GENERAL ARTICLES

Economic impact assessment of the

Agrometeorological Advisory Service of India
Parvinder Maini* and L. S. Rathore

A pilot study was conducted to assess the economic impact of weather forecast-based advisories
issued to 15 of the 127 Agrometeorological Advisory Service (AAS) units of the Ministry of Earth
Sciences, Government of India. Six seasons comprising three Kharif (summer) and three Rabi (win-
ter) during 2003–2007 were chosen. The major crops chosen for the study included food grains,
oilseeds, cash crops, fruit and vegetable crops. The sample set consisted of 80 farmers, comprising
40 responding and 40 non-responding farmers. The main aim was to study the percentage in-
crease/decrease in the yield and net return due to AAS. Results obtained suggest that the AAS farm-
ers accrued a net benefit of 10–15% in the overall yield and a reduction by 2–5% in the cost of
cultivation over the non-AAS farmers.

Keywords: Advisories, Agrometeorology Advisory Service, economic impact, yield.

AGRICULTURAL productivity largely depends upon From a farmer’s perspective, the forecast value increases
weather. Weather forecasts in all temporal ranges are de- if the weather and climate forecasts are capable of influ-
sirable for effective planning and management of agricul- encing their decisions on key farm management opera-
tural practices. The development of response strategy1 tions13–15. Thus, it becomes essential to relate with the
helped farmers realize the potential benefits of using requirements of farmers16, understand their needs and
weather-based agrometeorological information in mini- give the forecast in appropriate spatial and temporal
mizing the losses due to adverse weather conditions, range17–20. This ultimately helps in increasing the reliabil-
thereby improving yield, quantity and quality of agricul- ity of the forecast and thus in better adoption of the
tural productions. In fact, short and medium-range weather-based advisory21.
weather forecasts play a significant role in making short- The National Centre for Medium Range Weather Fore-
term adjustments in daily agricultural operations. casting (NCMRWF) under the Ministry of Earth Sciences
Some of the early works that appeared in the late 1960s (MoES), Government of India in collaboration with India
concentrated on effectiveness of agrometeorological Meteorological Department (IMD), Indian Council of
information2–4. Studies have also been carried out to Agricultural Research and State Agricultural Universities
determine the potential benefits in agricultural farm deci- had been providing Agrometeorological Advisory Ser-
sions from long-range weather predictions5–9, in particu- vices (AAS) at the scale of agroclimatic zone to the farm-
lar in areas where the El Niño/Southern Oscillation has ing community based on location-specific medium-range
marked impact on the regional climate10,11. However, weather forecast (MRWF)22. Since 2007, the entire
very little work has been done on the economic impact of framework of AAS, developed and successfully demon-
medium-range weather forecasts on farm-level decisions. strated by NCMRWF, has been relocated at IMD under
In general, it is difficult to assess the economic benefit of MoES for extending the service (in operational mode) to
any advisory service given to take measures against ca- districts under these agro-climatic zones. It is now called
tastrophes or life-threatening situations, but it is possible the Integrated Agrometeorological Advisory Service
to assess the economic benefit of the agrometeorological of MoES. Thus, the AAS set up exhibits a multi-
services12. This can be done if the scientific methods to institutional, multidisciplinary synergy to render an
be used for weather-based advisories have a direct rela- operational service for use of the farming community.
tionship with the traditional knowledge of the farmers11. These weather-based agro-advisories have been help-
ing the farming community to take advantage of prognos-
ticated weather conditions and thereby form a response
Parvinder Maini is in the Ministry of Earth Sciences, Prithvi Bhavan,
Lodhi Road, New Delhi 110 003, India; L. S. Rathore is in the India
strategy. This was also reported by the AAS units. But
Meteorological Department, Mausam Bhavan, Lodhi Road, New Delhi these were sporadic cases and could not be intercompared
110 003, India. mainly due to non-uniform use of the methodology.
*For correspondence. (e-mail: parvinder.maini@moes.gov.in) Hence, a detailed study was carried out at 15 AAS units

1296 CURRENT SCIENCE, VOL. 101, NO. 10, 25 NOVEMBER 2011

 GENERAL ARTICLES

with an objective to evaluate the use and value of the ser- wherein verification was done by the service provider
vice in economic terms. This also helped in assessing the (NCMRWF) as well as the user community (farmers).
usage pattern of AAS and identifying the strengths and Here the verification is presented during two main crop-
weaknesses of the service. The impact studies have indi- ping seasons, namely Kharif and Rabi for all the 15 units
cated that the weather based advisories issued by during the period of study. The evaluation of usability of
NCMRWF had a positive impact on the overall yield and forecast of quantitative precipitation and temperature
also helped in decreasing the cost of cultivation. One of (maximum (Tx); minimum (Tn)) was done using an error
the major achievements of the study was that it helped in structure formulated at NCMRWF25,26 and is given in
increasing awareness among farmers about the adoption Table 1. Table 2 give the skill in terms of ratio score (RS)
of weather based advisories and their positive impacts. and Hanssen and Kuipers Score (HKS) for rainfall fore-
The present article discusses the pilot study, its methodo- cast and correlation coefficient (CC) and root mean
logy and impact of AAS. square error (RMSE) for temperature forecast during the
period of study. Although the skill of forecast follows a
similar trend every year32–34, a distinct difference is seen
Preparation, dissemination and feedback
in the Kharif and Rabi seasons.
mechanism of the agro-advisory bulletin
It is observed from Table 2 that the RS of Yes/No rain-
fall forecast is around 90% during Rabi and 69% in
For issuing the weather based agro-advisories, a forecast-
Kharif. Similarly on an average the HKS is around 0.4
ing system for generating23 objective medium-range loca-
during Kharif and 0.5 during Rabi. Tx has a correlation of
tion-specific forecast of surface weather elements has
65–70% and an RMSE of 2–3°C in Kharif, whereas in
been evolved at NCMRWF24–26. Location-specific weather
Rabi the correlation is around 60%. On the other hand,
forecasts for six parameters, viz. rainfall, cloud cover,
the correlation of Tn forecast is around 50% in Kharif and
wind direction and speed, and minimum and maximum
about 65% in Rabi season. The RMSE of Tn is lower than
temperature were obtained twice a week from a T-80
that of Tx and is in the range 1–2.5°C during both sea-
General Circulation Model with a resolution of 150 km ×
sons.
150 km. These forecasts were further subjected to statis-
The usability of the temperature and rainfall forecast is
tical27,28 and synoptic interpretation by experts and 5-days
given in Table 3. While in the case of quantitative pre-
forecasts in quantitative terms were issued to the AAS
cipitation, the Rabi forecast (90–98%) is better than the
units twice a week. These forecasts were converted into
Kharif forecast (60–80%), in temperature forecast it is
farm-level advisories by AAS units and disseminated to
observed that the usability is good during both seasons,
the farmers in vernacular language through mass media22.
with maximum temperature having higher usability
Public awareness programmes in ‘kisan melas’ are regu-
in Rabi (50–90%) and minimum temperature in Kharif
larly held by agricultural universities to educate farmers
(60–95%).
on the usage of farm advisories. Figure 1 shows the com-
AAS units also reported periodic feedback on worthi-
plete flow diagram of the AAS of NCMRWF.
ness of forecasts and usefulness of advisories in eco-
In order to regularly assess the quality and skill of fore-
nomic terms. Although these studies gave an initial idea
cast, a verification mechanism was introduced24–31,
about the value of forecast, tangible conclusion could
be arrived at only by carrying out the study in greater
detail.

Table 1. Criteria for obtaining usability of quantitative precipitation

and temperature forecast

Observed rainfall Observed rainfall

≤10 mm >10 mm

Correct Diff ≤ 0.2 mm Diff ≤ 2% of obs

Usable 0.2 mm < Diff ≤ 2.0 mm 2% of obs < Diff ≤ 20% of obs
Unusable Diff >2.0 mm Diff > 20% of obs

Diff, Absolute difference of rainfall observed and forecast (in mm).

Obs, Observed rainfall (in mm).

Correct Diff ≤ 1°C

Usable 1°C < Diff ≤ 2°C
Unusable Diff > 2°C
Figure 1. Agrometeorological Advisory Service (AAS) of the Natio-
nal Centre for Medium Range Weather Forecasting (NCMRWF). Diff, Absolute difference of temperature observed and forecast (in °C).

CURRENT SCIENCE, VOL. 101, NO. 10, 25 NOVEMBER 2011 1297

GENERAL ARTICLES
Table 2. Skill of forecast during the study period

Kharif Rabi

Rain Tn Tx Rain Tn Tx

Station RS HKS RMSE CC RMSE CC RS HKS RMSE CC RMSE CC

Anand 74 0.45 1.59 0.68 1.97 0.87 92 0.37 2.35 0.69 1.54 0.8
Bangalore 57 0.19 1.29 0.17 1.68 0.7 84 0.25 1.94 0.64 1.64 0.37
Bhubaneshwar 65 0.3 1.65 0.54 2.7 0.74 98 0.41 2.35 0.32 1.93 0.41
Hisar 75 0.38 2.61 0.55 2.7 0.6 96 0.3 2.12 0.68 2.85 0.7
Coimbatore 60 0.13 1.67 0.29 2.33 0.33 87 0.42 1.89 0.58 1.89 0.61
Hyderabad 56 0.24 1.53 0.64 2.54 0.81 95 0.39 1.98 0.74 1.5 0.66
Jaipur 62 0.25 2.49 0.51 3.13 0.6 91 0.22 2.99 0.67 2.07 0.76
Jodhpur 80 0.48 2.78 0.45 2.97 0.65 96 0.32 2.34 0.69 1.68 0.74
Ludhiana 70 0.31 2.61 0.6 3.69 0.53 87 0.38 2.86 0.62 2.41 0.68
Nadia 78 0.33 1.6 0.28 2.24 0.35 93 0.74 2.4 0.61 2.64 0.51
Pantnagar 72 0.56 1.85 0.39 2.63 0.77 92 0.46 2.14 0.77 3.09 0.31
Pune 67 0.2 1.21 0.54 1.88 0.54 98 0.5 2.67 0.53 2.45 0.56
Raipur 67 0.33 1.79 0.58 2.36 0.81 94 0.41 2.43 0.52 2.5 0.6
Solan 70 0.42 1.87 0.56 2.23 0.82 87 0.56 2.13 0.77 2.86 0.69
Thrissur 82 0.5 1.22 0.2 1.66 0.6 89 0.22 1.31 0.62 1.76 0.22

RS, Ratio score; HKS, Hanssen and Kuipers score; RMSE, Root mean square error and CC, Correlation coefficient.

Table 3. Usability (in %) of rainfall and temperature forecast during the study period

Kharif Rabi

Station Rain Tn Tx Rain Tn Tx

Anand 66.78 82.84 70.89 99.23 68.48 89.34

Bangalore 68.02 97.61 77.28 98.66 74.67 88.57
Bhubaneshwar 54.34 89.07 58.58 98.65 56.58 71.25
Hisar 88.89 70.73 65.61 95.77 56.33 58.62
Coimbatore 85.88 82.68 75.21 95.95 72.73 85.19
Hyderabad 55.87 89.71 70.33 96.11 90.00 87.30
Jaipur 82.67 71.31 57.37 100.00 62.07 67.81
Jodhpur 75.89 75.23 60.33 100.00 65.49 71.33
Ludhiana 84.57 61.51 59.84 90.00 59.77 64.37
Nadia 65.43 84.03 63.73 100.00 53.33 65.34
Pantnagar 57.17 56.94 40.27 94.59 56.25 43.75
Pune 61.35 90.59 75.13 100.00 66.24 87.50
Raipur 67.44 77.20 65.99 96.34 65.52 79.31
Solan 60.25 78.78 64.94 96.92 65.38 60.26
Thrissur 66.44 95.84 87.50 100.00 90.25 82.98

Figure 2. Schematic diagram showing the steps involved in the assessment of the impact of AAS.

1298 CURRENT SCIENCE, VOL. 101, NO. 10, 25 NOVEMBER 2011

 GENERAL ARTICLES

Impact assessment of AAS for the project, helped formulate the study plan, including
devising the sampling method, preparation of the ques-
Agromet impact study paradigm tionnaire and monitoring its implementation. NCMRWF
on its part was responsible for conceptualizing and
There does not exist any general simulation model for the executing the study, providing grants and bringing out the
evaluation of the economic benefits of meteorological final report.
assistance to agriculture. However, the benefit can always
be assessed in terms of what is apparent or possible
and what can be the maximum possible benefit theoreti- Benefits or expectations from these studies
cally.
The schematic diagram in Figure 2 depicts the various It was envisaged that the project
stages of the study. As a first step, quantitative weather
forecast generated at NCMRWF was disseminated to • Will give an insight into forecasting skill and reach of
Agro Meteorological Field Units (AMFU) in quantitative the service and also its economic value in terms of
terms along with information of adverse weather, if any. money.
Weather information was then translated into farm level • Will help in conducting farm operation in tune with
action-oriented advice by the agricultural scientists at weather forecasts leading to energy-saving, enhancing
AMFU. It contained weather based advisories, including the efficacy of inputs such as fertilizer, pesticide, etc.,
time and method of sowing, time and amount of irriga- cutting the cost of cultivation and saving crop from
tion, time and method of fertilizer/pesticide application, adverse weather.
etc. Agriculture impacts realized included changes • Will help in assessing the impact of favourable weather
experienced by farmers in terms of a positive (a benefit on overall growth, development and final yield of the
effect) or negative (an undesired effect) effect, thereby crop.
helping them to decide on the selection of crop/variety, • Will give agromet advisories that will increase profits
sowing/harvesting time, irrigation management, fertilizer by consistently delivering actionable weather informa-
management, pest/disease management and other inter- tion, analysis and decision support for farming situa-
cultural operations. This formed the backbone of the tions such as pest management through forecast of
economic impact study carried out by NCMRWF in col- relative humidity, temperature and wind; water man-
laboration with AAS units. agement through rainfall forecasts and crop protection
from thermal stress through forecast of extreme
Preliminary outline temperature conditions.

Once the outreach of AAS and skill of the forecast were

established, it was pertinent to study the impact of the
service in terms of economic gain/loss. Although some of
AAS units reported the benefits accrued by the farmers,
they were generally qualitative. This helped in creating a
modest awareness about the impact of the weather-based
agro-advisories on the farming sector, but there was lack
of a clear and precise understanding of the impact. There-
fore, there was a need to carry out an impact assessment
study using specified impact criteria and with a uniform
pattern of study.
Assessing impacts of weather forecast application in
farm management sector is a stupendous task. It becomes
even more challenging if one is attempting to quantify the
value of weather forecast-based agro-advisories. It was
difficult to consider all crop and all agro-climatic situa-
tions. Hence a conscious decision was taken to undertake
the study at 15 representation sites (Figure 3) covering
principal crops (cereal, millets, fruits and vegetables, oil-
seeds and cash crops). The study period was spread over
3 years, comprising three Kharif (2004, 2005 and 2006)
and three Rabi seasons (2004–05, 2005–06 and 2006–07).
The National Centre for Agriculture Economics and Pol-
icy Research (NCAP), which was engaged as a consultant Figure 3. Network of stations chosen for the study.

CURRENT SCIENCE, VOL. 101, NO. 10, 25 NOVEMBER 2011 1299

GENERAL ARTICLES

Hence it was expected that through intelligent use of and computational skills. Since the main objective of the
weather-based farm advisories, the study would facilitate study was to assess the adequacy, use and impact of the
protection of natural resources and preservation of the medium-range weather forecasts, an analytical method
environment. focusing more on farm-level impact was considered to be
the most appropriate. Therefore, NCAP proposed the use
of simple farm-level indicators for the impact assessment.
Objectives of the study
The impact assessment framework included estimation of
accuracy of the forecast, adequacy and reliability of the
• To study the adoption of the forecast by the user com-
forecast from the farmers’ perspective, use of the forecast
munity and its realization, thus understanding the
and farm-level impacts.
linkages between information, users and impacts.
Table 4 describes the framework followed for assess-
• To assess the use, effectiveness and potential benefits
ing the usefulness of weather forecast through the survey
of AAS by taking into account the AAS contact and
and Table 5 gives the economic impact indicators consid-
non-contact farmers.
ered.
• To understand and formulate weather-based farming
strategies based on the economic impact of AAS.
• To recognize the needs of the farming community for Sample selection
increasing the farm produce.
• To assess the economic impact of AAS in various
Considering the importance of sampling in the study, care
crops under different agro-climatic conditions.
was taken to identify the sample which is the true repre-
sentative of the class. Thus farmers were selected based
It also included other components of AAS such as dis-
on their size of holding (small, medium, large), educa-
semination of the bulletins, outreach of the service, and
tional background, size of the family and type of crop
capacity of the user community in adopting the advisories
grown. The next section gives the demographic details of
by different sections of the society under varying educa-
the samples chosen by each unit.
tion, gender and socio-economic classes. The impact
Out of a total of 127 AAS units, 15 were selected based
assessment framework also encompassed the aspects
on the existence of an effective weather-based AAS at the
related to skill of weather forecast35, its reliability and
unit for a few years. From each unit, a representative dis-
adequacy, mechanism of flow of weather information,
trict where AAS unit was operating was selected for con-
and extent of use of weather-based advisories by farmers
ducting the farm survey. The selection of the district was
for economic gains. It however did not cover the evalua-
based on its similarity with the agro-climatic zone in
tion of the capacity and methods of weather forecasts,
terms of cropping pattern, irrigated area, and rainfall and
which were beyond the scope of the present study.
soil type. A list of villages, from the selected district,
having AAS contact farmers was prepared and two
Impact assessment analysis framework villages were chosen randomly from among these. From
each selected village, a list of all AAS contact farmers
The selection of the analytical method is determined by was prepared by category of their size of holding (small,
the objective of the study, availability of required data medium, large), educational background, family size,

Table 4. Use of weather forecasts

Impact area Indicator

Perception of stakeholders Reliability, dissemination, adequacy, value-addition

Awareness about AAS Farmers knowing about AAS (%)
Usefulness – perception of farmers Farmers considering it useful (%)
Use of information Farmers using weather forecasts (%)

Table 5. Economic impact indicators

Parameter Indicator

Yield Difference in yield of AAS and non-AAS farmers.

Cost • Difference between total paid out cost (per acre) of AAS and non-AAS contact farmers.
• Changes in cost per unit of output.
Profitability Difference in return over paid out cost (Rs/acre) of AAS and non-AAS contact farmers.
Utility Increase in utilization by farmer for manoeuvring cultural operations.

1300 CURRENT SCIENCE, VOL. 101, NO. 10, 25 NOVEMBER 2011

 GENERAL ARTICLES

Figure 4. Schematic diagram showing the sampling framework.

type of crop grown, etc. A total of 20 farmers were then following in the questionnaire: (a) family structure;
selected using random sampling technique. Thus a sample (b) literacy among farmers; (c) size of land holding;
size of 40 AAS contact farmers was selected from two (d) cropping pattern; (e) traditional methods used; (f)
villages. Similarly, the non-AAS farmers were chosen. awareness of AAS; (g) mode of receipt of AAS; (h)
Thus a sample of 80 farmers (40 AAS and 40 non-AAS) weather parameters required; (i) satisfaction from ser-
was chosen for the study as shown in Figure 4. Due to vice (reliability, timely availability, expected benefits,
certain constraints, at times the Nodal Officers carried out frequency) and (j) willingness to pay. Results from a
the study with fewer farmers. few are given in the next section.
• Quantity analysis of inputs used: (a) quantity of seed,
fertilizer, pesticide; (b) number of labour (human,
Survey and the questionnaire
machine) and (c) number of irrigations.
• Price analysis of inputs used: (a) price of seed, ferti-
The sampling method was designed to work directly with
lizer, pesticide; (b) cost of labour (human, machine);
the users of forecast and advisory information, and to
(c) cost of irrigation; (d) cost of product/by-product
meaningfully assess credible cost/loss estimates. The im-
and (e) any other associated cost.
portant issue was to develop an effective and meaningful
base for assessing the impacts of cost-cutting yield and
The major crops chosen for the study were: food grains
enhancing decisions. Cost-cutting measures can take a
(wheat, rice, millets, maize, red gram and chickpea);
variety of forms, some of which include saving in irriga-
oilseeds (mustard); cash crops (cumin, jute, cotton and
tion, reducing the loss of fertilizer, reducing pesticide
tobacco); fruit crops (apricot, peach and banana) and
application, etc. To obtain quantitative information,
vegetables (tomato and spinach).
working relationships between AMFUs and user farmers
were set up through periodic visits. Through such visits,
inputs from the farmers were obtained about use and Survey results of socio-economic features of
application of the advisory bulletins through a pre- farmers
devised questionnaire. Thus the sample survey was not
independently conducted by the agency which provided For the purpose of comparison of the socio-economic
the questionnaire and therefore, a certain amount of bias features of households, the 15 chosen units were divided
was inevitable. This is one of the limitations that has been into four zones: North (Ludhiana, Hisar, Pantnagar,
encountered during the study. Solan); West (Jaipur, Jodhpur, Anand, Pune); East
The survey gave special attention to dates of sowing, (Raipur, Nadia, Bhubaneshwar) and South (Bangalore,
planting, harvesting, spraying, irrigation and tillage Hyderabad, Coimbatore, Thrissur).
operation. Emphasis was given on collecting information
on the adoption of advisory by the farmers during such
operations and the benefit/loss accrued by them on Family structure of farmers
following/not-following the advisories related to such
crucial operations. The family structure in most cases was 3–7 members per
Based on the above methodology and impact assessment family. Educated and prosperous farmers had smaller
framework, the survey was done using three aspects. family size when compared to their counterparts. The pie
chart shown in Figure 5 depicts the age group of farmers
• Socio-economic status: The socio-economic status of in different zones of India. It is seen that in the South,
the farmers was surveyed using queries related to the more than 70% of the farmers are in the age group of 35

CURRENT SCIENCE, VOL. 101, NO. 10, 25 NOVEMBER 2011 1301

GENERAL ARTICLES

Figure 5. Pie chart depicting the age group of farmers in different zones of the country.

Figure 6. Pie chart depicting the educational level of farmers in different zones of the country.

1302 CURRENT SCIENCE, VOL. 101, NO. 10, 25 NOVEMBER 2011

 GENERAL ARTICLES

Figure 7. Pie chart showing the land holdings owned by different farmers of the country.

or more (83%); followed by the East (61%) and North and the rest 51% have less than 5 acres. In the East and
(57%). In the West 47% of the farmers are less than 35 South zones, the farmers generally have small to medium
years of age. On the average most of the farmers belong land holdings ranging between 2.5 and 5 acres (East –
to the middle-level age group. This implies that the 65%; South – 71%). About 22% of the farmers in the East
younger generation may not be interested to take up farm- and 6% in the South have land holdings greater than 10
ing as a profession. acres.

Literacy among farmers Sources of weather forecast

AAS farmers received weather bulletins from different

Figure 6 shows the education level of farmers in the four
sources, viz. radio, television, newspaper; AAS bulletin
zones of the country. It shows that 52% of the farmers in
in printed form/public notice, or through telephone/fax/
the North and West are at least matriculate followed by
personal contact with AAS. Among these, radio and tele-
the East (48%) and South (45%). Although the percent-
vision remained the most popular and easily accessible
age of illiterate farmers is less (about 0–8), it is highest in
media through which 60–70% farmers could get informa-
the East and nil in the West. Interestingly, about 6–17%
tion on weather forecast. Literate farmers comprising
of the farmers are college pass with the West, thus lead-
about 20% could read the detailed weather-based agro-
ing in this category.
advisory bulletin in newspapers and take corrective action.
On several occasions, they contacted agricultural exten-
Size of land holding sion workers and took their advice for various operations.

Figure 7 depicts the size of land holdings of the farmers

in the four zones. In general, farmers in the West have Frequency and reliability of forecast
large land holdings, with 12% of them having land hold-
ings greater than 25 acres, followed by 40% in the 10–25 It was concluded from the study that almost 90% of the
acre category. This is followed by stations in the North farmers gave more importance to forecast of rainfall and
where 23% farmers have land holdings greater than 10 temperature over wind speed, wind direction and cloud
acres; 26% have land holdings in the 5–10 acre category cover. Wind speed and direction and cloud cover play an
CURRENT SCIENCE, VOL. 101, NO. 10, 25 NOVEMBER 2011 1303
GENERAL ARTICLES

important role during infestation of pest and disease. taken, fertilizers applied, harvest technology adopted and
About 60–80% of the AAS farmers were keen to have various other issues. The data collected were eventually
weekly cumulative rainfall forecast for carrying out their used to assess the impact of AAS in economic terms. The
various farm operations. About 20% of the AAS farmers direct impact of AAS on the cost of cultivation, gross net
were interested to obtain the seasonal forecast of rainfall returns and impact on yield was an important part of the
in order to plan their sowing and harvesting operations in economic analysis. The overall analysis in terms of
advance. Almost all AAS farmers opined that the weather percentage of increase in yield and total input cost was
bulletin was received by them on time and 60–70% felt calculated for each of the crops considered36. For conven-
that the rainfall forecast and the temperature trends were ience of the present study, results from two representative
reliable. A large group of 70–80% farmers felt that the stations are presented for each category of crops, viz.
forecast of wind direction was bad. cereal, millets, oilseeds, cash crops, pulses, fruits and
vegetables.
Table 6 gives the cost of cultivation (Rs/acre), gross
Willingness to pay for the service
returns (Rs/acre) and B : C (benefit to cost) ratio for both
AAS and non-AAS farmers during the Kharif and Rabi
The credibility and worthiness of a service is realized by
seasons in all three years. The B : C ratio is higher for the
its acceptability in totality and the readiness on the part of
AAS farmers compared to the non-AAS farmers, though
the user to pay for the service. In the 10–15 years of its
for certain crops the difference is less. Least variation in
existence, AAS had established sufficient credibility and
the B : C ratio is seen for the cereal crops grown in the
reliability among the farming community. As a result, the
northern belt of India (Ludhiana, Pantnagar, Hisar, etc.).
farmers implemented the free advice given through the
This may be attributed to the large land holdings of the
service and also accrued substantial benefits. So it was
farmers who have access to assured irrigation in this part
pertinent to assess through this survey the willingness of
of the country. In general, the vegetable and fruit crops
the farmers to pay for the service. Therefore, a specific
have higher B : C ratio. This is because these crops are
question about the farmers ‘willingness to pay’ was also
highly weather-sensitive and a slight deviation of forecast
included in the survey. Though the reply to this particular
from that realized can have tremendous impact on the
question was neither forthcoming nor overwhelming, it
overall yield. In fact, the cost of cultivation is higher in
definitely helped in analysing the effectiveness and
Rabi 2004–05 for all vegetables. This may be attributed
worthiness of weather forecast and also the risk-taking
to the use of more fertilizers, irrigation and pesticides.
ability of the sample farmers. The situation under which a
The cash crops like cotton, jute and tobacco are generally
farmer was ready to pay included: (a) expected weather-
taken up by farmers who have high risk-taking capability.
induced losses, (b) risk-taking ability of a farmer (income
As a result, in addition to their own experience, they fol-
and assets) and (c) reliability of forecast.
low the weather advisories carefully and thus save on the
Though most of the AAS farmers were not ready to
overall loss to the crop. Thus not much variation is seen
pay and were willing to implement the weather-based
in AAS and non-AAS farmers, except in the case of jute
advisories free of cost, there was a small group of farmers
crop taken up by Kalyani farmers. The AAS farmers who
in Jaipur, Hyderabad and Pune, who were willing to pay
grow Rabi crops like wheat, gram and mustard in North
for the service. This group possesses medium to large
India show higher B : C score compared to crops grown
land holdings. They generally cultivate cash crops and
during Kharif. This again establishes the fact that the
are ready to pay for the service if the price is nominal and
overall skill of the NWP models is higher during the win-
the service is specific to their needs. Farmers with small
ter season.
land holding are unwilling to pay as they are generally
In order to statistically test if the yield of the AAS
poor and take huge loans against their holdings and so do
farmers is significantly higher when compared to the non-
not possess the risk-taking ability.
AAS farmers, we applied the unpaired t-test under the as-
sumption that the data are normally distributed and the
Survey results of quantity and price analysis variances of the two samples are equal. This is so because
there is no one-to-one correspondence between the AAS
Using the questionnaire designed by NCAP, the Nodal and non-AAS farmers. If μ1 and μ2 are the population
Officer carried out the survey at specified time schedules means (yield in this case) of the AAS and non-AAS farm-
that coincided with different operations like land prepara- ers, Y1 and Y2 the sample means, s12 and s22 the sample
tion, sowing, planting, irrigation scheduling, fertilizer variances, and N1 and N2 are the sample sizes, then we
applications, harvesting and post-harvesting operations. test the null hypothesis
The survey included questions on date of operation, the
action taken by the farmer in view of the impending H 0 : μ1 = μ2,
weather/advisory, cost of seed, labour applied in terms of
both machine and human, number of irrigations under- H a : μ1 ≠ μ2,

1304 CURRENT SCIENCE, VOL. 101, NO. 10, 25 NOVEMBER 2011

 GENERAL ARTICLES
Table 6. Impact of the AAS service during the study period

Total cost of cultivation Gross returns Benefit to

(Rs/acre) (Rs/acre) cost ratio
Crop Season Station AAS Non-AAS AAS Non-AAS AAS Non-AAS
Cereals
Wheat 2003–04 Ludhiana 4,908.5 4,902.8 13,136.2 11,921.7 2.68 2.43
2004–05 Ludhiana 5,201.0 5,208.1 16,221.9 14,279.7 3.12 2.74
2005–06 Ludhiana 4,735.2 4,842.7 10,969.9 10,269.1 2.32 2.12
2003–04 Raipur 3,701.4 3,772.1 14,375.6 13,215.0 3.88 3.50
2004–05 Raipur 4,172.9 4,788.0 10,949.4 10,510.5 2.62 2.20
2005–06 Raipur 3,561.7 3,135.7 13,500 11,583.6 3.79 3.69

Paddy 2003 Pantnagar 5,988.2 5,887.3 17,214.6 15,592.3 2.87 2.65

2004 Pantnagar 4,829.3 4,774.3 14,647.1 11,564.8 3.03 2.42
2005 Pantnagar 4,400.3 5,125.6 14,110.7 11,647.7 3.21 2.27
2003–04 Kalyani–Boro 5,955.6 6,028.6 12,851.2 11,677.9 2.16 1.94
2004–05 Kalyani–Boro 5,635.0 5,828.0 12,727.6 11,476.6 2.26 1.97
2005–06 Kalyani–Boro 3,722.0 3,844.2 16,860.3 13,173.2 4.53 3.43

Millets
Pearl millet 2004 Pune 4,332.3 3,446.2 5,435.2 3,743.9 1.25 1.09
2005 Pune 3,734.9 3,949.2 5,980.3 5,106.4 1.60 1.29
2006 Pune 3,686.7 3,492.6 4,864.4 4,372.6 1.32 1.25
2004 Jodhpur 1,151.0 1,021.0 3,397.0 2,182.0 2.95 2.14
2005 Jodhpur 1,569.0 1,326.0 4,976.0 3,708.0 3.17 2.80
2006 Jodhpur 2,686.0 2,241.0 6,016.0 4,681.0 2.24 2.09

Vegetables
Tomato 2004 Coimbatore 13,472.0 13,331.0 72,309.0 60,651.0 5.37 4.55
2005 Coimbatore 15,238.0 16,508.0 58,975.0 51,569.0 3.87 3.12
2006 Coimbatore 18,269.2 19,196.0 81,733.3 72,683.7 4.47 3.79
2003–04 Bhubaneshwar 14,936.1 13,143.5 19,235.5 12,787.7 1.29 0.97
2004–05 Bhubaneshwar 22,794.7 20,632.0 57,767.4 48,256.5 2.53 2.34
2005–06 Bhubaneshwar 11,252.8 11,964.9 34,874.5 33,001.8 3.10 2.76

Palak 2003–04 Hyderabad 10,804.3 10,241.8 29,381.3 20,265.0 2.72 1.98

2004–05 Hyderabad 33,652.1 35,963.3 102,906.3 82,588.1 3.06 2.30
2005–06 Hyderabad 4,811.6 7,750.4 3,541.1 2,246.0 0.74 0.29

Onion 2003–04 Pune 9,726.2 10,087.6 34,225.5 24,500.1 3.52 2.43

2004–05 Pune 10,288.2 8,870.3 24,741.0 20,562.0 2.40 2.32
2005–06 Pune 6,344.3 6,208.3 30,387.0 24,518.2 4.79 3.95

Cash crops
Cotton 2003–04 Coimbatore 10,623.8 12,274.8 15,543.2 16,352.3 1.46 1.33
2004–05 Coimbatore 9,061.8 9,469.4 8,486.4 7,105.1 0.94 0.75
2005–06 Coimbatore 8,628.5 9,201.9 15,687.5 15,304.3 1.82 1.66
2004 Hyderabad 5,330.9 9,243.7 7,743.5 8,333.1 1.45 0.90
2005 Hyderabad 9,736.2 11,959.5 22,063.3 22,031.6 2.27 1.84
2006 Hyderabad 6,378.7 8,780.5 7,743.5 8,333.1 1.21 0.95

Jute 2004 Kalyani 3,352.9 3,268.8 7,402.5 6,704.3 2.21 2.05

2005 Kalyani 4,767.6 5,890.5 21,323.7 19,952.2 4.47 3.39
2006 Kalyani 6,146.0 6,390.0 26,702.4 24,802.4 4.34 3.88

Tobacco 2004–05 Anand 5,504.0 4,740.1 7,713.5 6,205.2 1.40 1.31

2005–06 Anand 5,088.0 5,288.0 9,266.0 8,799.0 1.82 1.66
2006–07 Anand 5,296.0 5,014.0 9,034.0 8,091.0 1.71 1.61

Fruits
Banana 2003–04 Coimbatore 16,096.0 15,022.8 127,549.1 114,226.9 7.92 7.60
2004–05 Coimbatore 31,582.8 31,131.4 83,218.8 78,992.8 2.63 2.54
2005–06 Coimbatore 20,110.1 20,212.8 119,887.7 110,255.2 5.96 5.45
2003–04 Thrissur 54,995.6 53,198 186,370.4 165,484.8 3.39 3.11
2004–05 Thrissur 54,136 56,914.4 192,096 178,487.2 3.55 3.14
2005–06 Thrissur 79,372 74,915.2 186,732 169,052 2.35 2.26
(Contd)

CURRENT SCIENCE, VOL. 101, NO. 10, 25 NOVEMBER 2011 1305

GENERAL ARTICLES
Table 6. (Contd)

Total cost of cultivation Gross returns Benefit to

(Rs/acre) (Rs/acre) cost ratio
Crop Season Station AAS Non-AAS AAS Non-AAS AAS Non-AAS
Apricot 2003–04 Solan 3,976.8 3,541.2 57,743.1 48,518.7 14.52 13.70
2004–05 Solan 4,934.8 2,962.3 69,722.8 40,691.5 14.13 13.74
2005–06 Solan 5,307.3 3,826.5 69,722.8 40,691.5 13.14 10.63

Oil seeds
Mustard 2003–04 Kalyani 3,179.2 3,145.2 10,594.7 9,993.9 3.33 3.18
2004–05 Kalyani 3,085.4 3,163.4 9,896.8 9,187.0 3.21 2.90
2005–06 Kalyani 3,599.3 3,943.3 11,894.0 10,197.7 3.30 2.59
2003–04 Hisar 4,906.0 4,956.4 11,057.8 10,673.4 2.25 2.15
2004–05 Hisar 7,375.0 4,763.8 19,501.2 9,859.9 2.64 2.07
2005–06 Hisar 5,638.4 5,714.9 10,899.9 10,665.0 1.93 1.87

Pulses
Gram 2003–04 Jaipur 3,240.3 3,451.2 10,037.7 9,763.2 3.10 2.83
2004–05 Jaipur 3,871.2 3,961.9 11,927.7 11,070.8 3.08 2.79
2005–06 Jaipur 3,672.9 3,612.3 16,019.7 14,561.6 4.36 4.03

Red gram/tur 2004 Bangalore 3,745.0 4,515.0 12,895.0 9,835.0 3.44 2.18
2005 Bangalore 4,396.0 4,982.7 15,483.8 12,640.3 3.52 2.54
2006 Bangalore 4,448.3 4,997.5 14,427.4 13,577.1 3.24 2.72

using the test statistic: theless, considering the complexity in estimating and
quantifying the user response, the survey results as given
in Tables 6 and 7 provide information about the value of
Y1 − Y2
T= . forecast-cum-advisory products.
s12 /N1 + s22 /N 2 In general in quantitative terms, it is seen that the AAS
farmers were able to reduce the cost of cultivation by
2–5%, except in the case of fruits where the cost of culti-
Under the null hypothesis the statistics follows a t- vation has increased. This shows that the right selection
distribution with N1 + N2 – 2 degrees of freedom. This is of fertilizers and seeds due to organization of awareness
then compared with the tabulated t values at 5% level of programmes in the villages and spraying of appropriate
significance. pesticides according to the advisory saved the input costs.
Table 7 shows the net yield (quintals/acre) of the AAS It was also observed that the yield increased by almost
and non-AAS farmers for each crop, the calculated t val- 10–15% in most of the crops, with maximum benefit in
ues and the tabulated t values for all the stations for each the fruit crops. Undertaking timely field operations by
category of crop, viz. cereal, millets, oilseeds, cash crops, adoption of agro-advisories being disseminated twice a
pulses, fruits and vegetables. It is seen that in general the week by NCMRWF, helped in increasing the yields of
calculated t is more than the tabulated t for majority of various crops. The economic impact assessment is crop-
the crops. This shows that the difference in the yield specific, region-specific and season-specific.
between the AAS and non-AAS farmers is statistically Case studies of specific operations have also been cited
significant at 5% level of significance. The difference is with the gain/loss in economic terms in Table 8. It is seen
more for fruit and horticulture crops. This is also cor- that extreme weather can have severe impact on certain
roborated from Table 6. highly weather-sensitive crops. For example, at Thrissur,
Tables 6 and 7 show economic impact of weather- the banana crop is highly sensitive to heavy rainfall and
based advisories on different crops cultivated by weather- wind speed. Therefore, when heavy rainfall occurred in
sensitive users. Indirectly it assesses what the impacts Thrissur District against a prediction of light to moderate
might have been had the forecasts-cum-advisories not rainfall, the farmers incurred a loss of 12% in the overall
been available. Though the sampling method was devised yield. On the other hand, in Anand the tobacco farmers
to determine the direct and indirect impacts of weather- gained by Rs 667/acre by following the AAS of fertilizer
related costs and losses it was difficult to completely seg- application, as no rainfall was predicted. Table 8 gives a
regate the weather-related impacts from other factors that preliminary indication of the overall role of weather advi-
may also be responsible for the impact on yield. Never- sories in the end yield accrued by the farmer.

1306 CURRENT SCIENCE, VOL. 101, NO. 10, 25 NOVEMBER 2011

 GENERAL ARTICLES
Table 7. Test for significance for the difference in the yield of AAS/ non-AAS farmers

Mean yield (quintals/acre)

Crop Season Station AAS Non-AAS df Calculated t Tabulated t
Cereals
Wheat 2003–04 Ludhiana 32.40 29.70 78 3.93 1.98
2004–05 Ludhiana 31.11 28.30 78 4.80 1.98
2005–06 Ludhiana 33.64 30.33 78 5.27 1.98
2003–04 Raipur 18.06 16.51 78 2.22 1.98
2004–05 Raipur 15.90 14.73 78 4.09 1.98
2006–07 Raipur 12.32 13.42 78 2.96 1.98
2005–07 Raipur 13.80 12.50 78 2.76 1.98

Paddy 2003 Pantnagar 19.12 16.40 78 2.09 1.98

2004 Pantnagar 19.62 15.27 78 3.12 1.98
2005 Pantnagar 23.21 19.32 78 4.21 1.98
2003–04 Kalyani–Boro 46.58 43.32 78 6.22 1.98
2004–05 Kalyani–Boro 40.10 43.63 78 4.12 1.98
2005–06 Kalyani–Boro 35.57 31.17 78 2.62 1.98

Millets
Pearl millet 2004 Pune 23.47 18.58 38 1.72 2.02
2005 Pune 23.18 15.58 38 2.84 2.02
2006 Pune 12.91 11.64 38 3.34 2.02
2004 Jodhpur 9.62 6.72 72 2.62 1.99
2005 Jodhpur 13.68 10.72 72 1.62 1.99
2006 Jodhpur 20.07 18.01 72 2.16 1.99

Vegetables
Tomato 2004 Coimbatore 130.63 110.48 78 2.83 1.98
2005 Coimbatore 110.56 98.77 78 2.94 1.98
2006 Coimbatore 127.67 112.03 78 3.16 1.98
2003–04 Bhubaneshwar 83.61 58.14 78 2.48 1.98
2004–05 Bhubaneshwar 170.44 150.98 78 2.71 1.98
2005–06 Bhubaneshwar 86.28 81.83 78 2.69 1.98

Palak 2003–04 Hyderabad 4.48 3.21 78 2.32 1.98

2004–05 Hyderabad 16.09 12.37 78 2.78 1.98
2005–06 Hyderabad 16.11 15.14 78 1.51 1.98

Onion 2003–04 Pune 87.76 63.86 38 8.21 2.02

2004–05 Pune 76.15 65.75 38 4.40 2.02
2005–06 Pune 76.72 68.79 38 5.27 2.02

Cash crops
Cotton 2003–04 Coimbatore 7.89 6.98 78 3.44 1.98
2004–05 Coimbatore 7.20 6.21 78 3.61 1.98
2005–06 Coimbatore 8.25 7.81 78 3.22 1.98
2004 Hyderabad 5.11 6.11 78 1.99 1.98
2005 Hyderabad 10.31 10.37 78 2.34 1.98
2006 Hyderabad 8.65 7.99 78 2.38 1.98

Jute 2004 Kalyani 21.05 19.92 78 1.32 1.98

2005 Kalyani 36.00 30.00 78 2.38 1.98
2006 Kalyani 38.56 34.42 78 2.29 1.98

Tobacco 2004–05 Anand 8.58 7.57 78 5.17 1.98

2005–06 Anand 6.96 6.44 78 3.09 1.98
2006–07 Anand 5.43 5.16 78 3.31 1.98

Fruits
Banana 2003–04 Coimbatore 214.05 195.49 78 3.02 1.98
2004–05 Coimbatore 172.87 157.58 78 3.23 1.98
2005–06 Coimbatore 208.04 189.51 78 3.30 1.98
2003–04 Thrissur 80.50 74.00 78 2.15 1.98
2004–05 Thrissur 117.25 113.50 78 3.12 1.98
2005–06 Thrissur 98.25 94.00 78 2.03 1.98
(Contd)

CURRENT SCIENCE, VOL. 101, NO. 10, 25 NOVEMBER 2011 1307

GENERAL ARTICLES
Table 7. (Contd)

Mean yield (quintals/acre)

Crop Season Station AAS Non-AAS df Calculated t Tabulated t
Apricot 2003–04 Solan 47.54 39.40 78 2.04 1.98
2004–05 Solan 39.91 35.65 78 2.38 1.98
2005–06 Solan 38.27 32.56 78 4.56 1.98
Oil seeds
Mustard 2003–04 Kalyani 6.36 6.11 78 3.28 1.98
2004–05 Kalyani 5.87 5.65 78 1.88 1.98
2005–06 Kalyani 7.15 6.08 78 8.92 1.98
2003–04 Hisar 14.81 14.11 88 2.68 1.98
2004–05 Hisar 15.59 14.59 88 7.21 1.98
2005–06 Hisar 22.36 21.40 88 5.87 1.98
Pulses
Gram 2003–04 Jaipur 19.80 18.17 88 1.66 1.98
2004–05 Jaipur 17.91 16.23 88 2.23 1.98
2005–06 Jaipur 16.72 14.57 88 2.68 1.98
Red gram/tur 2004 Bangalore 7.24 6.99 48 3.16 1.98
2005 Bangalore 8.58 7.57 48 5.17 1.98
2006 Bangalore 7.80 7.35 48 2.99 1.98

Table 8. Case studies of the impact of AAS

Loss/gain achieved due to the

Date of AAS recommendation
recommendation in Whether AAS
Season/ Weather parameter light of the prevailing recommendation Total cost of cultivation Net returns
Station operation crucial to the crop weather for that operation was followed (Rs/acre) (Rs/acre)
Thrissur Kharif Cloudy weather, 13 and 27 June and 43% of farmers 367 850
Paddy 2006 high relative 4 July 2006 followed
Spraying humidity and low Recommendation:
temperature Infestation of leaf folder is
seen in paddy, use a thorny
stick and open the folded
leaves, spray Monocrotophos/
Quinalphos/Carbary
Hyderabad Kharif Cloud cover, 29 September 2006 Followed 3682 4361
Cotton 2006 rainfall Recommendation: (26%) (39%)
Pesticides No rain forecast, so spray
spraying crop with Monocrotophos
Thrissur Rabi Heavy rainfall 20 and 27 June 2006 Followed Heavy rainfall occurred. Due to
Banana 2006–07 and wind speed Recommendation: this extreme rainfall event
(irrigated) Strengthening Light to moderate rainfall is there was 12% yield loss and
of propping expected in and around 20.3% loss in net return
and drainage Thrissur District
Jaipur Rabi Minimum 23 January 2007 Followed 250 Frost
Gram 2005–06 temperature Recommendation: occurred.
Plant Drop in minimum temperature Increase
protection by 3–4°C. Adopt protection by 10%
against frost
Anand Kharif Rainfall 15 August 2005 Followed 1535 667
Tobacco 2005 Recommendation:
Fertilizer No rainfall forecast, apply
application recommended basal dose of
fertilizer
Jaipur Kharif Rainfall 31 August 2004 Followed Light rainfalls Negative
Pearl 2004 Recommendation: occurred. Top impact
millet Fertilizer No rainfall forecast. dressed fertilizer on net
application Top dressing of urea is suggested wasted, loss of returns
in view of dry weather Rs 114.6/acre

1308 CURRENT SCIENCE, VOL. 101, NO. 10, 25 NOVEMBER 2011

 GENERAL ARTICLES

Limitations of the study this, first, the overall skill and reliability of the meteoro-
logical forecast should increase with special emphasis on
As the study was designed and conducted in the most im- extreme weather events. Secondly, efforts should be
partial way, there is a likelihood that some unexpected made to establish the acceptability factor among the
but unavoidable bias might have percolated into the sur- farming community about the use of weather informa-
vey. This includes sampling bias; obtaining a mutually tion-based farm advisories and the overall benefits of the
exclusive set of AAS and non-AAS farmers regarding service. Moreover, for successful implementation of this
their awareness about weather-based agro-advisories, service, scientific agromet forecasts need blending with
partial/incorrect information collected during the survey; local technologies like traditional methods so that farmers
willfully concealing information about the actual benefits can readily adopt and be benefited from these forecast.
accrued by the farmer; fictitious information regarding India experiences a range of weather phenomena that
the losses suffered owing to weather, in order to obtain may or may not be hazardous. These vary in both time
funds from the government. Though due care was taken and space scales. MoES has a key responsibility to pro-
to avoid such types of shortfalls/deficiencies while ana- vide weather warnings/forecasts on various weather phe-
lysing the data collected during the survey, it is impossi- nomena. While the costs of providing these services are
ble to evade all of them and therefore, some of these quantifiable through the budget of MoES, little effort has
might have influenced the final results. been made to quantify the economic benefits of the mete-
orological services. The success of the survey presented
in this study gives an impetus to carry out similar studies
Conclusion
in other user sectors of meteorological services such as
aviation, power, etc.
The AAS programme of MoES is an innovative inter-
departmental extension service, with a goal to deliver
weather-wise management of agriculture. Although initial 1. Stewart, J. I., Response Farming in Rainfed Agriculture, Wharf
evaluation of AAS has been quite favourable, they have Foundation Press, California, USA, 1988, p. 103.
2. Bagrov, N. A., On economic value of the forecasts. Meteorol.
been quantitative in nature and are based on descriptive
Gidrol., 1966, 2, 3–12 (in Russian).
analyses of results of structured surveys. Hence more 3. Ehrendorfer, M. and Murphy, A. H., On the relationship between
work needs to be carried out. forecast quality and forecast value in the standard cost–loss ratio
It may be concluded that AAS of MoES has helped in situation. In Programme on Short- and Medium-Range Weather
bringing out substantial awareness among farmers about Prediction Research, World Meteorological Organization,
Geneva, Switzerland, 1987, 25, 157–162.
adoption of weather-based advisories, their timely avail-
4. Gandin, L. S. and Zhukovsky, E. E., On the rational use of pre-
ability and quality of service. It has also helped in dicted and climatic information in economic decision-making.
encouraging the adoption and use of modern agricultural Meteorol. Gidrol., 1972, 2, 18–26 (in Russian).
production technologies and practices, in promoting 5. Hansen, J. W., Integrating seasonal climate prediction and agricul-
weather-based irrigation management, pest/disease man- tural models for insights into agricultural practice. Philos. Trans.
R. Soc. London, Ser. B, 2005, 360, 2037–2047.
agement, etc. along with greater use of post-harvest tech-
6. Hansen, J. W. and Indeje, M., Linking dynamic seasonal climate
nologies and commercial marketing of commodities. forecasts with crop simulation for maize yield prediction in semi-
Despite positive effects of AAS on adoption of im- arid Kenya. Agric. For. Meteorol., 2004, 125, 143–157.
proved production technologies and practices, marginal 7. Hansen, J. W., Challinor, A., Ines, A., Wheeler, T. and Moron, V.,
differences were found in the yield obtained by AAS and Translating climate forecasts into agricultural terms: advances and
challenges. Climate Res., 2006, 33, 27–31.
non-AAS farmers for some crops. This may be attributed
8. Meinke, H. and Stone, R. C., Seasonal and inter-annual climate
to certain other factors like shortage of capital, shortage forecasting: the new tool for increasing preparedness to climate
of irrigation water, lack of adequate farmland, unfavour- variability and change in agricultural planning and operations.
able weather patterns and problems of pests and diseases. Climatic Change, 2005, 70, 221–253.
These are a few aspects that highlight that the quality of 9. Sivakumar, M. V. K., Climate prediction and agriculture: current
status and future challenges. Climate Res., 2006, 33, 3–17.
advisory services is not the only vital factor that influ-
10. Harrison, M., Kanga, A., Magrin, G. O., Hugo, G., Tarakidzwa, I.,
ences technology adoption and productivity, and that Mullen, C. and Meinke, H., Use of seasonal forecasts and climate
there is urgent need for complementary progress in other prediction in operational agriculture. CAgM Report No. 102,
areas as well. WMO/TD No. 1344, World Meteorological Organization, Com-
mission for Agricultural Meteorology, Geneva, Switzerland, 2007,
p. 81.
Scope for future work 11. Patt, A. and Gwata, C., Effective seasonal climate forecast appli-
cations: examining constraints for subsistence farmers in Zim-
babwe. Global Environ. Change, 2002, 12, 185–195.
The economic benefits of AAS should be made an inte-
12. Nicholls, J. M., Economic and social benefits of climatological
gral part of AAS of MoES to recognize and acknowledge information and services: a review of existing assessments,
the impact of weather on different crops grown in differ- WMO/TD-No. 780, World Meteorological Organization, Geneva,
ent agro-climatic zones of the country. To accomplish Switzerland, 1996, p. 38.

CURRENT SCIENCE, VOL. 101, NO. 10, 25 NOVEMBER 2011 1309

GENERAL ARTICLES
13. Everingham, Y. L., Muchow, R. C., Stone, R. C., Inman-Bamber, 26. Maini, P., Development of statistical–dynamical models for loca-
G., Singels, A. and Bezuidenhout, C. N., Enhanced risk manage- tion specific weather forecast, Ph D thesis, Andhra University,
ment and decision-making capability across the sugarcane industry 2006, p. 242.
value chain based on seasonal climate forecasts. Agric. Syst., 27. Maini, P., Kumar, A., Singh, S. V. and Rathore, L. S., Statistical
2002, 74(3), 459–477. interpretation of NWP products in India. Meteorol. Appl., 2002, 9,
14. Gadgil, S., Seshagiri Rao, P. R. and Narahari, K., Use of climate 21–31.
information for farm-level decision-making: rainfed groundnut in 28. Wilks, D. S., Statistical Methods in the Atmospheric Sciences. An
southern India. Agric. Syst., 2002, 74(3), 431–457. Introduction, Academic Press, 1995, p. 467.
15. Ingram, K. T., Roncoli, M. C. and Kirshen, P. H., Opportunities 29. Murphy, A. H. and Daan, H., Forecast evaluation. In Probability,
and constraints for farmers of west Africa to use seasonal precipi- Statistics and Decision Making in the Atmospheric Science (eds
tation forecasts with Burkina Faso as a case study. Agric. Syst., Murphy, A. H. and Katz, R. W.), Westview, Boulder, CO, USA,
2002, 74(3), 331–349. 1985, pp. 379–437.
16. Hansen, J. W., Realising the potential benefits of climate predic- 30. Schaefer, J. T., The critical success index as an indicator of warn-
tion to agriculture: issues, approaches, challenges. Agric. Syst., ing skill. Weather Forecast., 1990, 5, 570–575.
2002, 74(3), 309–330. 31. Woodcock, F., The evaluation of yes/no forecasts for scientific
17. Hammer, G. L., Hansen, J. W., Phillips, J. G., Mjelde, J. W., Hill, and administrative purposes. Mon. Weather Rev., 1976, 83, 249–
H., Love, A. and Potgieter, A., Advances in application of climate 253.
prediction in agriculture. Agric. Syst., 2001, 70(2/3), 515–553. 32. NCMRWF/DST, Monsoon-2004: Progress, Performance Predic-
18. Hansen, J. W., Applying seasonal climate prediction to agricul- tion and Agro-Meteorological Advisories (eds Gupta, A. K. et al.),
tural production. Agric. Syst., 2002, 74(3), 305–307. National Center for Medium Range Weather Forecasting (Depart-
19. Nicholls, N., Advances in long-term weather forecasting. In Cli- ment of Science and Technology), 2004, p. 197.
matic Risk in Crop Production: Models and Management in the 33. NCMRWF/DST, Location-specific forecast including SI forecast
Semi-Arid Tropics and Subtropics (eds Muchow, R. C. and and its evaluation. In Monsoon-2005: Performance of the
Bellamy, J. A.), CAB International, Wallingford, CT, 1991, pp. NCMRWF Global Assimilation Forecast System, National Center
427–444. for Medium Range Weather Forecasting (Department of Science
20. Nicholls, N., Opportunities to improve the use of seasonal climate and Technology), 2006, pp. 125–139.
forecasts. In Applications of Seasonal Climate Forecasting in 34. NCMRWF/MoES, Location-specific forecast including SI forecast
Agricultural and Natural Ecosystems: The Australian Experience and its evaluation. In Monsoon-2006: Performance of the
(eds Hammer, G. L., Nicholls, N. and Mitchell, C.), Kluwer, NCMRWF Global Assimilation Forecast System, National Center
Dordrecht, The Netherlands, 2000, pp. 309–327. for Medium Range Weather Forecasting (Department of Science
21. Stone, R. C. and Meinke, H., Weather, climate, and farmers: an and Technology), 2007, pp. 103–115.
overview. Meteorol. Appl., 2006, 13(1), 7–20. 35. Katz, R. W. and Murphy, A. H., Economic Value of Weather and
22. NCMRWF/DST, Guide for Agrometeorological Advisory Service Climate Forecasts, Cambridge University Press, 1997, p. 222.
(eds Singh, S. V., Rathore, L. S. and Trivedi, H. K. N.), 36. Rathore, L. S. and Maini, P., Economic impact assessment of
NCMRWF (DST), MoES, GoI, New Delhi, 1999, p. 201. agro-meteorological advisory service of NCMRWF, Report no.
23. Das Gupta, M. and Rizvi, S. R. H., An overview of meteorological NMRF/PR/01/2008, 2008, p. 104.
data decoding and pre-processing operational at NCMRWF.
NCMRWF Technical Report No. NMRF/TR/1/2001, 2001, ACKNOWLEDGEMENTS. We thank Dr Shailesh Nayak, Secretary,
p. 33. Ministry of Earth Sciences, Government of India, for encouragement
24. Kumar, A., Maini, P., Rathore, L. S. and Singh, S. V., An opera- and valuable suggestions, and Drs Rahul Nigam, Sunil Kaushik and
tional medium range local weather forecasting system developed Girdhar Dewal for help in analysing the field data. We also thank our
in India. Int. J. Climatol., 2000, 20, 73–87. colleagues at NCMRWF for their help and support during the course of
25. Maini, P., Kumar, A., Singh, S. V. and Rathore, L. S., Operational study, and the Nodal Officers and research scholars who helped in the
model for forecasting location specific quantitative precipitation collection of data for carrying out this study.
and probability of precipitation over India. J. Hydrol., 2004, 288,
170–188. Received 3 September 2009; revised accepted 30 September 2011

1310 CURRENT SCIENCE, VOL. 101, NO. 10, 25 NOVEMBER 2011