In decadal analysis ANOVA test was found significantly
different between all the zones at significant level (p=0.05). The p-values of minimum
temperature, computed for zone3b and zone4 are found most significant (p<0.05) but for
zone1, zone2 and zone3a, they are non-significant because of higher p-values.
Orogrphically speaking zone1 and zone3a are mostly mountainous areas lying in north
59
and west of the country respectively. These two zones play an important role in the water
availability of the country. Several tributaries of the two big rivers i.e. Indus and Kabul
originate from areas included in zone1. These two rivers are fed by the snow and glacial
melt derived from zone1 as most stations of these zones receive a lot of snowfall in
winter and the pre- monsoon dry season (April-May). The freshwater supply derived
from snow and ice melt serves as a critical resource for irrigation and hydropower
generation in the country. Forests present in these two zones are the main source of
timber supply and medicinal plants of the country (Shah, 2008).
Similarly the zonal analysis for maximum temperature indicates that zone3b is
most significant but zone4 also cannot be ignored as it has higher probability (84%) or p-
value. The remaining areas such as zones 1, 2, and 3a, show similar results as were
observed for the minimum temperature. Zone2 and zone3b are plain areas of the country
and play a vital role in the agriculture sector of the country, especially for production of
cotton, rice, wheat, sugarcane and mangoes. These zones contain large cities like Lahore,
Multan, Sialkot, Faisalabad, Bahawalpure and Peshawar. These cities are highly polluted
because of congestion traffic, industry and fewer plantations. Zone4 mainly constitutes
the coastal belt of the Arabian Sea and most southerly part of the country. It is most
significant for minimum temperature and slightly significant for maximum temperature.
This is largely because of the Karachi station, which has been recorded with significant
increasing trend in temperature.
Analysis of minimum temperature carried out on the basis of two 15-years periods
indicate that with the exception of zone3a all zones are most significant having smaller p-
values i.e. zone1 (p= 0.00), zone2 (p= 0.00), zone3b (p= 0.047) and zone4 (p= 0.048) but
zone3a (p= 0.635> 0.05) is non-significant. Similarly in terms of maximum temperature,
only zone3b (p= 0.04) can be considered most significant while other four zones show a
non-significant trend. Single station analysis shows that Islamabad (0.55oC), Peshawar
(0.67oC), Karachi (2.09oC), Lahore (1.04oC), Multan (0.40oC) and Khanpur (0.63oC)
recorded a significant increase in minimum temperature while Karachi (0.82oC), Quetta
(0.67 oC) and Nawabshah (0.57 oC) recorded a rising trend in maximum temperature.
These stations are located in different regions of the country, mainly from zone2, zone3a,
zone3b and zone4. Most of these are big cities having diverse microclimatic conditions,
60
huge population, heavy traffic, industrial estates and scarce vegetation. All these
environmental factors could be the causes of rising trend of temperatures in these cities
(For details see references, Munich Report, 2004 and Shirazi and Ali. 2009) and require
further research works which are not possible in the present study. Khanpur (0.63oC),
Sibbi (0.36oC), Jacobabad (0.32oC) and Kalat (0.76oC) are relatively smaller cities of
Pakistan with less urbanization, population, transportation and industrial effects. Two
stations, Parachinar and Jewani, from zone1 and zone4 have been observed with
significant decrease in the temperature i.e. -4.09oC and -0.46oC, respectively albeit both
stations are located in entirely different regions, latitude, longitude and atmospheric
conditions. Jewani is situated in the coastal area at lower latitude in the south bordering
of Arabian Sea while Parachinar – the capital of Kurram Agency in Federally
Administrated Tribal Area (FATA) is a valley in high mountainous range of Koh-e-
Sufaid along Pak-Afghan border in the northwest of the country. In both stations, the
abrupt decline of minimum temperature during the second half (1990-2005) indicates
their abnormal behavior and requires further work to investigate other meteorological
variables beyond the scope of the present study. The remaining stations such as Astor,
Chitral, Said-u-Sharif, Skardu, Kakul, Khuzdar also show a certain but relatively smaller
drop in the minimum temperature during the 2nd half of study period. However, their
significance values are larger than the critical value of (p=0.05) therefore, no change in
the trend is considered. Similarly Gilgit, Dir, Chilas, Cherat, Sialkot, Quetta, Rohri,
Hyderabad, Nawabshah, Muzaffarabad, Zhob, D.I.Khan and Bahawalpure have shown no
significant change in the minimum temperature. While Chilas (-0.47), Kalat (0.55 oC),
Khuzdar (0.66 oC) and Zhob (0.60 oC) from zone1 and zone3a were noted with slightly
higher value of significance (p> 0.05), but the probability for rejecting the null hypothesis
is greater (90%). These stations are small mountainous towns, located along western
border of the country. No station was observed with significant decreasing trend in the
maximum temperature except Chilas town located in the Gilgit Baltistan province.
Although variations were noticed among the five zones but generally the overall
trend was found increasing. The p-value computed for the whole study area trend shows a
decrease in the mean minimum temperature of zone1 (-0.33 oC) during 1991-2005 period,
which looks anomalous when compared with the increasing trend of other zones. A
61
similar decreasing trend has been advocated by other workers for the western Himalaya
(Kumar et al. 1994; Yadav et al. 2004). The overall increase in minimum temperature of
the country is 0.29 oC from 1976-90 to 1991-05 periods which can be inferred as 0.10 oC
per decade. The p-value for maximum temperature of the country is significant (p<0.05)
for 0.36 oC increase between the two consecutive 15-years time periods and 0.12 oC per
decade. Zone3a (0.62 oC) and zone4 (0.55 oC) have recorded highest increase in the
maximum temperature.
It is evident from this study that all climatic zones of the study area show a
consistently rising trend in the minimum and maximum temperatures except certain high
elevation areas situated in the north, northwest and west of the country e.g. zone1 and
zone3a. However, zone2, zone3b and zone4, situated in the plains, southwest and coastal
areas show a rising trend. This study has determined an increase of 0.1oC per decade in
mean minimum temperature, 0.12 oC per decade in mean maximum temperature and 0.11
o
C in mean temperature. These values are higher than the global mean value of 0.018oC
for minimum temperature and 0.050 oC for maximum temperature (Easterling et al. 1997)
but lower than 0.296 oC, 0.287°C and 0.296°C per decade for mean, maximum and
minimum temperatures respectively as shown by Russel et al (2004).
Afzaal et al (2009) have also reported a sharp increase of 0.06°C per decade and a
cumulative rise of 0.64°C over the period 1901-2007 by merging datasets of the Climatic
Research Unit (CRU) of University of East Anglia and Pakistan Meteorological
Department real time series (PMD).
Dash et al (2007) have determined an increase of 0.25 oC/decade in the mean
annual minimum temperature in India for post monsoon and winter seasons. Likewise
Shrestha et al (2005) have found an increase of 0.04oC from 1975 - 2005 in the mean
temperature of Nepal. Similarly Islam et al (2007) have observed increase of 0.61oC for
Bangladesh from 1961-1990 while comparing observational data with PRECIS modeling
data. In Turkey, Mikdat (1997) has indicated an increase of 1.1oC during their study
period from 1939 to 1989. A similar analysis was conducted by Zhou et al (2004) for
southeast China for period 1979-1998 where they show an increase of 0.6oC per decade
of minimum temperature, a rate faster than that of maximum temperature. Similarly for
other parts of the world an increase of 0.3 to 0.6oC for different seasons has been
62
observed by Russel et al (2004). It means the results of the rising temperature trends in
Pakistan are in harmony with other published works in south Asia as well globally.
4.2. RAINFALL
4.2.1. Introduction
Like the rising trend of global temperature in 21st century, no clear evidence with
regard to long-term change in the global precipitation has been found (IPCC, 2001; Karl
and Trenberth, 2003; Allen and Ingram, 2002). Recent studies show a rising trend of only
9mm, which is approximately 0.98% per decade for 20th century (New et al. 2001).
However, the above mentioned increase has been observed over high latitude areas and
tropical oceans with a decrease over tropical lands (New et al., 2001; Kumar et al., 2003;
Bosilovich et al., 2005). It is obvious that climate warming leads to increased moisture-
holding capacity of air and cause changes in the characteristics of precipitation events,
including their frequency, intensity, amount and duration Karl et al., 1995; Karl and
Knight, 1998; Trenberth, 1999; Treydte et al, 2006). Other evidence for increasing
precipitation rates can be found in Japan (Iwashima and Yamamoto, 1993), China (Zhai
et al., 1999), Australia (Suppiah and Hennessy, 1996), and South Africa (Mason et al.,
1999).
According to the IPCC (2001) report, it is estimated that although total rainfall
increase may be smaller (1.0 - 3.4%) but its intensity was slightly larger (Allen and
Ingram, 2002). Recent studies, based on climate models and past observational records,
predict a future increase in droughts in the south of Europe as a result of increased evapo-
transpiration and a relatively slow decrease of rainfall and precipitation frequency
(Kostopoulou and Jones, 2005; Vicente-Serrano and Cuadrat-Prats, 2007).
4.2.2. Data Collection and Analysis

In order to analyze spatial and temporal variations in the mean rainfall of study
area similar statistical treatment of the data was likewise carried out as mentioned in the
previous section on Temperature analysis.
63
4.2.3. Results
4.2.3.1. Descriptive statistics and ANOVA results
Mean monthly and mean annual precipitation data of the same 30 stations of the
study area for 30 years has been statistically analyzed. It was observed that the
distribution of precipitation is not normal on the basis of large inter-annual variations.
Therefore, rainfall data was analyzed using ANOVA and Dunnett T3 for multiple
comparison of alpha 0.05 between the decadal analyses. The principle feature of these
methods is that they can be used for analyses of datasets which do not display a normal
distribution. In order to show deviation of annual zonal data from average values,
statistical indicators such as standard error, skewness, percentile, 5% trimmed mean and
median were calculated. Results of the descriptive statistical analysis are presented in
Table 4.13 including ANOVA result for rainfall of the study area.
It is evident from these statistical indicators that highest zonal means and medians
were observed for zone1 (66.64 mm, 57.22 mm) and zone2 (66.99 mm, 57.05 mm) for 30
years average rainfall while zone 3b shows the lowest value of mean (22.65 mm) and
median (19.29 mm) respectively. Similarly other statistics such as standard error, 95%
confidence interval and quartile all are also higher for zone1 and 2 as compared to other
zones. The skewness shows positive values for all five zones with zone3a and 4 being
higher. Results of the ANOVA test are most significant (F= 71.64, p= 0.00) not only
within each zone but also for inter-zonal analysis on decadal and inter-decadal scales. In
Dunnett T3 test (Fig. 4.8 & Table. 4.6 in Annexure) all zones are quite different from
each other but not significant except zone3a and zone4 which show significance (p<
0.05) in the last two decades i.e. 1986-1995 and 1996-2005.
64
Table. 4.13 Descriptive statistics and ANOVA result of rainfall for all climate zones.
Rainfall (F= 71.64, p= 0.00)
Stations 10 6 4 6 4
30 Years Valid/Missing values 297/3 175/5 116/4 172/8 118/2
5%trimmed Mean 64.72 64.42 31.74 21.76 29.62
Median 57.22 57.05 30.47 19.29 25.04
Std. Error 2.70 2.77 1.39 1.05 2.35
95%CI Lower Bound 61.31 61.52 30.16 20.58 27.45
Upper Bound 71.96 72.46 35.68 24.72 36.75
Percentile 25 23.25 39.02 24.16 13.80 13.02
50 57.22 57.05 30.47 19.29 25.04
75 107.18 90.34 39.59 29.16 48.16
Skewness 0.46 1.12 1.56 1.15 1.45
80
70
60
50
mm
40
30
20
10
0
1976-1985 1986-1995 1996-2005
Figure: 4.8. Mean annual rainfall in all climate zones of the study area. Zone3a and Zone4 show
significance at p= 0.05 level.
65
The overall trend for the study area is most significant for intra-decadal and inter-
decadal comparison with (p<0.05) as shown in Table 4.14 and Fig.4.9 (Table.4.7 in
Annexure).
Table. 4.14 Descriptive statistics and ANOVA result for the whole study area
Whole Study area 95% CI for M ean

N F-value Sig. Mean Std. Error Upper Lower
Mean Rainfall 30 3.65 0.04 48.99 1.38 46.16 51.82
60
53.03 mm
49.30 mm
50 44.64 mm
40
mm
30
20
10
0
1976-1985 1986-1995 1996-2005
Figure: 4.9. Mean annual rainfall in the whole study area. Show significant decrease at p=0.0.5 level.
4.2.3.2. Zonal analysis
Similar calculations have been carried out for rainfall data analysis, as were done
for the maximum and minimum temperature, except the Dunnett T3 test because rainfall
data show unequal variances in homogeneity test (mainly used for equality of means).
4.2.3.2.1. Zone1
Probability for zone1 (F=1.93 p=0.15) is not significant as the computed values
are higher than the critical value (p> 0.05). However, different stations of zone1 such as
Chitral (p=0.00), Dir (p=0.05), Muzaffarabad (p=0.02), Skardu (p=0.04) and Said-U-
Sharif (p=0.00) show most significant results (Table 4.15). Similarly, results of Dunnett
T3 test for multiple comparison (Table 4.16) show most significant trend (p<0.05) in the
mean rainfall of Chitral, Dir, Muzaffarabad, Skardu and Said-U-Sharif stations for all
three decades.
66
Table. 4.15 Descriptive statistics and ANOVA result for zone 1
Zone1 (F= 1.93, p= 0.15) 95% C I for Mean
Astor 29 0.08 0.92 43.83 2.49 38.72 48.94
Chilas 30 2.24 0.13 19.33 2.01 15.22 23.45
Chitral 30 10.40 0.00 40.29 2.20 35.78 44.79
Dir 30 3.47 0.05 123.98 4.44 114.90 133.07
Gilgit 30 0.22 0.81 14.82 1.37 12.03 17.62
Kakul 28 2.23 0.13 116.84 3.64 109.36 124.32
M-abad 30 4.59 0.02 132.52 4.39 123.55 141.49
P-Chinar 30 0.07 0.94 67.53 2.29 62.84 72.23
Skardu 30 3.81 0.04 20.84 1.50 17.76 23.91
S-Sharif 30 10.76 0.00 88.96 4.46 79.84 98.09
Table 4.16. Dunnett T3 test for multiple comparison

Stations (I) Decade (J) Decade Mean Difference (I-J) Sig.
Astor 1 2 2.24 0.98
3 2.21 0.99
2 1 -2.24 0.98
3 -0.03 1.00
3 1 -2.21 0.99
2 0.03 1.00
Chilas 1 2 -5.73 0.25
3 -9.98 0.24
2 1 5.73 0.25
3 -4.25 0.79
3 1 9.98 0.24
2 4.25 0.79
Chitral 1 2 -18.84 0.00
3 -12.45 0.03
2 1 18.84 0.00
3 6.39 0.42
3 1 12.45 0.03
2 -6.39 0.42
Dir 1 2 -19.80 0.12
3 5.38 0.92
2 1 19.80 0.12
3 25.17 0.11
3 1 -5.38 0.92
2 -25.17 0.11
Gilgit 1 2 -2.03 0.94
3 -1.87 0.87
2 1 2.03 0.94
3 0.17 1.00
3 1 1.87 0.87
2 -0.17 1.00
67
Kakul 1 2 -2.19 0.99
3 15.08 0.28
2 1 2.19 0.99
3 17.27 0.25
3 1 -15.08 0.28
2 -17.27 0.25
M-abad 1 2 -26.16 0.03
3 -1.94 1.00
2 1 26.16 0.03
3 24.22 0.07
3 1 1.94 1.00
2 -24.22 0.07
Parachinar 1 2 1.78 0.98
3 -0.11 1.00
2 1 -1.78 0.98
3 -1.88 0.98
3 1 0.11 1.00
2 1.88 0.98
Skardu 1 2 -9.30 0.02
3 -4.72 0.42
2 1 9.30 0.02
3 4.58 0.58
3 1 4.72 0.42
2 -4.58 0.58
S-Sharif 1 2 -38.21 0.00
3 -11.50 0.42
2 1 38.21 0.00
3 26.71 0.01
3 1 11.50 0.42
2 -26.71 0.01
4.2.3.2.2. Zone2
ANOVA test applied on zone2 data (Table 4.17) display non significant results
for all six stations (p> 0.05 or 0.1). Similarly zone2 is also observed with non-significant
trend for the mean rainfall. No significant change is noted in Dunnett T3 test (Table 4.18)
between decadal and inter-decadal analysis of the stations.
Zone2 (F= 0.91, p= 0.41) 95% CI for Mean

Cherat 27 0.46 0.64 54.35 2.80 48.59 60.11
DIK 30 0.54 0.59 29.58 1.62 26.27 32.88
Isl-bad 30 1.74 0.20 110.50 5.33 99.60 121.39
Lahore 30 2.61 0.09 69.35 6.15 56.78 81.92
Peshawar 28 0.17 0.84 42.61 2.43 37.63 47.60
Sialkot 30 0.05 0.95 92.67 5.34 81.75 103.59
68
Table. 4.18 Dunnett T3 for multiple comparison
Cherat 1 2 5.23 0.82
3 6.40 0.83
2 1 -5.23 0.82
3 1.17 1.00
3 1 -6.40 0.83
2 -1.17 1.00
D.I.Khan 1 2 -4.00 0.65
3 -0.97 0.99
2 1 4.00 0.65
3 3.03 0.82
3 1 0.97 0.99
2 -3.03 0.82
Isl-bad 1 2 20.09 0.42
3 20.99 0.32
2 1 -20.09 0.42
3 0.90 1.00
3 1 -20.99 0.32
2 -0.90 1.00
Lahore 1 2 31.74 0.18
3 22.41 0.49
2 1 -31.74 0.18
3 -9.33 0.66
3 1 -22.41 0.49
2 9.33 0.66
Peshawar 1 2 1.79 0.98
3 -1.91 0.99
2 1 -1.79 0.98
3 -3.70 0.94
3 1 1.91 0.99
2 3.70 0.94
Sialkot 1 2 -4.26 0.98
3 -1.19 1.00
2 1 4.26 0.98
3 3.07 0.99
3 1 1.19 1.00
2 -3.07 0.99
4.2.3.2.3. Zone3a
Two stations i.e. Khuzdar (p=0.04) and Quetta (p=0.02) are most significant to
the change while Kalat and Zhob show no trend in the 30 years mean rainfall as shown in
Table 4.19. The ANOVA result for zone3a (F=3.81, p=0.03) is found most significant to
69
the change. Multiple comparisons for the decades show no significance for Kalat & Zhob
while Quetta and Khuzdar (p< 0.05) are significant in the last two decades (Table 4.20).
Table. 4.19 Descriptive statistics and ANOVA result for zone3a
Zone3a (F= 3.81, p= 0.03) 95% C I for Mean

Kalat 26 0.94 0.41 28.47 3.12 22.04 34.90
Khuzdar 30 3.67 0.04 31.12 2.29 26.45 35.80
Quetta 30 4.88 0.02 40.47 3.46 33.40 47.55
Zhob 30 1.85 0.18 31.02 1.58 27.79 34.25
Table. 4.20 Dunnett T3 for multiple comparisons

Kalat 1 2 -10.94 0.44
3 -9.08 0.56
2 1 10.94 0.44
3 1.85 0.99
3 1 9.08 0.56
2 -1.85 0.99
Khuzdar 1 2 -5.82 0.65
3 8.06 0.18
2 1 5.82 0.65
3 13.88 0.08
3 1 -8.06 0.18
2 -13.88 0.08
Quetta 1 2 12.73 0.41
3 23.48 0.06
2 1 -12.73 0.41
3 10.75 0.04
3 1 -23.48 0.06
2 -10.75 0.04
Zhob 1 2 -5.60 0.42
3 1.18 0.99
2 1 5.60 0.42
3 6.78 0.21
3 1 -1.18 0.99
2 -6.78 0.21
70
4.2.3.2.4. Zone 3b
Like zone2, ANOVA Table (4.21) for zone3b also show no significant trend (p>
0.05) for all stations of the zone. Besides, the test for analysis of variance of the zone is
also found non-significant (F=1.53, p=0.22). Table 4.22 depicts no significant trend in the
Dunnett T3 test for multiple comparisons of the decades.
Table. 4.21 Descriptive statistics and ANOVA result for zone3b
Zone 3b (F=1.53,p= 0.22) 95% CI for Mean

B-Pur 30 0.65 0.53 23.19 2.75 17.57 28.81
Jaco-bad 23 1.20 0.32 21.65 3.36 14.68 28.62
Khanpur 29 1.08 0.36 16.00 1.96 11.98 20.03
Multan 30 0.36 0.70 25.64 2.06 21.43 29.85
Rohri 30 0.89 0.42 25.77 3.05 19.53 32.01
Sibbi 30 0.09 0.92 23.17 1.94 19.20 27.15
Table. 4.22 Dunnett T3 for multiple comparison

Stations (I) Decade (J) Decade Mean Diff (I-J) Sig.
B-Pur 1 2 6.35 0.82
3 7.02 0.75
2 1 -6.35 0.82
3 0.67 1.00
3 1 -7.02 0.75
2 -0.67 1.00
Jaco-bad 1 2 1.81 1.00
3 11.13 0.35
2 1 -1.81 1.00
3 9.32 0.73
3 1 -11.13 0.35
2 -9.32 0.73
Khanpur 1 2 -5.22 0.52
3 1.48 0.99
2 1 5.22 0.52
3 6.70 0.49
3 1 -1.48 0.99
2 -6.70 0.49
Multan 1 2 -0.56 1.00
3 3.45 0.89
2 1 0.56 1.00
3 4.01 0.81
3 1 -3.45 0.89
2 -4.01 0.81
71
Rohri 1 2 -9.52 0.50
3 -2.13 0.98
2 1 9.52 0.50
3 7.39 0.77
3 1 2.13 0.98
2 -7.39 0.77
Sibbi 1 2 -1.54 0.98
3 -1.94 0.97
2 1 1.54 0.98
3 -0.40 1.00
3 1 1.94 0.97
2 0.40 1.00
4.2.3.2.5. Zone 4
The ANOVA Table for zone4 (4.23) shows most significant result (F=5.92,
p=0.00) compared to other climatic zones of the country. Karachi station is observed with
most significant trend for thirty years average rainfall. The rest of stations i.e. Hyderabad,
Jewani and Nawabshah are not found significant for change during decadal analysis in
the Dunnett T3 test (Table 4.24).
Table 4.23 Descriptive statistics and ANOVA result for zone4
Zone 4 (F= 5.92, p= 0.00) 95% CI for Mean

H-abad 30 1.53 0.24 36.01 6.12 23.50 48.52
Jewani 30 1.97 0.16 25.93 2.90 20.00 31.86
Karachi 29 4.87 0.02 35.20 3.93 27.16 43.25
N-Shah 29 2.20 0.13 31.34 5.20 20.69 41.98
72
Table.4.24 Dunnett T3 test for multiple comparison
Stations (I) Decade (J) Decade Mean Diff (I-J) Sig.

H-abad 1 2 -13.79 0.75
3 11.93 0.63
2 1 13.79 0.75
3 25.72 0.39
3 1 -11.93 0.63
2 -25.72 0.39
Jewani 1 2 -8.57 0.58
3 4.91 0.80
2 1 8.57 0.58
3 13.48 0.21
3 1 -4.91 0.80
2 -13.48 0.21
Karachi 1 2 16.80 0.20
3 25.77 0.01
2 1 -16.80 0.20
3 8.97 0.73
3 1 -25.77 0.01
2 -8.97 0.73
N-Shah 1 2 -16.13 0.40
3 9.57 0.79
2 1 16.13 0.40
3 25.70 0.25
3 1 -9.57 0.79
2 -25.70 0.25
4.2.3.3. Single Station Analysis
By dividing the 30-years rainfall data into two 15-years periods (1976-1990 and
1991-2005) and computing change from the 1st half period to the 2nd, for each station,
zone and the country, the results are outlined below. The average rainfall in the country
was 50.48 mm during 1976-90 period while 47.5 mm was recorded for the 1991-2005
period. This shows a decline of 2.98 mm from the first half with non significant mean
variance (F=1.273, p= 0.259).
The trend analysis for individual stations is presented in Table 4.8 (Annexure) and
shown in Fig.4.10. The intensity of the map color shows mean inter-zonal rainfall
variation from 134.19 - 16.09 mm in Fig.4.10 (A) to 130.19 - 14.26 mm (B) for different
zones from the first half study period to the second half. Among the stations Chitral
73
shows a significantly higher trend with 11.23 mm increase between the two 15-years
periods while rest of the stations show an increasing trend in the range of 1 – 6 mm but
with no significance value (p>0.05). On the other hand rainfall data for Quetta (-
13.77mm) and Karachi (-14.72 mm) stations indicate a significant decrease in the mean
rainfall during the second half period. Similarly Dir, Gilgit, Kakul, Muzaffarabad,
Islamabad, Lahore, Kalat, Khuzdar, Bahawalpure, Khanpur, Hyderabad, Jewani, and
Nawabshah stations also show decrease in the average rainfall, albeit with a non-
significant trend. The overall difference in mean rainfall for each individual station is
observed with a decreasing trend in the 1991-2005 period as compared to 1976-90. In the
zonal analysis of rainfall, these zones show opposite trend to that as observed in case of
minimum and maximum temperatures, having lower values of means (22.65 and 32.10
mm) and medians (19.29 and 25.04 mm) for zone3b and zone4, respectively.
(A)
74
(B)
Figure 4.10: Map of Pakistan showing significant increase, decrease and no change in the mean
annual rainfall. (A): For period 1976-1990; (B): For period 1991-2005.
4.2.3.4. Whole Study Area Analysis
It is evident that all five zones in Fig. 4.11 (A & B) and Table 4.9 (Annexure)
show a decreasing trend in average rainfall between the two 15-year periods. Spatial
variation in mean rainfall across the country ranges from 68.23 - 23.37 mm during 1976-
90 period while it decreases to 66.41- 21.91 mm during 1991-2005 period. ANOVA test
for the two periods is found most significant only for zone4 (F=3.719, p-0.05) which
indicates a negative change in the mean rainfall (-8.96mm) followed by zone3a (-
4.33mm) that also shows a higher decreasing trend with a relatively higher p-value (F=
2.439, p= 0.1) than the significant tabulated value.
75
(A)
(B)
Figure 4.11: Whole study area analysis showing significant increase, decrease and no change in
Mean annual rainfall (A): For period, 1976-1990; (B): For period, 1991-2005.
76
It can be summarized from the above analysis that the overall trend calculated for
average rainfall is decreasing on both decadal and 15-years scales. Karachi and Quetta
stations located in different zones show a significant downward tendency in rainfall while
only Chitral station is found with an increasing trend. In the decade-wise analysis stations
of zone1 like Dir, Muzaffarabad, Skardu, and Said-u-Sharif show a significant trend
(p<0.05) but in terms of comparison between the two 15-years periods, all these stations
show a non-significant trend (p>0.05). Similarly zone3a and zone4 are most significant in
decadal analysis but in 15-years analysis zone3a is not much significant. ANOVA result
for the country into two time periods is found non-significant (F=1.273, p= 0.26).
4.2.3.5. Entire Study Period Analysis

In Figure (4.12) the time series data of mean annual rainfall for the entire study
period is observed with decreasing trend. Mean annual rainfall is showing up and down
fluctuation but in last decade observed with sudden downfall. The highest mean annual
rainfall was recorded in the year 1992 (63.9 mm) while lowest mean annual rainfall of
32.8mm - 35.8mm were observed in 1998-2002. The trend line generated through the
scattered data point also indicates a downward trend for rainfall in the entire study period
from 1976-2005.
Time series of Mean Annual Rainfall of the whole study area 1976-2005
70.0
65.0
60.0
55.0
mm
50.0
45.0
40.0
35.0
30.0
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
Time series
Figure: 4.12. Mean annual rainfall trend for the entire study period from 1976-2005.
77
4.2.4. Discussion
The analysis of rainfall data shows that zone1 (mean 66.64 mm & median, 57.22
mm) and zone2 (mean 66.99 mm and median 57.05 mm) show higher standard error
(2.7mm). In the ANOVA test for decadal analysis, zone3a and zone4 were found most
significant (p<0.05) and zone1 slightly significant, with probability approximately 85%.
The single station analysis suggests that Chitral was found significantly higher with mean
difference of 11.23mm in the two 15-year periods while rest of the stations were
observed with a increasing trend in the range of 1 to 6 mm but with no significance value
(p>0.05) except for Chilas (p=0.09>0.05) with approximately 90% of probability. Quetta
(-13.77 mm) and Karachi (-14.72 mm) are observed with significantly lower mean value
and Jacobabad (-12.2 mm) slightly significant mean value in the second half period.
Similarly Dir, Gilgit, Kakul, Muzaffarabad, Islamabad, Lahore, Kalat, Khuzdar,
Bahawalpure, Khanpur, Hyderabad, Jewani, and Nawabshah have shown decrease in the
average rainfall with non significant trend. But in the decadal analysis Dir, Muzaffarabad,
Skardu, and Said-u-Sharif were observed with a significant trend (p<0.05). The overall
mean difference in the average rainfall for each individual station and for the country
shows a decrease in from 1976-90 to the 1991-2005 periods. Zone4 was found most
significant in comparison between the two time series (p<0.05) with a negative change in
mean rainfall (-0.97 mm) followed by zone3a with slightly higher p-value approximately
(p=0.1). It is also clear that zone1, zone3a and zone4 situated in different regions of the
country showed an overall positive trend while plain areas and southwest of the country
have been recorded with no significant trend during the last two decades.
The negative difference observed in rainfall data for the whole country is (-3.55
mm) between the two 15-years intervals, which works out to -1.18 mm per decade. This
observational decrease in rainfall data is supported by the published data in IPCC (2001)
report, which talks about 0.3% average decrease in rainfall per decade for the subtropical
land areas as opposed to tropical lands with 0.3% increase per decade. Similarly most
parts of Europe and East Asia are observed with a positive trend in the annual maximum
consecutive days having rainfall below 1.0 mm and a negative trend in the number of
rainy days during 1950-1995 period (Kiktev et al, 2003). The results of Kiktev et al
(2003), were found relevant with the research conducted on the Mediterranean areas by
78
Trigo et al (2000), and Alpert et al (2002), during the period 1951-1995. Their research
works, in the Mediterranean areas, indicate a larger frequency of drought periods, with
associated impacts on agriculture, water resources and socio-economic activities.
Decreasing trends in annual mean rainfall are observed in Russia, North-East and North
China, coastal belts and most parts of North-East India, Indonesia, Philippines and some
areas in Japan. Annual mean rainfall exhibits increasing trends in Western China, South-
Eastern coast of China, Arabian Peninsula, Bangladesh and along the western coasts of
the Philippines (IPCC, 2007). Liu et al (2005), analyzed heavy precipitation events in
China over the period 1960-2000. They found that the increased frequency of heavy
precipitation events in China contributed 95% of the total increase of precipitation and
only 2% increase in total precipitation was observed over that time period. Otherwise
total rainy days trend was observed negative. For India, Roy and Balling (2004), found
about two thirds of increasing trend for precipitation extremes during all the study period
from 1910–2000 and also observed some regions with significant increase and decrease
all over India.
4.3. RIVER FLOWS

4.3.1. Introduction
Global warming induced changes in the temperature and rainfall trends and their
impact on river flows are already evident in many parts of the world, particularly in South
Asia. Heating of the earth is not only due to natural processes but mainly because of non-
environment friendly anthropogenic activities (Rasul et al., 2006). Because of this
phenomenon certain unpredictable changes occurred in the climate of Pakistan. Hazards
like floods, droughts, cyclones and others adverse consequences are being experienced
more frequently than ever before. Rising temperature and uneven temporal and spatial
distribution of rainfall is causing floods on one hand and droughts on the other. These
extreme weather events have altered the hydrological cycle globally and resulted in the
intensification of available water resources. (Lai et al., 1998; Vörösmarty et al. 2000;
Allen and Ingram, 2002). Changes in water resources mainly depend on changes in
precipitation and temperature. These climate variables have a direct effect on watersheds
or rivers as changes in water flow or discharge resemble the overall nature of
79
precipitation in large part because precipitation is the primary factor in determining
surface runoff. The increased flow of water has been noted in higher latitude regions and
decreases in sub-tropical areas. Demand for agricultural and domestic water in particular
increases significantly at hotter and drier times of the year. Agriculture has always been
the dominant end-user of surface water for irrigation. Meanwhile, demands for industry
are also expected to rise and become a greater issue in competition with agricultural and
domestic water supply.
4.3.2. Data Collection and Analysis

Temporal variability in river discharges was investigated for six gauge stations of
five relevant rivers of the country, comprising 25-years data length.. ANOVA test was
used to identify whether the time series has a declining or rising trend or no significant
trend at all in decades and fifteen year periods.
4.3.3. Results
4.3.3.1. Descriptive statistics and ANOVA results
Duncan test was applied to the dataset of river flows also because of its equal
variances. Table 4.25 indicates descriptive statistics for river discharges in the study area.
Among the different rivers included in this study, the Indus river being the major river of
the country shows the highest mean discharge and median values at both of its gauging
stations i.e. Chilas (1786 & 1748 m3/s) and its upstream station Skardu (1091 & 1052
m3/s) respectively. The smallest mean and median discharge among the studied rivers is
shown by the Swat River (87 & 89 m3/s) and Astor River (138 & 135 m3/s). Swat River
also has the lowest value of standard error i.e. 2.33 m3/s and its distribution negatively
skewed in comparison to other rivers with positively skewed.
The result of analysis of variance on decadal scale is most significant for Astor
river (F= 7.134, p= 0.00) and in decreasing order for Swat river (F= 5.488, p= 0.01) and
Indus river at Skardu (F= 5.4, p= 0.01) as shown in the Fig (4.13). ANOVA test shows
that Chitral river (F= 1.643, p= 0.22), Indus at Chilas (F= 0.992, p=0.39), and Gilgit river
(F= 0.7, p= 0.51) are non-significant. The homogenous subset of alpha (0.05) in Duncan
test (Table 4.10 in Annexure), decadal means of all rivers are significantly different from
each other and non-significant at all.
80
Table 4.25 Descriptive statistics and ANOVA result for ensuing rivers
Ensuing Rivers
Rivers Astor Chitral Gilgit Indus/Chilas Indus/Skardu Swat
25 years Valid/Missing 25/0 25/0 25/0 25/0 25/0 25/0
values
Mean All value used 137.65 276.59 612.9 1786.19 1091.19 87.24
5% trimmed Mean 136.94 277.03 611.9 1771.17 1087.06 87.52
Median 134.99 275.06 608.9 1748.08 1052.47 88.79
Std. Error 5.9 5.48 12.32 51.96 37.57 2.33
95%CI Lower Bound 125.47 265.28 587.5 1678.94 1013.64 82.43
Upper Bound 149.84 287.89 638.3 1893.44 1168.74 92.05
Percentile 25 116.11 257.99 569.4 1534.93 919.04 80.24
50 134.99 275.06 608.9 1748.08 1052.47 88.79
75 162.27 294.59 652.9 1952.78 1251.6 94.19
Skewness 0.26 0.03 0.19 0.73 0.36 -0.46
ANOVA F-value 7.134 1.643 0.7 0.992 5.4 5.488
Sig. 0.00 0.22 0.51 0.39 0.01 0.01
Analysis of River discharge on decadal basis

2500
1740 m 3 /s
2000 1932 m 3 /s
1760 m 3 /s
1141 m 3 /s
1500
1236 m 3 /s
m 3 /s
623 m 3 /s
969 m 3 /s
1000 3
284 m /s
159 m 3 /s 617 m 3 /s 95 m 3 /s
584 m 3 /s
500
3
285 m 3 /s
118 m 3 /s 135 m /s 87 m 3 /s
80 m 3 /s
265 m 3 /s
0
Astor Chitral Gilgit Indus/Chilas Indus/Skardu Swat
1st Decade 2nd Decade 3rd Decade
Figure 4.13: Single river discharge on decadal basis
4.3.3.2. Single Gauge-Station Analysis

Temporal variations in river discharge during the study period were examined by
dividing the dataset into two unequal periods i.e. 15-years from 1976-90 and 10-years
81
from 1991-2000. Analysis shown in Table 4.26 indicates that three out of six stations are
most significant, having p- values less than 0.05, whereas the remaining are non-
significant (p>0.05). The highest increasing trend in discharge was observed for Indus
river at Skardu followed Chitral river with mean differences of 175.72 and 22.76 m3/s
respectively from the 1976-90 period to the 1991-2000 period.
Table. 4. 26 Single Gauge station data analysis for trend determination

River F-value Sig. Mean (1976-1990) Mean (1991-2000) Mean Diff
Astor 3.07 0.09 129.54 149.82 20.27
Chitral 4.80 0.04 267.48 290.25 22.76
Gilgit 0.11 0.74 616.30 607.80 -8.51
Indus/Chilas 4.05 0.06 1705.74 1906.86 201.13
Indus/Skardu 6.44 0.02 1020.90 1196.62 175.72
Swat 4.96 0.04 83.31 93.14 9.83
The analysis of ensuing rivers has revealed mixed results. It is evident that both
Indus/Skardu and Chilas and Swat rivers show significant increase and almost similar
trend was observed for the Astor River. Further, on 15-years scale Chitral River shows a
significant increasing trend but on decadal scale it shows no significance.
4.4. HYDRO-METEOROLOGICAL CHARACTERISTICS

Hydro-meteorological characteristics have been determined through water yield
or contribution per annum in catchment‘s basin and each river basin analysis. In Table
(4.27) Swat river shows highest value (1,366,352 m3/s) of water contribution per annum
in the catchments‗s area while Indus river at Skardu Station shows the lowest value
(305,434 m3/s) of water yield per annum.
Table. 4.27 Water Yield per Annum in Catchment’s Basin
Ensuing Rivers (m3/s)
Rivers Astor Chitral Gilgit Indus/Chilas Indus/Skardu Swat
Mean 138 277 613 1786 1091 88
Catchment‘s Area
(Km2) 4040 11396 25900 142709 112665 2020
Water Yield Per
Annum 1,074,488 766,622 746,271 394,714 305,434 1,366,352
82
4.4.1. River Basin Analysis
In this section characteristics of each river basin along with their temperature and
rainfall record have been described to examine their influence on the river flow.
4.4.1.1. Astor River
Astor River, a tributary of the Indus River and rises in a glacier on the north-
facing slopes of the great Himalayan, range near the Burzil Pass in the Ladakh region of
Jammu and Kashmir. It flows in a northwesterly direction and joins Gilgit River at
coordinates between 34°00′ and 74°41′ (Wikipedia, 2012). Many small snow-fed streams
originating from different depressions on the great Himalayan range join the Astor River
in its short course. Its catchment‘s area consists of 4040 km2 and having U-shaped. As
shown in Table (4.25) the mean annual discharge rate of Astor River is approximately
138 m3/s, whereas the maximum discharge of 400 m3/s takes place in July while the
minimum discharge or base flow is recorded in February i.e. approx. 29 m 3/s (Table
4.28). A comparison of the mean monthly maximum temperature of Astor station for the
month of July with the discharge of Astor river for the same month corresponds very well
because the temperature of July (27.21 oC) is highest among all the months. In has been
noted that although maximum rainfall takes place in the month of August but it does
cause a significant increase in the discharge of Astor River. The table also shows that
49% flow takes place in the monsoon period i.e. Jul-Sep while 36% flow is received in
the pre-monsoon period (Apr-Jun). The flow duration curve for Astor River (Fig. 4.14)
indicates that approximate 35% of the time the discharge remains above the mean while
65% of the time the discharge is less than the mean. It also shows that up to 5% of the
time the discharge in access of 200 to 500% of the mean flow. The time series plot for
Astor River (Fig. 4.15) shows that the highest annual discharge took place in 1996 (200
m3/s) whereas the lowest discharge of 88 m3/s was recorded in 1985 and 1997. The trend
line (best fit curve) generated by the software through the scattered data point also
indicates a clear increasing in the discharge rate from 1976 to 2000. The same is
corroborated by the ANOVA result presented in Table 4.3.
83
Table. 4.28 Hydro-meteorological characteristics of Astor River Basin
Astor River
Months Mean Monthly Mean Monthly Mean Mean Monthly Mean
Min Temp Max Temp Monthly Discharge Monthly
(oC) (oC) Rainfall (m3/s) Discharge
(mm) (%age)
Jan -6.63 2.53 46.02 31.53 2
Feb -6.14 4.25 44.12 28.78 2
Mar -1.50 8.50 88.93 29.92 2
Apr 4.10 15.2 78.93 56.71 3
May 7.40 19.7 66.02 183.92 11
Jun 11.12 24.40 27.76 372 23
Jul 13.42 27.21 26.13 408.72 24
Aug 14.30 27.05 33.34 261.48 16
Sep 9.94 23.44 28.05 130.26 8
Oct 4.71 17.68 23.63 65.04 4
Nov -1.10 11.30 27.54 46.71 3
Dec -4.60 5.67 35.18 36.76 2
600
Flow Duration Curve for Astore River 1976-2000
500
Percent of Mean Flow
400
300
200
100
0
99% 95% 90% 85% 80% 75% 70% 65% 60% 55% 50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 1%
Percent of Time
1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988
1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
Figure: 4.14. Flow duration curve for Astor River from the period 1976-2000
84
Time Series of Mean Annual Discharge of Astor River
220
200
180
160
m3/s
140
120
100
80
60
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
00
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
20
Time Series 1976-2000
Figure: 4.15. Hydrograph of mean annual discharge of Astor River at Doyian station.
4.4.1.2. Chitral River

Chitral River is one of the major indirect tributaries of the Indus River arises near
the Baroghil pass in Hindukush Mountains. The upper most section of the Chitral River is
known as Yarkhum River, between Mastuj and Chitral towns and downstream it is called
Chitral River located between 35o50 and 71o48. The drainage area of Chitral River at
Chitral is 7080 km2. According to WAPDA (1997b) that precipitation rate in Chitral river
catchment‘s area is comparatively lows, especially the influence of the monsoon rainfall,
which makes the more northern parts of the valley quite arid. Because of high altitude,
precipitation mostly comes in the form of snowfall that feeds snowfields and glaciers of
Hindukush Mountains. The dataset of Chitral River shows the mean annual discharge of
277 m3/s (Table 4.25) with a strong monthly variance ranging from a maximum of 841.90
m3/s in July to a minimum of 66.44 m3/s in March (Table 4.29). It has also been noted
from the weather record of Chitral station that on average July is the hottest and driest
month of the year having mean maximum temperature of 36 oC and 8mm of average
rainfall whereas the mean maximum temperature of March is about 15oC and an average
of 117mm rainfall. Hence it is clear that temperature has a profound influence on the
snow/glacial melt in comparison to rainfall. The discharge of Chitral River increases
almost 12 times from March to July. The analysis also shows that maximum flow (60%)
85
takes place in the monsoon period i.e. Jul-Sep followed by 23% flow in pre-monsoon
period. The flow duration curve (Fig. 4.16) indicates that the discharge remains above the
mean approximately 30% of the time while 60% of the time the discharge is less than the
mean value. It also shows that up to 10% of the time the discharge in access of 300 to
900% of the mean flow. It means that snow/glacial cover of the area melt or retreats with
respect to the climate and respond in similar way to seasonal and annual variations in
discharge as suggested by Ferguson (1984). Plotting discharge values as time series
(Fig.4.17) shows that the highest annual discharge took place in 1988 and 1992 (320
m3/s) whereas the lowest discharge of 220 m3/s was recorded in 1982. The trend line
indicates a clear increasing trend in the discharge rate from 1976 to 2000. The same
significance trend is shown by the ANOVA result presented in Table 4.3.
Table. 4.29 Hydro-meteorological characteristics of Chitral River Basin.
Chitral River
Months Mean Mean Mean Mean Mean Monthly
Monthly Min Monthly Max Monthly Monthly Discharge Discharge
Temp (oC) Temp Rainfall (%age)
(mm) (m3/s)
(oC)
Jan -0.51 9.45 47.82 77.15 2
Feb 0.80 10.68 57.19 68.36 2
Mar 3.91 14.83 117.27 66.44 2
Apr 8.28 22.60 70.45 90.61 3
May 12.12 28.31 43.28 191.34 6
Jun 16.92 34.14 14.61 463.97 14
Jul 19.99 36.02 8.39 841.90 25
Aug 18.48 34.56 6.49 758.03 23
Sep 13.10 31.73 14.14 386.19 12
Oct 7.16 25.29 18.35 172.03 5
Nov 2.59 18.94 30.40 112.77 3
Dec 0.23 12.53 39.36 90.27 3
86
Flow Duration Curve for Chitral River 1976-2000
900
800
700
600
500
400
300
200
100
0
99% 95% 90% 85% 80% 75% 70% 65% 60% 55% 50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 1%
Percent of Time
1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988
1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
Figure: 4.16. Flow duration curve for Chitral River from the period 1976-2000
Time Series of Mean Annual Discharge of Chitral River
340
320
300
280
m3/s
260
240
220
200
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
00
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
20
Figure: 4.17. Hydrograph of mean annual discharge of Chitral River at Kelash station.
4.4.1.3. Gilgit River

Gilgit River is a tributary of the Indus River, and flows past the town of Gilgit. It
is located in the Northern Areas of Gilgit–Baltistan, Pakistan. The Gilgit starts from
87
Shandur Lake. Gilgit River at Alam Bridge is a junction point between Indus and Gilgit
rivers at coordinates between 35o 45‘ and 74 o
37‘. The drainage area of Gilgit river at
Alam bridge is 25900 Km2. Likewise, the Gilgit River shows the maximum discharge
rate in July i.e., approximately 2030 m3/s with highest mean maximum temperature of
36oC while minimum discharge rate is recorded in March (95 m3/s) with mean maximum
temperature of 18oC as shown in the Table 4.30. Further mean annual discharge rate from
1976-2000 is found 613 m3/s as indicated in the Table 4.25. The table also shows that
63% flow takes place in the monsoon period i.e. Jul-Sep while 25% flow is received in
the pre-monsoon period (Apr-Jun). The flow duration curve for Gilgit River (Fig. 4.18)
indicates that approximate 25% of the time the discharge remains above the mean while
65% of the time the discharge is less than the mean. It also shows that up to 15% of the
time the discharge in access of 200 to 500% of the mean flow. The time series plot for
Gilgit River (Fig. 4.19) shows that the highest annual discharge took place in 1978 (753
m3/s) whereas the lowest discharge of 494 m3/s was recorded in 1997. The trend line
through the scattered data point also indicates a clear increasing in the discharge rate
from 1976 to 2000. The same is corroborated by the ANOVA result presented in Table
4.3.
Table.4.30 Hydro-meteorological characteristics of Gilgit River Basin
Gilgit River
Months Mean Mean Monthly Mean Mean Mean Monthly
Monthly Min Max Temp Monthly Monthly Discharge
Temp (oC) (oC) Rainfall Discharge (%age)
(mm)
(m3/s)
Jan -1.98 10.06 4.57 115.90 2
Feb -0.06 12.44 7.16 102.06 1
Mar 5.48 17.66 13.08 95.33 1
Apr 8.81 24.45 26.29 124.85 2
May 11.67 28.91 31.90 422.81 6
Jun 14.62 33.96 9.96 1222.36 17
Jul 17.41 36.43 15.59 2029.36 28
Aug 17.20 35.33 19.63 1791.04 24
Sep 12.44 32.09 11.84 821.46 11
Oct 6.28 25.54 9.53 310.19 4
Nov 0.24 18.52 5.68 181.43 2
Dec -1.63 12.18 8.75 138.02 2
88
Flow Duration Curve for Gilgit River 1976-2000
600
500
400
300
200
100
0
99% 95% 90% 85% 80% 75% 70% 65% 60% 55% 50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 1%
Percent of Time
1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988
1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
Figure: 4.18. Flow duration curve for Gilgit River from the period 1976-2000
Time Series of Mean Annual Discharge of Gilgit River
800
750
700
m3/s
650
600
550
500
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
00
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
20
Figure: 4.19. Hydrograph of mean annual discharge of Gilgit River at Alam Bridge.
4.4.1.4. Indus River/Chilas

Indus river at Chilas station coordinates at 35o 44‘ to 74o 37‘ over the old
suspension bridge, historically known as Partab Pul, it forms part of Astor district of the
Gilgit-Baltistan region. The Gilgit River joins the Indus a few kilometers above Bunji,
while the Astor River falls into the Indus a little below Bunji from the south-east. The
89
total drainage area of Indus River at Chilas is 142709 Km2. The data of Indus River at
Chilas station indicates the mean annual discharge of 1786 m3/s (Table 4.25) with
monthly variance ranging from a maximum of 5657 m3/s in July to a minimum of 307
m3/s in March (Table 4.31). It has also been observed that July is the hottest and driest
month of the year having mean maximum temperature of 39 oC and 14mm of average
rainfall whereas the mean maximum temperature of March is about 19 oC and an average
of 27mm rainfall. Hence it is clear that maximum flow (62%) takes place in the monsoon
period i.e. Jul-Sep followed by 25% flow in pre-monsoon period. The flow duration
curve (Fig. 4.20) indicates that the discharge remains above the mean approximately 25%
of the time while 65% of the time the discharge is less than the mean value. It also shows
that up to 15% of the time the discharge in access of 200 to 500% of the mean flow.
Plotting discharge values as time series (Fig.4.21) shows that the highest annual
discharge took place in 1977 (2412 m3/s) whereas the lowest discharge of 1472 m3/s was
recorded in 1986. The trend line indicates a clear increasing trend in the discharge rate
from 1976 to 2000. The same significance trend is shown by the ANOVA result
presented in Table 4.3.
Table.4.31 Hydro-meteorological characteristics of Indus River Basin at Chilas
Indus River at Chilas

Months Mean Mean Mean Mean Monthly Mean Monthly
Monthly Min Monthly Max Monthly Discharge Discharge
Temp (oC) Temp Rainfall (m3/s) (%age)
o
( C) (mm)
Jan 1.66 11.10 14.53 344.94 2

Feb 4.29 14.44 15.30 312.72 1
Mar 8.42 19.18 27.11 306.63 1
Apr 14.26 26.05 40.50 413.71 2
May 18.80 31.15 26.30 1249.12 6
Jun 24.03 36.98 14.49 3659.72 17
Jul 27.49 39.48 14.42 5675.64 27
Aug 26.48 38.43 11.08 5230.40 24
Sep 22.64 35.08 7.31 2424.24 11
Oct 14.79 28.24 15.31 875.02 4
Nov 7.70 20.88 17.31 532.50 3
Dec 2.38 13.90 16.27 409.61 2
90
Flow Duration Curve for Indus River at Chilas 1976-2000
600
500
400
300
200
100
0
99% 95% 90% 85% 80% 75% 70% 65% 60% 55% 50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 1%
Percent of Time
1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988
1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
Figure: 4.20. Flow duration curve for Indus River/Chilas from the period 1976-2000
Time Series of Mean Annual Discharge of Indus River at Chilas
2600
2400
2200
2000
m 3 /s
1800
1600
1400
1200
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
00
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
20
Figure: 4.21. Hydrograph of mean annual discharge of Indus river/Chilas at Partab Bridge
4.4.1.5. Indus River/Skardu

Indus River at Skardu located between coordinates 35o 27‘ and 75 o 25‘ on right
bank 32.18Km downstream from Skardu town, 3.22Km upstream Ayub bridge on
91
Kachura Kuwardo road. The drainage area of Indus River near Kachura is 112665 Km 2.
As shown in Table (4.25) the mean annual discharge rate of Indus River at Skardu is
approximately 1091 m3/s, whereas the maximum discharge of 3461 m3/s takes place in
July with mean monthly maximum temperature 32 oC and average rainfall 11mm while
the minimum base flow is recorded in March i.e. approx. 184 m3/s with mean monthly
temperature 12 oC (Table 4.32). The table also shows that 63% flow takes place in the
monsoon period i.e. Jul-Sep while 24% flow is received in the pre-monsoon period (Apr-
Jun). The flow duration curve for Indus River (Fig. 4.22) indicates that approximate 20%
of the time the discharge remains above the mean while 60% of the time the discharge is
less than the mean. It also shows that up to 20% of the time the discharge in access of 200
to 500% of the mean flow. The time series plot for Indus River (Fig. 4.23) shows that the
highest annual discharge took place in 1994 (1442 m3/s) whereas the lowest discharge of
811 m3/s was recorded in 1978. The trend line (best fit curve) generated by the software
through the scattered data point also indicates a clear increasing in the discharge rate
from 1976 to 2000.
Table.4.32 Hydro-meteorological characteristics of Indus River Basin at Skardu

Indus River at Skardu
Months Mean Mean Mean Mean Monthly Mean Monthly
Monthly Min Monthly Max Monthly Discharge Discharge
Temp (oC) Temp Rainfall (m3/s) (%age)
o
( C) (mm)
Jan -7.76 2.99 34.30 200.82 2

Feb -4.42 5.93 27.10 186.25 1
Mar 1.60 11.99 37.35 184.34 1
Apr 6.62 19.23 28.22 260.90 2
May 9.93 23.75 25.24 821.58 6
Jun 13.12 28.83 11.15 2122.84 16
Jul 16.46 32.07 11.38 3460.96 26
Aug 15.74 31.35 15.75 3234.76 25
Sep 11.44 27.75 11.89 1532.32 12
Oct 4.19 20.60 16.37 519.72 4
Nov -1.51 13.64 14.53 325.52 3
Dec -5.14 6.94 21.06 244.27 2
92
Flow Duration Curve for Indus River at Skardu 1976-2000
600
500
400
300
200
100
0
99% 95% 90% 85% 80% 75% 70% 65% 60% 55% 50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 1%
Percent of Time
1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988
1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
Figure: 4.22. Flow duration curve for Indus River/Skardu from the period 1976-2000
Time Series of Mean Annual Discharge of Indus River at Skardu
1500
1400
1300
1200
m 3 /s
1100
1000
900
800
700
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
00
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
20
Figure: 4.23. Hydrograph of mean annual discharge of Indus River/Skardu at Kachura station.
4.4.1.6. Swat River

The main source of Swat River is in the Hindukush Mountains, from where it
flows through the Kalam Valley and Swat District. From Lower Dir District it flows
93
through Malakand District to enter the Kabul River at Charsadda in the Peshawar valley.
The river is located at 35o30‘and 72o 35‗. The guage station is on the right bank of the
river about 3.28 Km downstream from the Kalam village. The data of Swat River
indicates the mean annual discharge of 87m3/s (Table 4.25) with monthly variance
ranging from a maximum of 261 m3/s in July to a minimum of 16 m3/s in March (Table
4.33). It has also been observed that maximum flow (48%) takes place in the monsoon
period i.e. Jul-Sep followed by 41% flow in pre-monsoon period. The flow duration
curve (Fig. 4.24) indicates that the discharge remains above the mean approximately 20%
of the time while 60% of the time the discharge is less than the mean value. It also shows
that up to 20% of the time the discharge in access of 200 to 500% of the mean flow.
Plotting discharge values as time series (Fig.4.25) shows that the highest annual
discharge took place in 1991 (110 m3/s) whereas the lowest discharge of 60 m3/s was
recorded in 1982 and 2000. The trend line indicates a clear increasing trend in the
discharge rate from 1976 to 2000. The same significance trend is shown by the ANOVA
result presented in Table 4.3.
Table. 4.33 Hydro-meteorological characteristics of Swat River Basin
Swat River
Months Mean Monthly Mean Monthly Mean Mean Monthly Mean
Min Temp Max Temp Monthly Discharge Monthly
(oC) (oC) Rainfall (m3/s) Discharge
(mm) (%age)
Jan 1.71 13.88 84.23 15.35 1
Feb 4.06 15.57 107.86 14.15 1
Mar 7.84 19.31 179.21 16.17 2
Apr 12.17 26.40 120.36 47.04 5
May 16.58 32.50 69.35 131.43 13
Jun 20.25 36.42 47.23 244.02 23
Jul 22.79 34.32 144.05 261.45 25
Aug 21.72 32.86 121.54 164.04 16
Sep 17.79 31.77 75.17 77.48 7
Oct 12.05 27.62 54.53 35.09 3
Nov 6.40 22.06 31.69 23.03 2
Dec 3.42 16.50 53.09 17.63 2
94
Flow Duration Curve for Swat River 1976-2000
500
450
400
percent of Mean Flow
350
300
250
200
150
100
50
0
99% 95% 90% 85% 80% 75% 70% 65% 60% 55% 50% 45% 40% 35% 30% 25% 20% 15% 10% 5% 1%
Percent of Time
1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988
1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000
Figure: 4.24. Flow duration curve for Swat River from the period 1976-2000
Time Series of Mean Annual Discharge of Swat River
120
110
100
90
m3/s
80
70
60
50
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
00
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
19
20
Figure: 4.25. Hydrograph of mean annual discharge of Swat River at Kalam station.
4.4.2. Discussion
Variability in river flows was determined for six gauge stations of those rivers
whose headwater regions lie in zone1. Among these, results of ANOVA test are observed
most significant (p<0.05) for Indus (at Skardu) and Swat rivers while Chitral and Gilgit
rivers along with Indus at Chilas are non-significant (p>0.05) both on decadal as well as
95
15-years scale. Astor River appears considerably significant on decadal scale but on 15-
years scale it shows no significant increasing trend. From the single river basin analysis,
it is clearly manifested by the data that mean monthly maximum temperature has strong
influence on river discharge and impact of rainfall is relatively small because large bulk
of river discharge occurs during May to September due to snow/glacial melt. The present
study is found in contrast with the global study like Karl and Riebsame (1989) have
found fluctuations in the decadal analysis of mean temperature up to 1oC annually and
2oC seasonally and changes in precipitation up to 10% in most areas of United States
during the modern climate record in the last 60 years. They also investigated the impact
of these fluctuations on the runoff of 82 rivers across the United States. Their research
has revealed that recent temperature increase has minimal impact on river flows but the
impact of precipitation is relatively large i.e. up to 10%, because of climatic influences
and dependence on the location, size and storage capacity of the drainage basin. Similarly
many researchers have found in their study that the melting of glaciers do not effect
annual discharge greatly but have impact on seasonal discharge. This is evident from the
research work of the Thayyen et al (2007) on Himalayan watershed. They found that
approximately 29% of the annual flow comes from snowmelt, but up to 70% in specific
months. Furthermore, by plotting the data into time series it is observed that the trend line
(best fit curve) generated by the software through the scattered data point also indicates a
clear increasing trend in the discharge rate from 1976 to 2000 for all of included rivers.
According to the study of Gleick (1986, 1987) annual river flow is greatly affected by
precipitation changes, not by temperature, while the seasonal distribution of river flow is
affected by changes in mean monthly temperature. During the past two decades
numerous studies have been conducted on the hydrological effects of climate change
such as Nemec and Schaake (1982), Mather and Feddema (1986), Gleick (1986; 1987;
1989; 1999), McCabe and Ayers (1989), Lettenmaier and Gan (1990), Schaake (1990)
and Ayers et al. (1994) from different regions of the world. These studies mainly focus
on the effects of changes in temperature and precipitation on mean monthly, seasonal, or
annual river flows. Similarly for global studies Probst and Tardy (1987, 1989), examined
discharge records of 50 large rivers for influence of rainfall on flows. They observed
increasing river flows over 1910–1975 in Africa and America, decreasing flow in Europe
96
and Asia, and 3% linear increase worldwide over the study period, which they attributed
to the global warming. Similarly Labat et al. (2004), analyzed runoff records of 221 large
rivers, for the periods 1900–1975 and 1925–1994. They observed increasing flow rates in
Africa, America and Asia and decreasing flows only in Europe, a global increase of 3%
over 1900–1975 and 8% over 1925–1994, because of accelerated hydrologic cycle caused
by global warming.
4.5. TRENDS FORECASTING

After 30 years analysis it is clear that the overall change in temperature is not
found uniform during the study period. Figure 4.26 (A & B) is the combine result of
minimum and maximum temperatures showing the regular increase in the last decade
with some significant decreasing fluctuations in 1976 to 1990. Overall results in figure
4.26 (C) show that mean temperature increased with fluctuations but have the significant
trends of rise in temperature during the whole study period but this change in two decade
is not more significant than the last decade. Mean annual rainfall in the fig. 4.26 (D)
show abrupt downfall after 1997 up to 2003 because of droughts (1998-2001) in the
country. In 2004 showing considerable rising trend but again in 2005 moving downward.
Therefore, the observed data of Fig.(4.26 E) forecasts an upward trend up to 50mm till
2035 with excluded drought period from the real time series.
As consequence of trend forecasting, the Time Series Modeler procedure
estimates exponential smoothing models for time series data of minimum, maximum,
mean temperatures and mean annual rainfall and produces forecasts. The procedure
includes an Expert Modeler that automatically identifies and estimates the best-fitting
ARIMA (Autoregressive Integrated Moving Average) or exponential smoothing model
for one or more dependent variable series. Models type and forecast of the mean
minimum, maximum, mean temperature and mean rainfall trend in the future are
presented in the Table (4.34) and model fit descriptive statistics Table (4.11) is attached
in the Annexure.
97
Table. 4.34 Show the trend forecasting of the climatic variables through different climatic models.
Forecast of the trend

Variable Model 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
Simple
Minimum Model Forecast 14.72 14.72 14.72 14.72 14.72 14.72 14.72 14.72 14.72 14.72 14.72 14.72 14.72 14.72 14.72
UCL 15.49 15.59 15.68 15.77 15.85 15.92 15.99 16.05 16.11 16.17 16.23 16.28 16.34 16.39 16.44
LCL 13.96 13.86 13.76 13.68 13.6 13.53 13.46 13.4 13.33 13.28 13.22 13.16 13.11 13.06 13.01
Holt
Maximum Model Forecast 28.44 28.46 28.49 28.52 28.55 28.58 28.61 28.63 28.66 28.69 28.72 28.75 28.77 28.8 28.83
UCL 29.43 29.46 29.49 29.52 29.55 29.58 29.61 29.64 29.67 29.71 29.74 29.77 29.8 29.83 29.86
LCL 27.45 27.47 27.5 27.52 27.55 27.57 27.6 27.62 27.65 27.67 27.7 27.72 27.75 27.78 27.8
Simple
Mean temp Model Forecast 21.48 21.48 21.48 21.48 21.48 21.48 21.48 21.48 21.48 21.48 21.48 21.48 21.48 21.48 21.48
UCL 22.24 22.29 22.34 22.38 22.43 22.47 22.51 22.54 22.58 22.62 22.65 22.68 22.72 22.75 22.78
LCL 20.71 20.66 20.61 20.57 20.53 20.49 20.45 20.41 20.37 20.34 20.3 20.27 20.24 20.2 20.17
Rainfall ARIMA Forecast 48.99 48.99 48.99 48.99 48.99 48.99 48.99 48.99 48.99 48.99 48.99 48.99 48.99 48.99 48.99
UCL 64.47 64.47 64.47 64.47 64.47 64.47 64.47 64.47 64.47 64.47 64.47 64.47 64.47 64.47 64.47
LCL 33.51 33.51 33.51 33.51 33.51 33.51 33.51 33.51 33.51 33.51 33.51 33.51 33.51 33.51 33.51
UCL: stands for Upper Critical Limit, LCL: stands for Lower Critical Limit.
98
(A)
(B)
99
(C)
(D)
100
Figure: 4.26. (A): Mean minimum temperature trend from 1976-2035; (B): Mean maximum
temperature trend from 1976-2035; (C): Mean Annual Temperature trend from 1976-2035; (D):
Mean Annual Rainfall trend from 1976-2035 with included drought period from 1998-2001; (E)
Mean Annual Rainfall trend from 1976-2035 with excluded drought period from 1998-2001.
4.5.1. Discussion
Analysis revealed that observed change in river flow is insufficient to explain the river
flow increasing trend because of less number of gauge stations and 25 years of data
record length. Therefore, for the prediction of the trend in the future, data of river
discharge was not accounted for consideration. Exponential Smoothing Model for
temperatures data and Autoregressive Integrated Moving Averages for rainfall data
analyses indicated a dominant periodicity of between (2006-2035) years within the study
area. The time series modeler estimated upward moving trend for minimum, maximum
and mean temperatures and downward moving trend for rainfall data. This declining
trend of rainfall data is due to the presence of drier period from 1998 to 2001 in which
Pakistan has faced severe drought, mainly in the southern and central parts of the country.
According to Awan (2002) Pakistan has experienced several droughts in the past, of
which the most severe one occurred in 1998-2002. Surface water availability was reduced
by over 30% and the total surface discharge in the major rivers declined from 162 billion
cubic meters (BCM) to 109 BCM, while rainfall was below normal. Furthermore
irrigation- based agriculture suffered to a great extent during the drought period, up to
101
51% of water supply through canal diversions than normal years supply. It means that
downward trend in the data of rainfall is just seasonal fluctuation or variability and not be
consider a trend of average pattern change. In study, rainfall shows non significant
declining trends in the last decade, but the overall analysis supports decreasing trend in
rainfall amount as result of rising temperature. The issue of detecting a climate change
impact on ensuing river flow data is very complex because the process of river flow is the
integrated result of several factors, such as rainfall amount and pattern, catchments area
and capacity of storage water and evaporation losses (Solow et al. 1989: Solow and
Broadus 1989: Lins and Slack 1999).
4.6. Environmental Effects of Climate Change in Pakistan

Like other countries of the world Pakistan is also experiencing direct and indirect
effects of climate change. A rise of 0.1oC per decade in the mean temperature, computed
through this study is comparable to other parts of the globe. The rising trend differential
and related impacts of climate change in Pakistan vary from zone to zone in the whole
study area. However, in general terms the most significant environmental effects include
extreme weather events such as heat waves in June 2007 and May 2010 in most parts of
Punjab, Sindh and Baluchistan, and heavy torrential rains across Pakistan in Jul. – Aug.
2010 and Aug – Sep 2011 leading to heaviest floods and land sliding in the history of the
country. Among other impacts felt or perceived are reduction in crop yields, local food
supplies, soil erosion and species extinctions. Although, Pakistan contributes very little in
GHG emission (up to 0.43%) but rated as the 12th most vulnerable country in the world in
terms effects of climate change (Maplecroft, 2007). Agriculture sector is one example
where reduced productivity, water shortage and different crop management practices are
further worsening the situation. Agriculture is the largest income and employment-
generating sector of Pakistan‘s economy with about two thirds population of the country
living in rural areas and earning livelihood directly or indirectly from agriculture and
cattle farming (Ali, 2000). Agriculture contributes about 24% of the GDP of Pakistan and
provides an employment to 48% of the labor force (EAW, 2003) where industries and
80% exports are almost agro-based, where the forest cover is low (4.5%) while
deforestation rate is 0.2- 0.4 % per annum (World Bank, 2006). The reports of Pakistan
102
Agricultural Research Council (2003) show that temperature increase in winter months
(Dec-Mar) have negative impacts on the big crops of food grains, sugarcane and fruits.
The production volume and quality have been adversely effected because of changes in
reproductive periods (Spring/late winter season), irrigation water requirements and its
supply for Kharif and Rabi crops, altering soil characteristics, and increasing the risk of
pests and diseases (TFCC, 2010).
Additional challenges imposed by climate change include changes in water
resource availability as demand for water has been steadily rising both in domestic and
agriculture sectors. In the pre-monsoon period (April-June) surface water requirements of
the country are mostly fulfilled by the perennial snow-fed rivers but due to shortening of
the winter season or increase in the minimum temperature snow/glacial melting is likely
to start early and may lead to reduced water availability at the time of heaviest demands.
This will have major impacts not just on hydropower generation but also on urban water
supplies and agricultural sectors (Roohi, 2004). Reduction in river flows will negatively
affect hydroelectric power generation. In a country already infested by the worst load
shedding of the history this impact will lead to an increase in fossil fuel combustion and
hence more GHG emissions. In addition, higher temperature, particularly in the summer
season has already increased demand for electricity for air-conditioning and refrigeration
in domestic and commercial sectors. This demand is likely to go further up and hence
necessitates additional generation capacity. According to the World Bank (2006),
Pakistan is among the 17 countries that are already facing water shortages and is among
the 36 countries where there is a serious threat of food crisis.
Farooqi et al (2005) have outlined several environmental effects of climate
change on Pakistan. Among many other effects such as strengthening of monsoon
circulation, intensification of tropical cyclones and diminishing flows in Indus basin, they
have also identified ecological impacts on wetlands and mangrove forests. Pakistan could
face the loss of mangrove forests - a vital source of fuel wood and seafood for local
consumption and exports.
Similarly, accelerated melting of glaciers and seasonal snow cover in Himalayas,
Karakoram and Hindu Kush ranges is threatening the natural habitat of rare animals such
as the Markhor and Ibex and hence, hundreds of rare plants and animals are in danger
103
of extinction (UNEP, 1995). The glaciers present in the Himalayan region are reported to be
melting faster than in any other part of the world and fears have been expressed that they
might disappear till 2035 (Rees and Collins 2004; WWF 2005; IPCC 2007). Further, the
studies of National Institute of Oceanography indicate that the sea level along the coast of
Pakistan has been rising approximately at 1.2 mm per year since 1960. Besides these,
weather related disasters hit the countries regularly like cyclones, hurricanes (Katrina and
Nargis), heat waves, super floods, droughts and intense rainfall. According to a World
Bank Report (2006), the country loses nearly $4.5 billion annually from environmental
disasters.
In most areas of the country, rainfall patterns have become very unreliable and
unpredictable, making it difficult for communities to make necessary arrangements for
their safety, crops and livestock. For intense, on the one hand the super flood of 2010 was
caused by very intense and extended rainfall that affected extensive parts of the country
but on the other hand droughts have become more evident with the worst drought hitting
the country during the period 1998-2001. According to PMD, ―in recent years there has
been a slow but steady change in monsoon pattern, particularly in locations where
Pakistan's major rainfalls concentrate. In past, monsoon rains fell most intensely over
central and northern Punjab, Kashmir and parts of Hazara and Swat areas but slowly and
steadily the concentration of rainfall has moved north and west to the Khyber
Pakhtunkhwa‖.
Furthermore, Margalla Hills would continue to suffer from some impacts related
to climate change. During summers most of the forest fire incidents have been reported
from the Margalla Hills. Although the cause of the fire is yet to be ascertained, however,
officials of the Capital Development Authority (CDA) attribute these fires to scorching
weather. Sadly summer fires at Margalla hills have become an annual feature, which not
only destroy the natural beauty but also harm the wildlife (The Nation, 2012).
From the above facts and perceived threats it becomes imperative to protect the
country from such adverse impacts of climate change through adaptation and mitigation. It is,
however, essential to continue scientific research not only to further ascertain climate change
in Pakistan but also on aspects of timing, frequency and intensity of the projected and
perceived threats so as to be able to predict and mitigate hazardous events before they strike.
104
Chapter # 5
CONCLUSION
Climate change, variability and its adverse impacts on the environment are
growing concern worldwide. The major cause of these adverse consequences is the
increase concentration of GHG. According to the IPCC (1996) the burning of fossil fuels
is the major factor of contributing the more atmospheric CO2 concentrations than double
their pre-industrial levels within the next 100 years. Similarly, there is the growing
consensus among the world‘s scientists that climate warming would effect precipitation
pattern and modify the hydrological cycle and induced stressed in freshwater supply.
In the context of climate change, the present study finds a closer examination of
the relationship between temperatures, rainfall and river discharge and analyzed the
climate fluctuation between these variables and associated river discharge rate over the
past 30 years from 1976-2005 in Pakistan. The dataset consist of 30 stations from all over
the country and divided into five different climatic zones according to the Kööpen based
classification for different microclimate of the world. All of the five climate zones are
different from each other in their climatic conditions, elevation from the sea level,
longitude, latitude, geography and in orography (e.g. mountainous study). A series of
statistical tests were applied on the data in different combination of ten, fifteen and thirty
years for making comparison in trends and find climate variability among hydro-
meteorological parameters such as minimum, maximum and mean temperatures and
rainfall with river flow rate. The statistical tests applied for data verification are
skewness, outlier and ANOVA. For equal variances of means Duncan test is applied and
for unequal variances Dunnett T3 is used. Besides theses analysis, the data is also
computed for monthly analysis and future trend of forecasting for the said variability,
using Ms Excel sheets and SPSS. The visualization of the data through mapping is
carried out by using the Arc GIS tool. The observed trend is then matched with global
studies for making comparison easy.
In the light of results and discussion (chapter 4), it is concluded that the trend of
temperature is showing increase such as 0.11oC per decade, especially in those zones
where big cities of country situated. From the analysis it is clear that all the climate zones
105
showed a consistent rising trend in minimum and maximum temperatures except those
zones situated in the North, North West and West of the country e.g. zone1 and zone3a.
Whereas those Zones (zone2, zone3b and zone4) situated in the plain, South West and
coastal areas showing great rising trend. The overall rising trend in minimum temperature
for the country is observed approximately (0.29 oC) for two time periods from (1976-
1990; 1991-2005) and (0.10 oC) per decade which is found non-significant at p-value
such as (p= 0.25>0.05). Similarly, for maximum temperature, p- value of the country is
observed significant (p= 0.02<0.05) with (0.36 oC) increase in temperature in two time
intervals and per decade (0.12 oC). The overall trend in mean temperature of the country
is found most significant (p=0.03< 0.05) in decadal analysis. Zone 3a (0.62 oC) and zone4
(0.55 oC) are recorded with highest value of mean maximum temperature difference into
two time periods. While rainfall data showing overall declining trend approximately (-
1.18mm) per decade and found most significant (p=0.04<0.05). In contrast to the mean
annual rainfall, river discharge data is also recorded with increasing trend, especially for
Indus River at Chilas and Skardu Stations, Swat Rivers and Astor River. Hence, it is clear
from the river basin analysis that temperature has a profound influence on the
snow/glacial melt in comparison to the mean monthly rainfall. This is because majority of
the ensuing rivers derived from zone1 are faded by the snow and glacial melt as most
stations of this zone receive a lot of snowfall in winter (Dec-Feb) and the pre- monsoon
dry season (April-May) Furthermore, models used for trend forecasting show overall
increasing trend in mean minimum, maximum and mean temperature of the whole study
area till 2030 while rainfall data depicts declining trend because of drought from 1998-
2001. By excluding the drought period from the real time series of the mean annual
rainfall data, the observed data forecasts an upward trend up to 50mm till 2030. There
will be no forecasting model used for river discharge data because of short record length.
However, different studies on climate change scenario revealed that short and long term
changes in the climate may lead to environmental challenges in the future, like water
shortage, dryness of land, change in land use pattern like shifting of plants and animals
on elevation, threats to biodiversity and ecosystems and effects on agricultural practices
etc. Similarly increase in population growth and associated demands for fresh water, food
and energy would be a matter of concern for the future climate scenario. Climate change
106
is a global issue but its impacts can be felt severely in South Asian countries, especially
in Pakistan, with weak economy and vulnerability to extreme climate events like
droughts, sever cold and heat waves, and flash flooding. Pakistan economy is mainly
based on agriculture sector. Due to increased temperature and shortening of the winter
season, the snow will start early melting which can greatly affect the water management
sector by reducing the average flow of snow fed rivers, urban water supply and
agriculture sector. Although this change (0.11oC/ decade) in mean temperature is not
found too much, but if steps advised by IPCC in its 4th assessment report, are not taken at
the right time for the mitigation of the excessive anthropogenic activities, the situation
can be worse in the next two to three decades. To minimize or reduce any actual or
potential adverse effect of climate change, key consideration should be given in disaster
mitigation strategies and developmental schemes like in planning, designing, and
implementation, in terms of minimizing economic losses.
107
REFERENCES
Adams, W.T., Hipkins, V. D., Burczyk, J., and Randal, W. K., 1997. Pollen
contamination trends in a maturing Douglas-fir seed orchard. Canadian of forestry
research, 27: 131-134.
Afzaal, M., Haroon, M. A., and Zaman, Q., 2009. Inter- decadal Oscillations and
the Warming Trend in the Area-Weighted Annual Mean Temperature of Pakistan.
Pakistan Meteorology, 6(11).
Aguilar, E., I. Auer, M. Brunet, T.C. Peterson and J. Wieringa, 2003. Guidelines
on Climate Metadata and Homogenization. WCDMP-No. 53, WMO-TD No.
1186, World Meteorological Organization, Geneva, 55.
Ahemad, S., Bari, A., and Muhammed. A., 2003. Climate change and water
resources of Pakistan: Impacts, vulnerabilities and coping mechanism.
Proceedings of the Year-End Workshop on ―Climate Change and Water
Resources in South Asia‖ Kathmandu, Nepal, Muhammed, A., (Eds.), 7-9: 138-
201.
Alam, M., T., 2009. Impact of climate change over Pakistan, Environmental
Prediction into the Next Decade: Weather, Climate, Water and the Air We breath.
Presentation, Incheon, Republic of Korea, 16-17.
Ali, S., 2000. Productivity Growth in Pakistan‘s Agriculture 1960-1996. Ph. D
dissertation, Department of Economics, Simon Fraser University, Canada.
Allen, M. R., and Ingram, W. J., 2002. Constraints on future changes in climate
and the hydrological cycle. Nature, 419: 224–232.
Alpert, P., Ben-Gai, T., Baharad, A., Benjamini, Y., Yekutieli, D., Colacino, M.,
Diodato, L., Ramis, C., Homar, V., Romero, R., Michaelides, S., and Manes, A.,
2002. The paradoxical increase of Mediterranean extreme daily rainfall in spite of
decrease in total values. Geophysical Research Letters, 29. DOI:
10.1029/2001GL013 554.
Archer, D. R., and Fowler, H. J., 2006. Conflicting signals of climate change in
the upper Indus basin. Climate, 19: 4276 – 4293.
108
Archer, D., 2003. Contrasting hydrological regimes in the upper Indus Basin.
Hydrology, 274 : 198-210.
Arnell, N. W., 1999a. The Effect of Climate Change on Hydrological Regimes in
Europe. In: A Continental Perspective, Global Environment Change, 9: 5–23.
Arnell, N. W.: 1999b. Climate Change and Global Water Resources. Global
Environment Chang, 31– 49.
Association of British Insurers (ABI), 2005. Financial risks of climate change. Summary
Report, 40. http://www.abi.org.uk.
Awan, S. A., 2002. The Climate and Flood Northern Areas of Pakistan, Risk
Potential of Commission on Science and Technology for Sustainable
Development in the South. (http://www.comsats.org.pk/main_introduction.htm).
Ayers, M.A., Wolock, D.M., McCabe, G. J., Hay, L.E., and Tasker, G. D., 1994.
Sensitivity of water resources in the Delaware River basin US Geological Survey
Water-Supply. 2422 US, Geological Survey, Reston, VA.
Barry, R.G., 2002. Mountain climate change and cryospheric
responses: A review. In Berger, T., (Eds) “Mountain of the World.”
Proceedings of the World Mountain Symposium (WMS 2001), 7pp. CD-
ROM. Swiss Agency for Development and Cooperation, Bern,
Switzerland.
http://www.wms2001.ch/cd/publications/publication_10.html
Bart -Nijssen, G., M. O‘Donnell, Alan, F., Hamlet, Dennis, P., and Lettenmaier,
C., 2001. Hydrologic sensitivity of global rivers to climate change climate
change. Kluwer Academic Publishers. Printed in the Netherlands, 50: 143-175.
Batima, P., 2003. Climate change: pasture–livestock. Synthesis Report. Potential
Impacts of Climate Change, Vulnerability and Adaptation Assessment for
Grassland Ecosystem and Livestock Sector in Mongolia. Admon Publishing,
Ulaanbaatar, 36–47.
Batima, P., L. Natsagdorj, P. Gombluudev and B. Erdenetsetseg, 2005a. Observed
climate change in Mongolia. AIACC Working Paper, 13: 25.
109
Bhutiyani, M. R., Kale, V. S., Pawar, N. J., 2007. Long term trends in maximum,
minimum and mean annual air temperature across the North-western Himalayan
during the 20th century. Climate Change, 85(1-2): 159-177.
Boer, R., and A. Faqih, 2004. Current and Future Rainfall Variability in Inland
Vulnerability in Watershed Areas and Communities in Southeast Asia. START
Secretariat, Washington, District of Columbia.
Bonan, G., 2002. Ecological Climatology. In: Concepts and Applications,
Cambridge University Press, Cambridge, 678.
Bonsal, BR, Zhang X, Vincent LA, Hogg WD., 2001. Characteristics of daily and
extreme temperatures over Canada. Climate, 14:1959-1976.
Bosilovich, M. G., Schubert, S., and Walker, G., 2005. Global changes of the
water cycle intensity. Climate, 18: 1591–1607.
Braganza, K., D.J., Karoly, and Arblaster, J.M., 2004. Diurnal temperature range
as an index of global climate change during the 20th century. Geophysical
Research Letter, 31. DOI: 10.1029/ 2004GL019998.
Bruce, J., Frome, M., Haites, E., Janzen, H., lal, R., and Paustian, 1998. Carbon
sequestration in soil, soil water conserves, 54: 382-389.
Brunetti, M., Maugeri, M., Nanni, T., 2001. Changes in total precipitation, rainy
days and extreme events in northeastern Italy. International Journal of
Climatology, 21: 861–871
Cavallo, E., and Ilan Noy, 2010. The economics of Natural Disasters, Survey,
IDB working paper series, No.IDB-WP-124, Inter-American Development Bank.
Chaudhary, Q. Z., 1994. Pakistan summer monsoon rainfall‘s association with
global and regional circulation features and its seasonal prediction. Proceeding of
International Conference on Monsoon Variability and Prediction, Triest, Italy, 9-
13: 136-145.
 Chaudhry, Q. Z., 2001. History‘s Worst Drought Hit Pakistan. Farming Outlook.
Soil, Agriculture, Fertilizer and Environment (SAFE) Foundation, Pakistan.
 Chaudhry, Q. Z., 2002. Drought Monitoring and Early Warning System.
Proceedings of the SAARC Workshop on Drought and Water Management
110
Strategies (September 16-18, 2002 Lahore-Pakistan). Pakistan Council of
Research in Water Resources, Islamabad, Pakistan.
Chaudhry, Q. Z., and Rasul, G., 2007. Development of a meso-scale convective
system over the foothills of Himalaya into a severe storm. Mountains witnesses of
Global Changes, Elsevier, Chapter (33): 301-311.
Christy, J.R., R.W. Spencer, and W.D. Braswell. 2000. MSU Tropospheric
Temperatures: Dataset Construction and Radiosonde Comparisons. Atmospheric
and Oceanic Technology, 17: 1153-1170. A reanalysis of the MSU channel 2
tropospheric temperature record. Climate, 16: 3650-3664.
Collins, D. A., Della-Marta, P.M., Plummer, N., and Trewin, B.C., 2000. Trends
in annual frequencies of extreme temperature events in Australia. Australian
Meteorological Magazine, 49: 277-292.
Cruz, R.V., Harasawa, H., Lal, M., Wu, S., Anokhin, Y., Punsalmaa, B., Honda,
Y., Jafari, M., Li, C., and Huu Ninh, N., 2007. Impacts, adaptation and
vulnerability. In: Climate change. Contribution of Working Group II to the 4th
Assessment Report of the Intergovernmental Panel on Climate Change. M.L.
Parry, O.F. Canziani, J. P. Palutikof, P.J. Vander Linden, C.E. Hanson, (Eds).
Cambridge: Cambridge University Press, 469-506.
Dai, A., K.E. Trenberth, and T.R. Karl, 1999. Effects of clouds, soil moisture,
precipitation, and water vapor on diurnal temperature range. Climatology, 12:
2451-2473.
Dash, S. K. and Hunt, J. C. R., 2007. Variability of climate change in India,
Current Science, 93(6): 782-788.
Dash, S.K., Jenamani, R.K., Kalsi, S.R., Panda, S.K., 2007. Some evidence of
climate change in 20th century India. Climatic Change, 85 (3-4): 299-321.
Diffenbaugh, S., Jeremy, S. Pal, Rober,t J. Trapp, and Giorgi,F., 2005. Fine-scale
processes regulate the response of extreme events to global climate change.
Proceedings of the National Academy of Sciences (PNAS), 102 (44): 15774-
15778. DOI:10.1073/pnas.0506042102
Duong, L.C., 2000. Lessons from severe tropical storm Linda. Workshop report:
The Impact of El Niño and La Niña on Southeast Asia, Hanoi.
111
Easterling, D.R., B. Horton P.D., Jones, T.C. Peterson, T.R. Karl, M.J. Salinger,
D. Parker, V. Razuvayev, N. Plummer, P. Jamason, and C.K. Folland, 1997.
Maximum and minimum temperature trends for the globe. Science, 277: 364-367.
EAW, 2002-03. Pakistan Economic Survey, Economic Advisory Wing, Finance
Division, Government of Pakistan.
Erbach, J. and Gaudet, J., 1998. Urbanization Issues and Development in Sub-
Saharan Africa, USAID Africa Bureau, Office of sustainable development.
Fan, D.D,. and C.X. Li, 2005. Complexities of Chinese coast in response to
climate change. Advance in Research on Climate Change,1: 111-114.
Farooq, A. B. and A. H. Khan, 2004. Climate change perspective in Pakistan: In:
Proceeding of the workshop on capacity building on global change research. June
8-10, Islamabad, Pakistan.
Farooqi, A.B., Khan, A.H. and Mir, H., 2005. Climate change perspective in
Pakistan. Pakistan Journal of Meteorology, Vol. 2(3): 11-21.
Ferguson, R. I., 1984. Sediment load of the Hunza River. In: Miller, K.J. (Ed.).
The International Karakoram project. Cambridge university, 2: 580-598.
Folland, C.K., Miller, C., Bader, D., Crowe, M., Jones, P., Plummer, N.,
Richman, M., Parker, D.E., Rogers, J. and Scholefield, P., 1999. Workshop on
indices and indicators for climate extremes. Asheville, NC, USA, 3-6 June 1997.
Breakout group C: Temperature indices for climate extremes, Climatic Change,
42: 31-43.
Folland, C.K., N. Rayner, P. Frich, T. Basnett, D. Parker and B. Horton, 2000.
Uncertainties in climate data sets - a challenge for WMO. WMO Bull., 49: 59-68.
Food and Agriculture Organization (FAO), 1987. Pakistan‘s Experience in
Rangeland Rehabilitation and Improvement, United Nation Organization, 70.
Food and Agriculture Organization (FAO), 2004. The State of Food Insecurity in
the World, United Nation, Rome, Italy, 32-34.
Gan, T. Y., 1998. Hydro climatic trends and possible climatic warming in the
Canadian Prairies. Water Resources, 34(11): 3009–3015. DIO:10.1029/98
WR01265.
112
GCOS, 2005a. GCOS Regional Action Plan for South and South-East Asia.
http://www.wmo.int/pages/prog/gcos/documents/GCOS_ESEA_RAP_FINALDR
AFT_ Sept2005.pdf
GCOS, 2005b. GCOS Regional Action Plan for South and South-West Asia.
http://www.wmo.int/pages/prog/gcos/documents/GCOS_SSWA_RAP_FINALD
RAFT_ Sept 2005.pdf
Gleick, E H., 1986. Methods for Evaluating the Regional Hydrological Impact of
Global Climatic Changes. Hydrology, 88: 97-116.
Gleick, E. H., 1987. Regional Hydrologic Consequences of Increases of
Atmospheric CO2 and Other Trace Gases. Climatic Change, 110: 137-161.
Gleick, P. H., 1999. Studies from the Water Sector of the National Assessment.
American. Water Resources Associates, 35: 1297-130.
Gleick, P.H., 1989. Climate change, hydrology, and water resources. Review of
Geophysics, 27: 329-344.
Groisman, P.Ya., Knight, R.W., Karl, T.R., Easterling, D.R., Sun, B. and
Lawrimore, J.H., 2004. Contemporary changes of the hydrological cycle over the
cotiguous United States. In: Trends derived from in situ observations.
Hydrometeorology,5: 64-85.
Gruza, G. and E. Rankova, 2004. Detection of changes in climate state, climate
variability and climate extremity. Proc. World Climate Change Conference,
Moscow, Y. Izrael, G. Gruza, S. Semenov and I. Nazarov, (Eds)., Institute of
Global Climate and Ecology, Moscow, 90-93.
Haines, A. A., McMicael, and Epstein, P. R., 2000. Environment and health.
Global climate change and health. Canadian Medicine Association, 16: 729-734.
Hajat, S, Ebi, K.L., Kovats, R.S., 2003. The human health consequences of
flooding in Europe and the implications for public health: A review of the
evidence. Applied Environmental Science and Public Health, 1 (1): 13-21.
Hansen, J., R. Ruedy, M. Sato and K. Lo., 2006. GIS and Surface Temperature
Analysis, Global Temperature Trends: 2005 Summation, NASA's Goddard
Institute for Space Studies.
113
Hoque, S., 1995. Flood forecasting and warning processes for river system in
Bangladesh. In: Proceedings of the International Seminar on Water Induced
Disasters, Kathmandu: Department of Water Induced Disaster Prevention, 227-
243.
Houghton, J. T., Y. Ding, D. J. Griggs, M. Noguer, P. J. Vander Linden, X. Dai,
K. Maskell, and C. A. Johnson, (Eds.), Climate Change, 2001. The Scientific
Basis, Cambridge University Press, New York, 881.
Hussain, S. S., Mudasser, M., Sheikh, M. M., and Manzoor, N., 2005. Climate
change and variability in mountain regions of Pakistan implications for water and
agriculture Pakistan. Meteorology, 2(4).
Intergovernmental Panel on Climate Change, ―Summary for Policymakers,‖ in
Climate Change 2007. The Physical Science Basis, Contribution of
Working Group I to the Fourth Assessment Report of the Intergovernmental Panel
on Climate Change, S. Solomon et al. (Eds), Cambridge, UK: Cambridge
University Press.
International Federation of the Red Cross and Red Crescent Societies, 1999.
World Disasters Report, Geneva, 7-26
IPCC, 1996. Climate Change 1995. The Science of Climate Change, Cambridge
University Press, Cambridge U.K.
IPCC, 2001. Climate change. The IPCC 3rd assessment report. The scientific
basis, Houghton, J.T., Ding, Y., Griggs, D., Noguet, M., Vander Linden, P., Dai,
X., Maskell, K., Johnson, C.A. (Eds). Cambridge: Cambridge, University Press.
Islam, N., Rafiuddin, M., Ahsanuddin and Kollic, R.K., 2007. Calibration of
PRECIS in employing future scenarios in Bangladesh. International Journal of
Climatology Royal Meteorological Society.
Iwashima, T., Yamamoto, R., 1993. A statistical analysis of the extreme events:
long-term trend of heavy daily precipitation. Journal of the Meteorological
Society of Japan 71: 637–740.
Izrael, Y.A., and Anokhin, 2001. Climate change impacts on Russia. Integrated
Environmental Monitoring. Nauka, Moscow, 112-127.
114
Jones, P.D., Horton, E. B., Folland, C.K., Hulme, M., Parker, D.E. and Basnett,
T.A., 1999. The use of indices to identify changes in climate extremes. In: Karl,
T.R., Nicholls, N. and Ghazi, A. (Eds), Weather and climate extremes: changes,
variations and a perspective from the insurance industry. The Netherlands:
Kluwer Academic Publications
Jones, P.J., 2001. Instrumental temperatures change in the context of the last 1000
years. In: Brunet. India, M. and Lopez Bonillo, D., (Eds). Detecting and modeling
regional climate change. Berlin, Springer-Verlag.
Jonnes, P. D., and A. Moberg., 2003. Hemispheric and large scale surface air
temperature variation. In: An extensive revision and an update to 2001. Climate,
16: 206-223.
Jonnes, P. D., and Moberg, A., 2003. Hemispheric and large scale surface air
temperature variation: An extensive revision and an update to 2001. Climate, 16:
206-223.
Kajiwara, M., Oki, T., and Matsumoto, J., 2003. Inter-annual variability of the
frequency of severe rainfall in the past 100 years over Japan. Extended abstract,
biannual meeting of the Meteorological Society of Japan.
Kalayci, S., and Kahya, E., 2004. Stream flow trends in Sakarya basin. ARI. The
Bulletin of the Istanbul Technical University, 53: 79–84.
Kanai, S., Oki, T., and A. Kashida, 2004. Changes in hourly precipitation at
Tokyo from 1890 to 1999. Meteorology Society. Japan, 82: 241-247.
Karl, T. R., Jones, P. D., Knight, R. J., Kukla, G., Plummer, N., Razuvayev, V.,
Gallo, K.P., Lindseay, J., Charlson, R. J., and Peterson, T. C., 1993. Asymmetric
trends of daily minimum and maximum temperature. American Bulletin of
Meteorological Society.
Karl, T., and K. Trenberth, 1999. The human impact on climate. Scientific
American. December, 100-105.
Karl, T.R. and Knight, R.W., 1998. Secular trends of precipitation amount,
frequency, and intensity in the United States. Bulletin of American Meteorology
Society, 79: 231-241.
115
Karl, T.R., and Riebsame, W.E., 1989. The impact of decadal fluctuations in
mean precipitation and temperature on runoff. A sensitivity study over the United
States. Climate Change, 15: 423–447.
Karl, T.R., Knight, R.W., Plummer, N., 1995. Trends in the high frequency
climate variability in the twentieth century. Nature 377: 217–220
Karl, T.R., Miller, C. D., and Murray, W. L., (Eds), 2006. In Temperature Trends
in the Lower Atmosphere: Steps for Understanding and Reconciling Differences.
A Report by the Climate Change Science Program and the Subcommittee on
Global Change Research, Washington, DC.
Kaser, G. 2004. Modern glacier retreat on Kilimanjaro as evidence of climate
change: observations and facts. International journal of Climatology
Kawahara, M. and Yamazaki, N., 1999. Long-term trend of incidences of extreme
high or low temperatures in Japan. Bi-annual meeting of the Meteorological
Society of Japan.
Kazi, S. A. and Khan, M. L., 1951. Variability of rainfall and its bearing on
agriculture in the arid and semi-arid zones of West Pakistan. Pakistan
Geographical Review, 6(1): 40-63.
Kenney, J. F. and Keeping, E. S. 1962. Skewness. Chapters (7.10) in
Mathematics of Statistics, (3rd Ed). Princeton, N. J, Van Nostrand, 100-101.
Khan, F. K., 2002. Pakistan Geography, Economy and People. Oxford University
Press, Karachi.
Khan, J. A., 1993. The Climate of Pakistan, Rehber Publishers, Karachi.
Khan, T. M. A., Singh, O.P., and Rahman, S., 2000. Recent sea level and sea
surface temperature trends along the Bangladesh coast in relation to the frequency
of intense cyclones. Marine Geology, 23: 103-116.
Kiktev, D., Sexton, D. M. H., Alexander, L., and Folland, C. K., 2003.
Comparison of modeled and observed trends in indices of daily climate extremes.
Climate, 16: 3560–3571.
Kirk, R. E. 1982. Experimental design: Procedures for the behavioral sciences.
(2nd Ed.), Belmont, CA: Brooks/Cole.
116
Knowlton, K., P.H. , Lynn, B., Richard, A., Goldberg, M.S., Rosenzweig, C.,
Hogrefe, C., Rosenthal, J. K., and Patrick L. Kinney, 2007. Projecting Heat-
Related Mortality Impacts under a Changing Climate in the New York City
Region. American Journal of Public Health, 97: 2028-2034.
DOI:10.2105/AJPH.2006.102947.
Kondratyev, K., 1998. Observing Global Climate Change. Taylor and Francis Ltd,
1-5
Kostopoulou, E., and Jones, P. D., 2005. Assessment of climate extremes in the
Eastern Mediterranean. Meteorology and Atmospheric Physics, 89: 69–85. DOI:
10.1007/s00703-005-0122-2.
Krakauer, N. Y., and Fung, I., 2008. Is stream flow increasing? Trends in the
Coterminous United States, European Geosciences Union.
Kumar A, Yang, F., Goddard, L., and Schubert, S., 2003. Differing trends in
tropical surface temperatures and precipitation over land and oceans. Climate, 17:
653–664.
Kureshy, K. U., 1998. Geography of Pakistan. National Book Service Lahore,
Pakistan
Labat, D., Godderis, Y., Probst, J. L., and Guyot, J. L., 2004. Evidence for global
runoff increase related to climate warming. Advances in Water Resources, 27:
631–642.
Lai, M., Whettori, P.H., Pittodi, A.B., and Chakraborty, B., 1998. The greenhouse
gas induced climate change over the Indian Sub-continent as projected by GCM
model experiments, Terrestrial. Atmospheric and Oceanic Sciences, TAO, 9(4):
663-66
Lal, M., 2001. Tropical cyclones in a warmer world. Current Society. India, 80:
1103- 1104.
Lal, M., 2002. Global climate change: India‘s monsoon and its variability. Final
Report under Country Studies Vulnerability and Adaptation Work Assignment
with Stratus Consultant‘s Contract of the U.S. Environmental Protection Agency,
58.
117
Lal, M., 2003. Global climate change: India‘s monsoon and its variability.
Environmental Studies and Policy,6: 1-34.
Lettenmaier, D. P., and Gan, T.Y., 1990. Hydrologic sensitivities of the
Sacramento-San Joaquin River basin. In: Caldornia, to global warming. Water
Resources Resident, 26: 69-86.
Lettenmaier, D. P., Wood, E. F., and J. R. Wallis, 1994. Hydro-Climatological
trends in the continental United States, 1948– 1988.Climatology, 7: 586– 607.
Lins, H.F., Slack, J.R., 1999. Stream flow trends in the United States. Geophysics
Research Letters, 26: 227–230.
Liu, W.T., X. Xie and W. Tang, 2005. Proceedings of the international
symposium in cloud- prone and rainy areas remote sensing (CARRS). Chinese
University of Hong Kong.
Luce, C. H., and Z. A. Holden, 2009. Declining annual stream flow distributions
in the Pacific Northwest United States from 1948–2006. Geophysical Research
Letters, 36: 16401. DOI: 10.1029/2009GL039407.
Luo, Q. and Lin, E., 1999, Agricultural vulnerability and adaptation in developing
countries: the Asia- Pacific region. Climatic Change, 43: 729-743.
Maplecroft , 2007. Climate Change Vulnerability Index, Country-by-country risk
analysis. Bath: Maplecroft.
Mandira S. Shrestha, S., Bajracharya, R., and Mool, P., 2008. Satellite Rainfall
Estimation in the HinduKush, Himalayan

Region. The International Centre for Integrated Mountain Development (ICIMOD) ,
Kathmandu.
Mann, M.E., and Jones, P.D., 2003. Global surface temperature over the past two
millennia. Geophysical Research Letters, 30(15).
Manton, M. J., Della-Marta, P.M., Haylock, M.R., Hennessy, K. J., Nicholls,
K.J., Chambers, L.E., Collins, D.A., Daw, G., Finet, A., Gunawan, D., Inape,
K., Isobe, H., Kestin, T.S., Lefale, P., Leyu, C. H., Lwin, T., Maitrepierre, L.,
Ouprasitwong, N., Page, C.M., Pahalad, J., Plummer, N., Salinger, M.J.,
118
Suppiah, R., Tran, V.L., Trewin, B., Tibig, I. and Yee, D., 2001. Trends in
extreme daily rainfall and temperature in Southeast Asia and the South Pacific:
1961–1998. International Journal of Climatology, 21: 269–284 (2001).DOI:
10.1002:joc.610
Mason, S. M., Waylen, P.R., Mimmack, G.M., Rajaratnam, R., Harrison, J.M.,
1999. Changes in extreme rainfall events in South Africa. Climatic Change 41:
249–257
Mastny, L., 2000. Melting of Earth‘s Ice Cover Reaches New High. World Watch
Institute, Washington, DC. www.Worldwatch.org/ alerts/000306.html.
Mather, J.R., and Feddema, J., 1986. Hydrologic consequences of increases in
trace gases and C02 in the atmosphere, in Effects of changes in stratospheric
ozone and global climate. Washington, DC, U.S. Environmental Protection
Agency. Climate change, 3: 251-271.
McCabe, G. J., and Ayers, M.A., 1989. Hydrologic effects of climate change in
the Delaware kver basin. Water Resource Bulletin, 25: 1231-1242.
McCabe, G.J. and Wolock, D.M., 2002. A step increase in stream flow in the
conterminous United States: Geophysical Research Letters, 29: 2185-2188.
Mikdat Kadioğlu, 1998. Trends in surface air temperature data over Turkey. The
Royal Meteorological Society. DOI: 10.1002/ (SICI) 1097-
0088(199704)17:5<511: AID-JOC130>3.0.CO;2-0.
Mimura, N., Nurse, L., Mclean, R.F., Agard, J., Briguglio, L., Lefale, P., Payet, R.
and Sem, G., 2007. Small islands. In Parry, M.L., Canziani, O.F., Palutikof, J.P.,
van der Linden, P.J. and Hanson, C.E. (Eds). Climate Change 2007: Impacts,
Adaptation and Vulnerability. Cambridge. Cambridge University Press, 687-716.
Min, S.K., W.T. Kwon, E.H. Park and Y. Choi, 2003. Spatial and temporal
comparisons of droughts over Korea with East Asia. International Journal
Climatology, 23: 223-233.
Mirza, E. M.Q., 2002. Global warming and changes in the probability of
occurrence of floods in Bangladesh and implications. Global Environment.
Chang, 12: 127-138.
119
Mirza, M.M.Q., 2003. Climate change and extreme weather events: Can
developing countries adopt? Climate policy, 3(3).
Mishers, A., 2004. Global Climatic Change. E-35/103, Jawahar Park Laxmi
Nagar, Dehli-110092, 1-25, 197-195
Mukhopadhyay, 1998. Severe heat wave over the Indian subcontinent in 1998 in
perspective of global climate. Current Society,75: 1308-1315.
Munich Report, (Eds.) 2004. Mega cities – Mega risks: Trends and challenges
for insurance and risk management. Munich, Germany
Munich reports 2002. An annual review of natural catastrophes. Munich
Reinsurance Company Publications.
Muzaharul, A., and Rabbani, G., 2007. Vulnerabilities and responses to climate
change for Dhaka. Environment and urbanization, 19(1): 81-97.
Nadia, R., and Islam, S., 2006. Future change in the frequency of warm and cold
spells durations over Pakistan simulated by the PRECIS regional climate model,
Young Scientist Conference, Beijing, China.
Natsagdorj, L., P. Gomboluudev and P. Batima, 2005. Climate change in
Mongolia. Climate Change and its Projections, P. Batima and B. Myagmarjav,
(Eds), Admon publishing, Ulaanbaatar, 39-84.
Nayava, J. L., 1980. Rainfall in Nepal. Himalayan Review, 12: 1–18.
NEF, 2001. The End of Development?.
http://www.neweconomics.org/gen/uploads/e_o_d.pdf
Nemec, J., and Schaake, J.S., 1982. Sensitivity of water resources systems to
climate variation. Hydrology Science, 27: 327-343.
New, M., Todd, M., Hulme, M, Jones, P., 2001. Precipitation measurements and
trends in the twentieth century. Climatology,21: 899–1922
Noah, D., Hall and Bret, B. Stuntz, 2007. How Climate Change Impacts Great
Lakes Water Resources, Great Lakes Natural Resource Center. National Wildlife
Federation 213 West Liberty St., Suite 200, Ann Arbor, MI 48104-1398.
Office of the High Representative for the Least Developed Countries, Landlocked
Developing Countries and Small Island Developing States(Un-ohrlls), 2009. The
impact of climate change on the development prospects of the least developed
120
countries and small island developing states. Room s-770; headquarters New
York, NY 10017, USA.
Olsen, J., Stedinger, J., Matalas, N., and Stakhiv, E.: 1999. Climate Variability
and Flood Frequency Estimation for the Upper Mississippi and Lower Missouri
Rivers. American Water Resources. Assoc, 35: 1509–1523.
Otto-Bliesner, B. L., Shawn J. Marshall, Jonathan T. Overpeck, Gifford, H.,
Miller, Aixue Hu, and CAPE Last Interglacial Project members 2006.
Simulating Arctic Climate Warmth and Ice field Retreat in the Last Inter-
glaciations. Science, 1751-1753. DOI:10.1126/science.1120808.
Overpeck, J., M. Sturm, J. Francis, D. Perovich, 2005. Arctic system on trajectory
to new, seasonally ice-free state. National Snow and Ice Data Center press
release, 86:309-313.
Oxfam Report on Climate Change, 2011. Six months into the floods resetting
Pakistan‘s priorities through reconstruction. 144 Oxfam Briefing Paper.
Palutikof, J.P., 2001.Validation of climate model outputs. In: Remote Sensing

and Climate Change: The Role of Earth Observation, A.P. Cracknell (Eds),
Praxis, Chichester, 185-203.
Pant G.B., and Rupa Kumar, K., 1997. Climates of South Asia. Johan Wiley and
Sons Ltd, Baffins Lane, Chichester, West Sussex PO191IUD, England, 219-224.
Parry, M. L. 1990. Climate Change and World Agriculture. London: Earth scan,
157.
Parthasarathy, G and Rajeevan, M., 2006. Trends in the rainfall pattern over India.
National Climate Centre India Meteorological Department Pune. India. Research
report No: 2/2006
Peduzzi, P., 2005. What is a disaster - Asking for disaster? Environment and
Poverty Times, working in collaboration with UNEP, highlighting IFRC World
Disasters Report for 2004, 03.
Philippine Atmospheric, Geophysical and Astronomical Services Administration
(PAGASA), 2001. Documentation and analysis of impacts of and responses to
extreme climate events. Climatology and Agrometeorology Branch Technical
121
Paper No. 2001-2, Philippine Atmospheric, Geophysical and Astronomical
Services Administration, Quezon City, 55.
Probst, J. L. and Tardy, Y. 1987. Long range stream flow and world continental
runoff fluctuations since the beginning of this century. Hydrology, 94: 289–311,
DOI: 10.1016/0022-1694(87)90057-6.
Probst, J.L. and Tardy, Y. 1989. Global runoff fluctuations during the last 80
years in relation to world temperature change. Science, 289: 267–285.
Przybylak, R., 2000. Diurnal temperature range in the Arctic and its relation to
hemispheric and Arctic circulation patterns. Climatology, 20: 231- 253.
Rashid, N., 2008. Climate changes in Peshawar and Abbott bad; issues, sources
and recommendation, BES thesis report. Department of Environmental Sciences
University of Peshawar, 24, 37, 57.
Rasul, G. and Chaudhry, Q.Z., 2006. Global Warming and Expected Snowline
Shift along Northern Mountains of Pakistan. Proceedings of 1st Asia climatic
Symposium. Yokohama, Japan.
Rees, G. and D. N. Collins, 2004. An assessment of the Potential Impacts of
Deglaciation on the Water Resources of the Himalaya. Technical Report, DFID
KAR Project No. R7890: Snow and Glacier Aspects of Water Resources
Management in the Himalayas (SAGAR MATHA), Centre for Ecology and
Hydrology, Oxfordshire, UK.
Rodo, X., 2003. Global Climate Current Research and Uncertainties in the
Climate System. University of Barcelona Climate Research Group Baldiri Rexac,
Torre D 08032 Barcelona Spain, 3-6.
Roohi, R., 2004. Farm mechanism options under climate change scenarios in
Pakistan, principle scientific officer; water resource research institute, National
agricultural centre Islamabad, Pakistan.
Rose, S., 2007. Decadal rainfall and runoff trends in the southeastern United
States. Georgia Water Resources Conference at the University of Georgia.
Rosenzweig, C., and Hillel, D., 1995. Potential Impacts of Climate Change on
Agriculture and Food Supply. Consequences, 1(2).
122
Roy, S. S., and Balling, R. C., 2004. Trends in extreme daily rainfall indices in
India. Climatology, 24: 457– 466.
Ruosteenoja, K., Carter, T.R., Jylhä, K., and. Tuomenvirta, H., 2003. Future
climate in world regions: an inter-comparison of model-based projections for the
new IPCC emissions scenarios. The Finnish Environment 644, Finnish
Environment Institute, Helsinki, 83.
Rupa Kumar K., Krishna Kumar, K. and Pant G.B., 1994. Diurnal Asymmetry of
surface temperature trends over India. Geophysical Research Letters, 21: 677-680.
Russell, S., Easterling, D.R., and Gleason. 2004. Maximum and minimum
temperature trends for the globe. NOAA National Climatic Data Center,
Asheville, North Carolina.
Ryoo, S.B., Kwon, W.T., and Jhun, J.G., 2004. Characteristics of wintertime
daily and extreme temperature over South Korea. Climatology, 24: 145-160.
Schaake, J.S., 1990. From climate to flow. In: Waggoner PE (Eds), Climate
change and US water resources. John Wiley, New York, 177-206.
Seshu, D.V., and Cady, E., 1984. Response of rice to solar radiation and
temperature estimated from international yield trials. Journal of Crop Science, 24:
649-654.
Shah, S.M. 2008. Impact of climate change on minimum temperature trends of
Pakistan. Ms Thesis report, Liverpool.
Shahid, S., 2010. Rainfall variability and the trends of wet and dry periods in
Bangladesh. International journal of climatology. 30: 2299–2313 DOI:
10.1002/joc.2053
Sheikh, M. M., 2005. Trends in Extreme Daily Rainfall and Temperature Indices
over South Asia. 1-year concluded project (APN2005-10-NSY) sent to Asia-
Pacific Network for Global Change Research (APN) in December, 2008 for
―Development and Application of Climate Extreme Indices and Indicators for
Monitoring Trends in Climate Extremes and their Socio-Economic Impacts in
South Asian Countries‖.
Shirazi and Ali, S., 2009. Patterns of Urbanization in Pakistan: A Demographic
Appraisal. Division, Department of Economic and Social Affairs, United Nations,
123
2009, ―Executive Summary‖. World Urbanization Prospects: The 2007 Revision.
New York: Population Division, Department of Economic and Social Affairs,
February 26, 2008.
Shrestha, A. B., Wake, C. P., Dibb, J. E. and Mayewski, P. A., 2000. Precipitation
fluctuations in the Himalaya and its vicinity. An analysis based on temperature
records from Nepal. Climate, 20: 317-327.
Shrestha, A. B., Wake, C. P., Mayewski, P. A., and Dibb, J. E., 1999. Maximum
temperature trends in the Himalaya and its vicinity. An analysis based on
temperature records from Nepal for the period 1971-94. Climate, 12: 2775-2787.
Shrestha, K. L., 2005. Global change impact assessment for Himalayan mountain
regions for environmental management and sustainable development. Global
Environmental Research, 9(1): 69-81.
Siddiqui, K. M., 2001. Review of Wood fuel Production and Marketing Reports
in Pakistan. In: Wood fuel production and Marketing in Pakistan National
Workshop, Sindh, Pakistan, 20-21.
Singh, N, and Sontakke, N. A., 1996. Climate variability over Pakistan and its
relationship to variation over the India region. In: Climate Variability and
Agriculture (Eds), Y. P. Abrol, S. Gadgil sand G. B. Pant, Narosa publishing
house, New Dehli, 69-97.
Small, D., Islam, S., and Richard, M. Vogel, 2006. Trends in precipitation and
stream flow in the eastern U.S. Paradox or perception? Geophysical Research
Letters, 33: l03403. DOI: 10.1029/2005gl024995.
Sneyers R., 1990. On The Statistical Analysis of Series of Observations. WMO
No: 415, Technical Note No: 143, Geneva.
Solow, A. R., and Broadus, J.M., 1989. On the detection of greenhouse warming.
Climate Change, 15: 449–453.
Solow, A.R., 1987. Testing for climate change: an application of the two-phase
regression model. Climate, Applied Meteorology, 26: 1401–1405.
Special Report on Emissions Scenarios (SRES) 2000. A Special Report of
Working Group III of the Intergovernmental Panel on Climate Control.
Cambridge, UK: Cambridge University Press. SL TD 885.5 G73 S7 2000.
124
Stern. N., 2006. The Stern Review on the Economics of Climate Change, British
report.HM,Treasury.
http://www.hmtreasury.gov.uk/independent_reviews/stern_review_economics_cli
mate_change/stern_review_report.cfm.
Stroeve, J., M., M. Holland, W. Meier, T. Scambos, and Serreze, M., 2007.
Arctic sea ice decline: Faster than forecast. Geophysical Research Letters, 34:
9501. DOI: 10.1029/2007GL029703.
Suppiah, R., Hennessy, K. J., 1998. Trends in the intensity and frequency of
heavy rainfall in tropical Australia and links with the southern oscillation.
Australian Meteorological Magazine 45: 1–17.
Takahashi, K., Y. Honda and S. Emori, 2007. Assessing Mortality Risk from
Heat Stress due to Global Warming. Risk Research, 10: 339-354.
TFCC, 2010. Climate change in Pakistan: TFCC (Task Force for Climate Change)
recommends various adaptations, mitigation measures.
Thayyen, R. J., Gergan, J. T., and Dobhal, D. P. 2007. Role of glaciers and snow
cover on headwater river hydrology in monsoon regime: Micro scale study of Din
Gad catchments (Dokriani Glacier), Garhwal Himalaya, India. Current Science
India, 92(3): 376–382.
The Nation Newspaper, 29th May, 2012. Islamabad.
Thorne, P.W., and Coauthors, 2003. Probable causes of late 20th century
tropospheric temperature trends. Climate Dynamics, 21: 573-591.
Tran, V.L., Hoang, D.C., and Tran, T.T., 2005. Building of climate change
scenario for Red River catchments for sustainable development and
environmental protection. Preprints, Science Workshop on Hydro meteorological
Change in Vietnam and Sustainable Development, Hanoi, Vietnam, Ministry of
Natural Resource and Environment, Hanoi, 70-82.
Trenberth KE., 1999. Conceptual framework for changes of extremes of the
hydrologica cycle with climate change. Climatic Change 42: 327–339.
125
Trenberth, K. E., Jones, P. D., Ambenje, P., Bojariu, R. Easterling, D., Klein
Tank, A., Parker, D., Rahimzadeh, F., Renwick, J. A., Rusticucci, M., Soden,
B., Zhai, P., 2007. Observations: Surface and Atmospheric Climate Change. In:
Climate Change 2007. The Physical Science Basis. Contribution of WG 1 to the
Fourth Assessment Report of the Intergovernmental Panel on Climate Change. S.
Solomon, D. Qin, M. Manning, Z. Chen, M. C. Marquis, K. B. Averyt, M. Tignor
and H. L. Miller (Eds), Cambridge University Press. Cambridge, U. K., and New
York, NY, USA, 235-336.
Treydte, K. S., Schleser, G. H., Helle, G., Frank, D. C., Winiger, M., Haug, G. H.,
and Esper, J., 2006. The twentieth century was the wettest period in northern
Pakistan over the past millennium. Nature, 440: 1179-1182. DOI:
10.1038/nature04743
Trigo, I. F., Davies, T. D., and Bigg, G. R., 2000. Decline in Mediterranean
rainfall caused by weakening of Mediterranean cyclones. Geophysical Research
Letters, 27: 2913–2916.
UN/FCCC, 1998. Kyoto Protocol to the United Nations Framework Convention
on Climate Change. UN/FCCC Document FCCC/CP/1997/7/Add.1, United
Nations, NewYork, Geneva.
UNEP, 1995. Global Biodiversity Assessment; Summary for Policy Makers: 46
UNEP, 2007. Global Environment Outlook – GEO4: Environment for
Development. United Nations Environment Programme, 540.
Vicente-Serrano, S. M. , Cuadrat- Prats, J. M., 2007. Trends in drought intensity
and variability in the middle Ebro valley (NE of the Iberian Peninsula) during the
second half of the twentieth century. Theoretical and Applied Climatology, 88:
247–258. DOI: 10.1007/s00704-006-0236-6.
Vörösmarty, C.J., Green, P., Salisbury, J., Lammers, R.B., 2000. Global water
resources: vulnerability from climate change and population growth. Science,
289: 284–288
WAPDA, 1997b. Golgen Gol hydel power project, Chitral: Feasibility study
report. Hydroelectric planning organization, Lahore, 2.
126
Wassmann, R., and Dobermann, A., 2007. Climate Change Adaptation through
Rice Production in Regions with High Poverty Levels. An Open Access Journal
published by ICRISAT, 4(1).
Watson, R. T., Zinyowera, M. C., Moss, R. H., and Dokken, D. J., 1997.
Summery for policy makers, the regional impacts of climate change: An
assessment of vulnerability. A special report of IPCC working group.
Webster, P.J., V.O. Magana, T.N. Palmer, J. Shukla, R.A. Tomas, M. Yanagi and
T.Yasunari, 1998. Monsoons: processes, predictability and the prospects for
prediction. Geophysical Research Letters, 103: 14451-14510.
Westerling, A. L. and B. P. Bryant, 2007. Climate Change and Wildfire in
California. Climatic Change.
WHO, 2000. Vegetation Fires. World Health Organization, Fact Sheet No. 254.
Aug 2000, 3. http://www.who.int/mediacentre/factsheets/fs254/en/index.html.
Winer, B. I, Brown, D. R., and Michels, K. M. 1991. Statistical principles in
experimental design (3rd Ed.). New York: McGraw-Hill.
WWF, 2005. An overview of glaciers, glacier retreat, and subsequent impacts in
Nepal, India and China. World Wildlife Fund, Nepal Programme.
Yadava, M. G., Ramesh, R. and Pant, G. B., 2004. The Holocene. 14(4): 517–
524.
Yan, Z. W., Jones, P.D., Davies, T. D., Moberg, A. Bergstrom, H., Camuffo, D.,
Cocheo, C., MAugeri, M., Demaree, G. R., Verhoeve, T., Thoen, E., Barriendos,
M., Rodriguez., R. Martin-Vide, J., and Yang, C., 2002. Trends of extreme
temperatures in Europe and china based on daily observation. Climate Change,
53: 355-392.
Yoshida, S., 1981. Fundamentals of rice crops science. Los Banos, Philippines,
IRRI.
Zhai, P. and X. Pan, 2003. Trends in temperature extremes during 1951-1999 in
China. Geophysical Research Letters, 30: 1913. DOI: 10.1029/2003GL018004.
Zhai, P.M., A. Sun, F. Ren, X. Liu, B. Gao and Q. Zhang, 1999. Changes of
climate extremes in China. Climatic Change, 42: 203-218.
127
Zhang, X., Zwiers, F.W., Hegerl, G.C., 2007. Detection of human influence on
twentieth-century precipitation trends. Nature, 448: 461-465.
Zhou, L., C.J. Tucker, R.K. Kaufmann, D. Slayback, Shabanov and R.B.
Mynemi., 2004. Variations in northern vegetation activity inferred from satellite
data of vegetation index during 1981-1990. Geophysics.
128
ANNEXURE
129
Table 3.1. Basic geographical information of climate zones of the study area.
S. No Station WMO Ref Altitude (m) Latitude Longitude Ensuing Rivers
1 Astor 41520 2168 35 o20'N 74 o 54'E Astor River

o o
2 Chilas 41519 1250 35 25'N 74 06'E Indus River
3 Chitral 41506 1499 35o51'N 71 o 50'E
4 Dir 41508 1375 35 o 12'N 71 o 51'E
5 Gilgit 43516 1459 35 o 55'N 74 o 20'E
6 Kakul 41535 1308 34 o 11'N 73 o 15'E
7 M-abad 43532 701 34 o 22'N 73 o 29'E
8 P-Chinar 41560 1725 33 o 52'N 70 o 05'E Kurram River
o o
9 S- Sharif 41523 961 34 44'N 72 21'E Swat River
o o
10 Skardu 43517 2209 35 18'N 75 41'E Shingar River
11 Cherat 41565 1372 33 o 49'N 71 o 33'E
12 DIK 41624 173 31 o 49'N 70 o 55'E Gomal River
o o
13 Isl-abad 41571 507 33 37'N 73 06'E
o
14 Lahore 41640 213 31 33'N 74 o 20'E Ravi River
o o
15 Peshawar 41530 359 34 01‘N 71 35'E Kabul River
o o
16 Sialkot 41600 255 32 31'N 74 32'E Chenab River
o o
17 Kalat 41696 2015 29 02' N 66 35' E
o
18 Khuzdar 41744 1231 27 50'N 66 o 38'E
19 Quetta 41660 1719 30o 15' N 66o 53' E
20 Zhob 41620 1405 31o 21'N 69o 28' E Gomal River
o o
21 B-pure 41700 110 29 20'N 71 47'E
o
22 J-abad 41715 55 28 18'N 68 o 28'E
23 Khanpur 41718 87 28 o 39'N 70 o 41'E
24 Multan 41675 122 30 o 12'N 71 o 26'E
25 Rohri 41725 66 27 o 40'N 68 o 54'E
o
26 Sibbi 41697 133 29 33'N 67 o 53'E
27 H-abad 41764 28 25 o 23'N 68 o 25'E
28 Jewani 41756 56 25 o 04'N 61 o 48'E
29 Karachi 41780 21 24 o 54'N 67 o 08'E
30 N-shah 41749 37 26 o 15'N 68 o 22'E
WMO Ref: Stands for World Meteorological Observety Reference Number
130
Table. 4.1 Duncan Test for homogenous subset of Alpha for all climate zones.
Minimum temp Subset for alpha = 0.05

Zone N Means of all climate zones
Zone 1 293 9.26
Zone2 120 10.27
Zone3a 180 15.83
Zone3b 180 18.92
Zone4 120 20.35
Sig. 893 1.00 1.00 1.00 1.00 1.00
Max temp Subset for alpha = 0.05

Zone N Means of all climate zones
Zone 1 300 22.89
Zone2 120 25.51
Zone3a 180 29.21
Zone3b 120 32.86
Zone4 180 33.70
Sig. 1.00 1.00 1.00 1.00 1.00
Table. 4.2 Duncan test for homogenous subset of Alpha for the whole Study Area.
Whole Study Area Subset for alpha = 0.05
st nd
HMSS 1 Decade 2 Decade 3rd Decade Sig (1) Sig (2)
Min Temp 10 14.03 14.12 14.33 0.12
Max Temp 10 27.75 27.89 28.35 0.51 1
Mean Temp 10 20.87 21.005 21.34 0.48 0.05
131
Table. 4.3 Stations are showing significant increase, decrease, and no change in
mean minimum temperature.
Increase in minimum temperature
Station F Sig. Mean (1976-1990) Mean (19912005) Mean Diff Zone
Isl-bad 7.253 0.01 14.18 14.73 0.55 2
Lahore 24.147 0.00 17.97 19.01 1.04 2
Peshawar 6.762 0.02 15.70 16.36 0.67 2
Khanpur 6.271 0.02 16.92 17.55 0.63 3b
Multan 5.556 0.03 18.02 18.42 0.40 3b
Karachi 13.599 0.00 19.68 21.77 2.09 4
Decrease in minimum temperature
P-Chinar 43.188 0.00 9.55 5.46 -4.09 1
Jewani 3.835 0.05 21.15 20.69 -0.46 4
No change in minimum temperature
Astor 0.238 0.63 3.63 3.39 -0.24 1
Chilas 0.081 0.78 14.52 14.57 0.05 1
Chitral 0.135 0.72 8.76 8.67 -0.09 1
Dir 0.35 0.56 7.88 8.05 0.17 1
Gilgit 0.476 0.50 7.66 7.78 0.12 1
Kakul 0.132 0.72 10.56 10.45 -0.11 1
M-abad 0.73 0.40 13.46 13.62 0.16 1
Skardu 0.205 0.65 5.21 5.08 -0.13 1
S-Sharif 0.864 0.36 12.28 12.02 -0.27 1
Cherat 0.282 0.60 12.69 12.81 0.13 2
DIK 1.01 0.32 16.71 17.03 0.32 2
Sialkot 1.642 0.21 16.23 16.49 0.27 2
Khuzdar 0.016 0.90 14.87 14.80 -0.07 3a
Quetta 0.309 0.58 8.34 8.48 0.14 3a
Zhob 1.097 0.30 12.18 12.70 0.51 3a
B-Pur 1.736 0.20 18.05 18.40 0.35 3b
Rohri 0.42 0.52 19.70 19.85 0.14 3b
H-bad 2.446 0.13 21.32 21.75 0.43 4
Kalat 3.041 0.09 5.03 5.80 0.76 3a
Jaco-bad 3.968 0.06 19.94 20.27 0.32 3b
Sibbi 3.174 0.09 19.80 20.16 0.36 3b
N-Shah 3.011 0.09 18.03 18.38 0.35 4
132
Table. 4.4 Stations are showing significant increase, decrease, and no change in mean maximum
temperature.
Increase in Maximum temperature

Station F Sig. Mean (1976-1990) Mean (19912005) Mean Diff Zone
Quetta 4.346 0.05 24.55 25.22 0.67 3a
Karachi 13.84 0.00 31.53 32.35 0.82 4
N-Shah 4.688 0.04 35.26 35.83 0.57 4
No change in Maximum temperature
Astor 0.889 0.35 15.46 15.74 0.28 1
Chilas 2.854 0.09 26.58 26.11 -0.47 1
Chitral 0.225 0.64 23.46 23.61 0.15 1
Dir 1.478 0.23 22.70 23.04 0.34 1
Gilgit 1.291 0.27 24.21 24.51 0.30 1
Kakul 1.846 0.19 22.45 22.76 0.31 1
M-abad 2.205 0.15 27.11 27.62 0.51 1
Skardu 2.155 0.15 18.65 19.16 0.51 1
S-Sharif 0.843 0.37 25.79 26.08 0.29 1
Cherat 0.052 0.82 20.71 20.78 0.07 2
DIK 0.049 0.83 31.59 31.65 0.06 2
Isl-bad 0.052 0.82 28.71 28.78 0.07 2
Lahore 0.844 0.37 34.76 35.02 0.25 2
Peshawar 0.069 0.80 29.76 29.83 0.07 2
Sialkot 0.635 0.43 29.36 29.56 0.20 2
B-Pur 2.074 0.16 32.70 33.07 0.38 3b
Jaco-bad 0.529 0.47 34.10 34.33 0.23 3b
Khanpur 0.229 0.64 33.39 33.52 0.13 3b
Multan 0.08 0.78 32.61 32.68 0.07 3b
Rohri 2.046 0.16 34.02 34.34 0.32 3b
Sibbi 0.467 0.50 34.73 34.92 0.19 3b
Jewani 1.746 0.20 29.92 30.28 0.36 4
P-Chinar 2.883 0.09 21.16 21.62 0.45 1
Kalat 3.501 0.07 21.92 22.46 0.55 3a
Khuzdar 3.486 0.07 28.66 29.32 0.66 3a
Zhob 2.982 0.09 25.69 26.28 0.60 3a
H-bad 3.317 0.08 33.60 34.08 0.49 4
133
Table. 4.5 Whole Study Area analysis of minimum and maximum temperatures
Minimum temp (Whole Study Area F= 0.408, p= 0.523)

Zones F Sig. Mean (1976-1990) Mean (1991-2005) Mean Diff
Zone1 0.67 0.00 9.43 9.10 -0.33
Zone2 2.91 0.00 15.58 16.07 0.49
Zone3a 0.227 0.635 10.11 10.44 0.34
Zone3b 3.998 0.047 18.74 19.11 0.37
Zone4 3.976 0.048 20.05 20.65 0.60
Maximum temp (Whole Study Area F= 0.762, p= 0.383)
Zone1 1.543 0.225 22.76 23.03 0.27
Zone2 0.441 0.512 29.15 29.27 0.12
Zone3a 7.225 0.012 25.20 25.82 0.62
Zone3b 1.323 0.26 33.59 33.81 0.22
Zone4 7.471 0.01 32.58 33.14 0.56
Table. 4.6 Dunnett T3 for homogenous subset of alpha.
Zone N Mean (I) Decade (J) Decade Mean Diff (I-J) Sig.
Zone 1 99 62.39 1 2 -11.65 0.25
3 -0.98 1.00
100 74.04 2 1 11.65 0.25
3 10.67 0.30
98 63.37 3 1 0.98 1.00
2 -10.67 0.30
Zone 2 59 72.11 1 2 8.66 0.52
3 6.71 0.73
60 63.44 2 1 -8.66 0.52
3 -1.96 0.98
56 65.40 3 1 -6.71 0.73
2 1.96 0.98
Zone 3a 36 35.12 1 2 -0.97 0.99
3 7.34 0.14
40 36.09 2 1 0.97 0.99
3 8.31 0.01
40 27.78 3 1 -7.34 0.14
2 -8.31 0.01
Zone 3b 59 23.15 1 2 -1.51 0.92
3 2.91 0.59
55 24.65 2 1 1.51 0.92
3 4.42 0.23
58 20.23 3 1 -2.91 0.59
2 -4.42 0.23
Zone 4 40 34.76 1 2 -5.48 0.70
3 13.04 0.02
38 40.24 2 1 5.48 0.70
3 18.52 0.01
40 21.71 3 1 -13.04 0.02
2 -18.52 0.01
134
Table 4.7 Dunnett T3 test for multiple comparison
Mean Rainfall
Decades (I) Decade (J) Decade Mean Difference (I-J) Sig.
1st Decade 1 2 -3.73 0.50
3 4.67 0.44
2nd Decade 2 1 3.73 0.50
3 8.40 0.04
3rd Decade 3 1 -4.67 0.44
2 -8.40 0.04
Table. 4.8 Single station data analysis for the determination of significant trend in two time periods
Increase in Rainfall
Station F Sig. Mean (1976-1990) Mean (1991-2005) Mean Diff Zone
Chitral 8.09 0.01 34.67 45.90 11.23 1
Decrease in Rainfall
Quetta 4.43 0.04 47.36 33.59 -13.77 3a
Karachi 3.87 0.05 42.82 28.10 -14.72 4
No change in Rainfall
Astor 0.04 0.84 43.31 44.32 1.01 1
Chilas 2.87 0.10 16.03 22.63 6.60 1
Dir 0.43 0.52 126.91 121.05 -5.86 1
Gilgit 0.56 0.46 15.85 13.79 -2.06 1
Kakul 1.68 0.21 121.19 111.83 -9.36 1
M-abad 0.25 0.62 134.76 130.28 -4.48 1
P-Chinar 0.35 0.56 66.16 68.90 2.74 1
Skardu 0.62 0.44 19.64 22.03 2.38 1
S-Sharif 0.02 0.91 88.42 89.51 1.10 1
Cherat 0.03 0.85 54.86 53.80 -1.06 2
DIK 0.27 0.61 28.73 30.43 1.70 2
Isl-bad 0.91 0.35 115.58 105.41 -10.18 2
Lahore 2.02 0.17 77.94 60.76 -17.18 2
Peshawar 1.73 0.20 39.68 46.00 6.33 2
Sialkot 0.03 0.87 91.78 93.56 1.79 2
Kalat 0.00 1.00 28.48 28.46 -0.02 3a
Khuzdar 0.09 0.77 31.83 30.42 -1.40 3a
Zhob 0.01 0.92 31.19 30.85 -0.34 3a
B-Pur 0.48 0.49 25.12 21.26 -3.86 3b
Jaco--bad 3.69 0.07 27.48 15.28 -12.20 3b
Khanpur 0.08 0.78 16.54 15.43 -1.11 3b
Multan 0.09 0.77 25.03 26.25 1.22 3b
Rohri 0.04 0.85 25.18 26.36 1.18 3b
Sibbi 0.59 0.45 21.67 24.67 3.00 3b
H-abad 0.77 0.39 41.39 30.63 -10.77 4
Jewani 1.57 0.22 29.53 22.33 -7.20 4
N-Shah 0.10 0.76 33.06 29.74 -3.32 4
135
Table. 4.9 Whole Study Area data analysis of Mean Rainfall.
Rainfall (Whole Study Area F=1.273, p= 0.26)

Zones F Sig. Mean (1976-1990) Mean (1991-2005) Mean Diff
Zone1 2.483 0.12 66.85 66.42 -0.43
Zone2 0.21 0.65 68.24 65.69 -2.55
Zone3a 2.439 0.1 35.16 30.83 -4.33
Zone3b 0.735 0.39 23.37 21.91 -1.46
Zone4 3.719 0.05 36.66 27.7 -8.96
Table 4.10 Duncan test for homogenous subset of alpha
Duncan Test Subset for alpha = 0.05

River HMSS 1st Decade 2nd Decade 3rd Decade Sig (1) Sig (2)
Astor 7.5 118.15 158.59 134.79 0.19 0.07
Chitral 7.5 264.75 284.2 285.04 0.18
Gilgit 7.5 616.85 623.43 583.95 0.26
Indus/Chilas 7.5 1759.59 1740.08 1931.6 0.19
Indus/Skardu 7.5 968.91 1141.24 1235.66 1 0.27
Swat 7.5 79.96 94.69 86.9 0.19 0.14
136
Table 4.11 Model Fit descriptive statistics for forecasting trend
Model Fit descriptive statistics

Percentile
Fit Statistic Mean SE Min Max 5 10 25 50 75 90 95
Stationary 0.296 0.33 9.81E-15 0.769 9.81E-15 9.81E-15 0.049 0.208 0.632 0.769 0.769
R-squared
R-squared 0.129 0.088 9.81E-15 0.194 9.81E-15 9.81E-15 0.038 0.162 0.189 0.194 0.194
RMSE 2.2 3.58 0.374 7.569 0.374 0.374 0.374 0.429 5.798 7.569 7.569
MAPE 4.581 5.77 1.475 13.226 1.475 1.475 1.494 1.811 10.437 13.226 13.226
MaxAPE 15.385 22.551 2.897 49.129 2.897 2.897 2.962 4.756 38.436 49.129 49.129
MAE 1.794 2.9 0.292 6.144 0.292 0.292 0.301 0.37 4.711 6.144 6.144
MaxAE 4.638 7.668 0.68 16.139 0.68 0.68 0.708 0.865 12.339 16.139 16.139
Normalized -0.194 2.919 -1.856 4.162 -1.856 -1.856 -1.855 -1.54 2.814 4.162 4.162
BIC
137

Climate Change and Variability Trends in Pakistan and Its Environmental Effects

Uploaded by

Copyright:

Available Formats

Climate Change and Variability Trends in Pakistan and Its Environmental Effects

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Climate Change and Variability Trends in Pakistan and Its Environmental Effects

Uploaded by

Copyright:

Available Formats

CLIMATE CHANGE AND VARIABILITY

TRENDS IN PAKISTAN AND ITS

Thesis submitted to Department of Environmental

Department of Environmental Sciences

List of Contents ------------------- i

Chapter#2 Literature Review 18

2.1. Studies related to the determination of Temperature trend 18

2.2 Studies related to the determination of Rainfall trend 21

2.3 Collective studies related to the determination of 22

3.2 Climate Zones of the study area 27

3.3. Limitation of the data 29

3.4 Data Analysis 30

3.4.3. 5% Trimmed Mean 30

3.4.5. Standard Deviation 31

3.4.6. Standard Error 31

3.4.7. 95% confidence interval (CI) for the mean 31

3.4.9. ANOVA test 32

3.5. Trend Visualization 34

3.5.2. Graphical Presentations 35

3.6. Trend Forecasting 35

Chapter # 4 Results and Discussion 37

4.1.2. Data collection and analysis 38

4.1.3. Results of temperature data 39

4.1.3.1. Descriptive statistics and ANOVA result of Mean Minimum 39

4.1.3.3. Single station analysis 49

4.1.3.5. Entire study period analysis 57

4.2.2. Data collection and analysis 63

4.2.3. Results of rainfall data 64

4.2.3.1. Descriptive statistics and ANOVA results 64

4.2.3.2. Zonal Analysis 66

4.2.3.3. Single station analysis 73

4.2.3.4. Whole study area analysis 75

4.2.3.5. Entire study period analysis 77

4.3. River flows 79

4.3.2. Data Collection and Analysis 80

4.3.3. Results of ensuing rivers 80

4.3.3.1. Descriptive statistics and ANOVA results 80

4.3.3.2. Single Gauge station analysis 81

4.4. Hydro-Meteorological characteristics 82

4.4.1 River basin analysis 83

4.4.1.2. Chitral River 85

4.4.1.3. Gilgit River 87

4.4.1.5. Indus River at Skardu Station 91

4.4.1.6. Swat River 93

4.5. Trend Forecasting 97

4.1. Descriptive statistics and ANOVA results of minimum and maximum 40

4.4 Duncan test for homogenous subset of Alpha for zone1 44

4.5. Descriptive statistics and ANOVA result for zone2 45

4.6. Duncan test for homogenous subset of Alpha for zone 2 45

4.7. Descriptive statistics and ANOVA result for zone 3a 46

4.8 Duncan test for homogenous subset of Alpha for zone 3a 46

4.9. Descriptive statistics and ANOVA result for zone 3b 47

4.11 Descriptive statistics and ANOVA result for zone 4. 49

4.12 Duncan test for homogenous subset of Alpha for zone4. 49

4.13 Descriptive statistics and ANOVA result for climate zones. 65

4.14 Descriptive statistics and ANOVA result for the whole 66

4.15 Descriptive statistics and ANOVA result for zone 1 67

4.16 Dunnett T3 test for multiple comparison 67

4.17 Descriptive statistics and ANOVA result for zone2 68