Science of the Total Environment 627 (2018) 782–791

Contents lists available at ScienceDirect

Science of the Total Environment

journal homepage: www.elsevier.com/locate/scitotenv

Effects of atmospheric stability and urban morphology on daytime

intra-urban temperature variability for Glasgow, UK
Patricia Drach a, Eduardo L. Krüger b,⁎, Rohinton Emmanuel c
a
Departamento de Arquitetura e Urbanismo, Universidade Estadual do Rio de Janeiro, Rio de Janeiro, Brazil
b
Departamento de Construção Civil, Universidade Tecnológica Federal do Paraná, Campus Curitiba - Sede Ecoville, Rua Deputado Heitor Alencar Furtado, 4900, 81280-340 Curitiba, PR, Brazil
c
Glasgow Caledonian University, School of the Engineering and Built Environment, Glasgow, UK

H I G H L I G H T S G R A P H I C A L A B S T R A C T

• Background atmospheric conditions

contributes to explain daytime intra-
urban temperature variability.
• Temperature variations tend to be more
accentuated in less stable atmospheric
classes.
• Variability in air temperature is mostly
noticed in urban canyons and less so in
open-air situations.

Relative air-temperature difference variations to the reference weather station for varying atmospheric conditions – determined
from the range of air-temperature differences for all 32 point locations

a r t i c l e i n f o a b s t r a c t

Article history: This study investigates the joint effect of atmospheric conditions and urban morphology, expressed as the Sky
Received 25 October 2017 View Factor (SVF), on intra-urban variability. The study has been carried out in Glasgow, UK, a shrinking city
Received in revised form 9 January 2018 with a maritime temperate climate type, and findings could guide future climate adaptation plans in terms of
Accepted 28 January 2018
morphology and services provided by the municipality to overcome thermal discomfort in outdoor settings. In
Available online xxxx
this case, SVF has been used as an indicator of urban morphology. The modified Pasquill-Gifford-Turner (PGT)
Editor: Scott Sheridan classification system was adopted for classifying the temperature monitoring periods according to atmospheric
stability conditions. Thirty two locations were selected on the basis of SVF with a wide variety of urban shapes
Keywords: (narrow streets, neighbourhood green spaces, urban parks, street canyons and public squares) and compared
Intra-urban temperature differences to a reference weather station during a total of twenty three transects during late spring and summer in 2013.
Atmospheric stability Maximum daytime intra-urban temperature differences were found to be strongly correlated with atmospheric
Urban morphology stability classes. Furthermore, differences in air temperature are noticeable in urban canyons, with a direct cor-
Climate change relation to the site's SVF (or sky openness) and with an inverse trend under open-air conditions.
Urban design
© 2018 Elsevier B.V. All rights reserved.

1. Introduction

Although the disadvantages of nocturnal overheating due to urban-

⁎ Corresponding author.
E-mail addresses: patricia.drach@uerj.br (P. Drach), ekruger@utfpr.edu.br (E.L. Krüger) ization in warm climate locations are well recognised, Urban Heat Island
, Rohinton.Emmanuel@gcu.ac.uk (R. Emmanuel). (UHI) effects in cold climate cities can still be a matter of dispute. For

https://doi.org/10.1016/j.scitotenv.2018.01.285
0048-9697/© 2018 Elsevier B.V. All rights reserved.
 P. Drach et al. / Science of the Total Environment 627 (2018) 782–791 783

example, London's UHI phenomenon had many advantageous connota- between the site's SVF and local warming was found to be more pro-
tions in the 1960s (e.g., longer growing season, lower heating require- nounced under given atmospheric conditions.
ments, less de-icing on railway tracks (Chandler, 1965)). Yet, after The effect of urban geometry, as expressed by the site's SVF, on mi-
b40 years, overheating caused by UHIs was recognised as a problem croclimate is one of the most studied aspects of UHIs. An early attempt
(Kolokotroni & Giridharan, 2008; Kolokotroni et al., 2006; Oikonomou to statistically relate urban geometry and microclimate was first re-
et al., 2012) and policies for mitigation and adaptation have been ported by Oke (1981)). Unger (2004) shows a detailed review of subse-
established (Davoudi et al., 2009; Carter, 2011). quent efforts.
In the context of climate change, the importance of tackling inadver- A review of the relationship between SVF and urban air tempera-
tent UHI effects has been increasingly recognised. This is due to both tures found it to be rather weak and contradictory (Unger, 2009). Strong
current urbanization trends and to the growing intensity of risks facing relationships have been reported in specific parts of a city (such as
cities, often affecting environmental conditions and the quality of urban downtown areas), in Gothenburg, Sweden (R2 = 0.78) (Svensson,
life. Appropriate urban planning options could help ameliorate the UHI 2004). Long-term measurements encompassing larger areas showed,
problem and urban microclimate, thereby reducing climate change- however, weaker relationships, as in a study analyzing the entire
related risks (Kleerekoper et al., 2012). In addition to mitigating climate urban area of Szeged, Hungary (R2 = 0.47) (Unger, 2004). A number
change by reducing greenhouse gas emissions, heat management is of studies on the relationship between the SVF and daytime thermal ef-
needed in cold climate cities in terms of using it as a resource in winter fects show divergences, some suggesting a relationship between urban
whereas ameliorating negative consequences in summer (Stone et al., geometry, defined by the SVF, and ambient temperature (Unger, 2004;
2012). However, the role of cities in climate change adaptation is only Svensson, 2004; Souza, 2007), others demonstrating negligible impacts
beginning to be addressed (Hebbert & Jankovic, 2013), with vague and of the SVF on local temperature (Eliasson, 1996; Upmanis & Chen,
uncertain city-specific urban climate change action plans. In particular, 1999).
studies that relate urban form to climate change are still scarce Eliasson & Svensson (2003) showed intra-urban temperature varia-
(Shimoda, 2003). tions reaching up to 9 °C based on data collected during an 18-month
To reduce the risk of overheating due to climate change and as a period at 30 sites in Göteborg, Sweden. Their work focused not only
means of mitigating negative consequences of UHIs, the effectiveness on nighttime variability but also on daytime differences (solar noon,
of certain urban forms needs to be explored when accounting for sea- or 12 pm, local time), thereby allowing comparisons between day and
sonal changes and background atmospheric conditions. Given the in- night conditions. Weather conditions at the times of the field observa-
creasing interest in climate change adaptation and the use of models tions followed a cloudiness/wind classification (clear, cloudy sky,
to evaluate the efficacy of various adaptations (Tomlinson et al., windy, calm). Monitoring sites ranged between dense urban locations
2012), such an assessment would further enhance their extent. to green spaces, with SVF ranging 0.7–1.0. Results showed that, in gen-
eral, temperature variations are more dependent on weather than sea-
2. Background son, the highest correlations (between air temperature and surface
cover) were found for clear, calm conditions, irrespective of season.
The UHI effect in cold climate cities is well known. For example, However, for daytime conditions, variability was more frequent during
comfort, energy and health implications of UHIs in cold climate cities cloudy conditions, irrespective of wind speed. Authors concluded that
(e.g., London) are well described by Mavrogianni et al. (2011). More in- statistically significant temperature differences do exist between
terest in this subject is likely to arise in the coming years due to growing densely built-up areas, large open areas and green areas during windy
changes in global climate. In this sense, Kershaw et al. (2010) developed and cloudy situations, both during day and night. For clear sky condi-
methods to predict the UHI effect in future climate projections for the tions, the lack of statistical significance can be attributed to the small
UK. amount of clear days during the monitoring period.
Exploring land use and meteorological aspects of UHIs in Szeged, Konarska et al. (2016) report findings from a field study involving
Hungary, Unger et al. (2001) reported strong relationships between long-term (two consecutive years) air temperature measurements in
urban thermal excess and distance from the city centre, and between Göteborg, Sweden. Intra-urban thermal variations were analyzed at
urban thermal excess and the built-up ratio. However, the authors ten fixed park and street sites characterized by different types and den-
were not able to relate significant correlations between meteorological sity of vegetation, building geometry, openness and surface cover. The
conditions and UHI intensity. study showed the importance of such spatial characteristics in analyzing
In recent years, there has been more recognition of the need for intra-urban variability in daytime and nighttime air temperature. Al-
more careful analysis of background atmospheric conditions (Lee though most of the sites had some vegetation, with consequent foliage
et al., 2009; Holmer et al., 2012) and synoptic weather patterns shading and evapotranspiration interfering effects (Park Cool Island
(Kolokotsa et al., 2009; Lai & Cheng, 2009) when carrying out UHI mea- ‘PCI’ effect), the relationship between openness and local air tempera-
surements. As observed in a study on the shortcomings of UHI monitor- ture during daytime in the warm half of the year (May–September)
ing and simulation techniques (Mirzaei & Haghighat, 2010), UHIs was positive i.e. followed a direct relationship.
develop from small-scale processes such as human metabolism and Scott et al. (2017) deployed a network of low-cost sensors
meso-scale interactions such as atmospheric forces. Kolokotsa et al. (‘iButtons’) to analyze the intra-urban temperature variability in Balti-
(2009) indicated that anticyclonic conditions greatly contribute to the more, United States. The amount of weather stations employed was
development of UHIs during summer; they used a classification of syn- quite significant (135+ sensors) in that study, though the variability
optic conditions developed by Kassomenos (2003)) as a reference to in- in terms of landscape was low, most sites with significant presence of
vestigate the UHI effect in Chania, Greece. vegetation. Results showed small intra-urban temperature variability
Krüger & Emmanuel (2013) estimated the effects of background at- (as regards daily minima). Authors concluded, from linear regression
mospheric conditions on UHIs and intra-urban air temperature in Glas- analysis that the presence/absence of vegetation is the only reliable pre-
gow, UK. They found that the range of intra-urban air temperature dictor of mean air temperature among the various aspects examined
variability and warming trends at specific urban locations relative to a such as albedo, tree-canopy cover, and distance to the nearest park.
rural condition were accentuated when atmospheric stability is taken The aim of the present study is to understand the effect of back-
into account during field observations. For understanding urban mor- ground atmospheric patterns on intra-urban temperature variability
phology effects on such relationship with a set of air temperature sta- under warm conditions in the maritime temperate climate city of Glas-
tions, authors used the Sky View Factor (SVF) as indicative of gow, UK. Previously published material on Glasgow's UHI (Emmanuel &
morphology attributes at each location surveyed. The relationship Krüger, 2012) point to a warming trend during the last decades.
784 P. Drach et al. / Science of the Total Environment 627 (2018) 782–791

Considering that local UHI still persists despite the fact that Glasgow can Mobile traverses were performed for measuring air temperature in
be classified as a shrinking city (with a significant decrease in popula- selected points in the central area with differing urban morphology at-
tion over time), the rationale behind the present study is to investigate tributes. A reference weather station (Davis Vantage Pro2) was used,
the intra-urban temperature variability and its main drivers as regards sited on the rooftop of a low-rise building (at approximately 10 m
urban morphology. From previous findings of an outdoor thermal com- above ground) on the central city campus of the Glasgow Caledonian
fort study carried out in downtown Glasgow, internationally established University (GCU).
comfort limits were found not to conform to thermal preference of local Air temperatures were measured on 23 days during spring and sum-
population (Krüger et al., 2013), suggesting a lower onset of thermal mer 2013 for a total of 32 visiting points. The urban area wherein tra-
discomfort due to heat expressed in terms of the human thermal com- verses took place was selected according to a previous study by
fort index PET (‘Physiological Equivalent Temperature’ (Höppe, Emmanuel & Loconsole (2015) where authors used the LCZ classifica-
1999)). Both facts, i.e. the need for the city to adapt to climate change tion (Stewart & Oke, 2012) for Glasgow from LIDAR data available
and the diminished tolerance of Glaswegians to warm climatic condi- with local authorities in order to identify potential problem areas as
tions, point to the relevance of the study. From the understanding of regards overheating in summer. The city centre area (Glasgow City Cen-
the true impact of morphology on intra-urban temperature variability tre West and Glasgow City Centre East) was categorized as the LCZ class
under clearly identified weather patterns, future analyses can more ap- ‘compact midrise’.
propriately address urban design strategies for improving microclimate Three main routes (eastbound, central and westbound, as shown in
and, thus, contribute to a rise of urban quality of life in the study Fig. 1) were defined to access all 32 points as expedite as possible within
location. the boundaries of the area of analysis. At each point, a two-minute mon-
itoring time was allowed, assumed to be sufficient to capture slight
changes in air temperature readings during the area covered on foot
3. Method and/or on a bicycle. Air temperature data were sampled every 10 s
and averaged after discarding the first minute in order to account for
The need for tackling possible health impacts of heat-wave episodes, the sensor's stabilization time. Mobile traverses started at 2:30 pm
which will likely increase in frequency due to global climate change, as local time (near the peak air temperature, typically around 3:00 pm)
pointed out by Koppe et al. (2004), requires effective interventions, and lasted approximately one hour. Eastbound and westbound routes
measures and policies to protect the health of vulnerable Europeans in were covered on bicycle while central route was done on foot.
summer. We investigate the relationship between urban morphology The reference weather station, a Davis Vantage Pro2 weather station
on local air temperature profiles accounting for background atmo- was set up at Glasgow Caledonian University and equipped with tem-
spheric conditions during the short periods of field monitoring. The re- perature and humidity sensors, a cup anemometer with a wind vane
search further contributes to findings obtained in a previous study, and pyranometer. For the traverses, we used small data loggers (Tinytag
where Krüger & Emmanuel (2013) evaluated the effect of atmospheric TGP-4500) enclosed in naturally ventilated radiation shields (ACS-5050,
stability on ambient temperature differences using a pair of stationary Stevenson Type Screen).
(urban vs. rural) weather stations and intra-urban differences with a Urban morphology was determined by the site's Sky-View Factor
set of temperature/relative humidity stations for the same city location. (SVF), as measured by fisheye-lens photographs (SIGMA 4.5 mm f 2.8
The present study was carried out in Glasgow, UK (55°52′N, 4°15′ EX) and computed using RayMan Pro (Matzarakis et al., 2010). In
W), a high-latitude location with climate type “5C” (ANSI/ASHRAE terms of SVF, the sampling points represented a wide variety of urban
Standard 90.2, 2004), cool, marine1 and with approximately 600,000 in- forms (narrow street canyons, green areas, urban parks, uniform and
habitants. According to the Koeppen-Geiger climate classification sys- non-uniform street canyons and public squares) (Table 1).
tem, Glasgow lies within a region with temperate climate with Atmospheric stability was classified using the modified Pasquill-
maritime influences (Cfb). The mean maximum temperatures in the Gifford-Turner (PGT) classification ((Turner, 1970), modified by
warmest season (July and August) is lower than 20 °C while mean Mohan & Siddiqui (Mohan & Siddiqui, 1998); Table 2) from data re-
daily temperature is over 10 °C at least during five months of the year trieved from the Weather Underground station “Glasgow EGPF”
(UK MET Office, n.d.). (https://www.wunderground.com/weather/gb/glasgow), comprising
As a continuation of past research initiatives, the study is served by wind speed and solar radiation data. For that, daytime classes were de-
two relevant databases: mobile air-temperature traverses carried out fined according to the mean wind speed and maximum solar radiation
during summer in 2013 and outdoor thermal comfort campaigns con- observed during the monitoring period.
ducted in 2011. The 2013 data were needed for assessing possible rela- Finally, in order to translate local air temperatures to felt tempera-
tionships between morphology and local air temperature whereas the tures in terms of comfort/discomfort, the database obtained from out-
2011 database allowed us to put obtained results into the perspective door comfort campaigns in Glasgow (Höppe, 1999) was used for
of thermal comfort, as perceived by local population as well as to classify relating thermal sensation of passersby to ambient temperatures. Mon-
most recent summer conditions in terms of stability classes. Results are itoring campaigns took place during late winter, spring and summer
structured as follows: 1) classification of atmospheric stability of the 2011, in pedestrian downtown streets within the circumscribed area
days with mobile traverses; 2) comparison between morphology attri- (Fig. 1), at six monitoring points. In those campaigns, meteorological
butes (expressed as the site's SVF) and air-temperature differences to variables were monitored according to ISO standard 7726 (ISO 7726,
the reference weather station; 3) correlation analyses of data subsets; 1998) with a Davis Vantage Pro2 weather station, equipped with tem-
4) assessment of thermal comfort/discomfort levels during the tra- perature and humidity sensors, cup anemometer with wind vane, sili-
verses; 5) representativeness of the PGT atmospheric classes when con pyranometer and gray-painted globe thermometer. The weather
compared to a longer series of summer data. station was positioned next to the passerby and variables were mea-
sured close to the person (approximately 1.5 m) except for the ane-
mometer and pyranometer, which were placed just above head height
(at about 2 m) so that both interviewer and respondent would not affect
1
wind speed readings. Surveys were conducted during daytime, usually
From weather files available at https://energyplus.net/weather. Weather data for
between 10 am and 1 pm. Excluding outliers and respondents who
Glasgow are assumed as at the nearest weather station in Oban, 56°25′N, 5°28′W, 4 m a.
s.l. (Note that the ASHRAE classification shown here is derived algorithmically from the did not meet the criteria of long-term and short-term acclimatization,
source weather data. It may not be indicative of the long-term climate for each location). the sample size for Glasgow was 567 respondents over 19 monitoring
campaigns. We used a thermal comfort questionnaire consisting of
 P. Drach et al. / Science of the Total Environment 627 (2018) 782–791 785

Fig. 1. Study area with monitoring points and traverses.

items related to gender, age, height, weight and clothing; residency at least 15 min); thermal perception and thermal preference.
time in Glasgow (exclusion criterion: less than six months) and expo- Concerning the perceptual evaluation (thermal sensation vote, or
sure time to the outdoors prior to the interview (exclusion criterion: TSV), the basic question was “How do you feel at this exact moment?”,

Table 1
Traverse points' specifications (SVF) and fisheye images.

Point SVF Fisheye image Point SVF Fisheye image Point SVF Fisheye image Point SVF Fisheye image

1 0,6 9 0,2 17 0,3 25 0,5

2 0,4 10 ~0.0 18 0,3 26 0,6

3 0,2 11 0,3 19 0,2 27 0,4

4 0,4 12 0,3 20 0,6 28 0,4

5 0,3 13 0,4 21 0,3 29 0,5

6 0,4 14 0,3 22 0,3 30 0,6

7 0,1 15 0,3 23 0,3 31 0,7

8 0,2 16 0,2 24 0,3 32 0,8

786 P. Drach et al. / Science of the Total Environment 627 (2018) 782–791

Table 2 The interpretation of temperature difference data, when accounting

PGT atmospheric stability classes. for atmospheric stability is shown in Fig. 3. For each PGT class, a mean
WS (m/s) Daytime SR (W/m2) Night time CC (octas) daytime pattern was determined from the various daytime periods be-
Higha Modb Lowc Cloudy Lowd Mode Highf
longing to that particular class.
The most stable condition (class C) shows less differentiation be-
≤2 A A–B B C G–F F D
tween point temperatures. The same finding is confirmed in respect of
2–3 A–B B C C F E D
3–5 B B–C C C E D D relative air temperature difference (to the reference weather station at
5–6 C C–D D D D D D GCU) variations among the 32 monitoring points (Fig. 4). Variations
N6 C D D D D D D are much more pronounced for the first two classes (A and A–B),
Legend: WS wind speed, SR global solar radiation, CC cloud cover. dropping consistently for the following categories. The explanation for
a
(N600). that lies in the fact that stronger radiation and weaker wind speed,
b
(300–600). which define the most unstable PGT classes, in conjunction with the
c
(b300).
d site's morphology, produce more noticeable effects on intra-urban tem-
(0–3).
e
(4–7). perature variability. In terms of morphology, the obtained effects mainly
f
(8), A (highly unstable or convective), B (moderately unstable), C (slightly unstable), include shading effects resulting from surrounding buildings on a given
D (neutral), E (moderately stable), and F (extremely stable), G (extremely stable, low site.
wind).

4.1. Strength of correlation due to morphology

using the symmetrical 7-degree two-pole scale ranging from −3
(“cold”) to +3 (“hot”), whose responses are, in this paper, solely corre- The ensemble of points with varying SVF attributes, yet with a par-
lated to ambient temperature concurrently monitored next to the ticular shape is analyzed separately. From Table 1, points 3, 5, 7, 10, 16
interviewed person. A thorough description of the survey conditions is and 23 are within street canyons with an almost E-W axis orientation.
shown in Krüger et al. (2013). Points 20, 30, 31 and 32 are located in open-air sites. The SVF variation
in the first set of points is within the range 0.04–0.31; for the second
4. Results subset within 0.55–0.78. Thus, the SVF variation in both subsets is sim-
ilar, averaging 0.25.
The arrangement of the daytime periods according to their PGT clas- Correlations between SVF variations and intra-urban temperature
ses is given in Table 3, along with temperature, measured at the weather differences have been drawn for subsets of the data sample under the
station at GCU, wind speed and solar radiation attributes during the full ensemble of atmospheric conditions, namely for the six points lo-
transects, as measured at the Weather Underground station ‘Glasgow cated in E-W street canyons and for four points which had greater sky
EGPF’. openness (SVF N 0.5) (Fig. 5). As such correlations refer to a limited
Inter-comparisons between monitoring points yielded diverse rela- number of points, absolute values do not possess statistical strength,
tionships, with no discernible pattern or consistency over the various though overall trends (direct or inverse relationship to SVF) can be
days of observation. Local temperatures (i.e. at each point location) regarded as indicative of morphology effects on local air temperature.
are interpreted in terms of relative differences to the reference station For the points located in E-W street canyons, correlations between rel-
located at GCU. Fig. 2 shows all differences obtained, superimposed ative air temperature differences to the reference weather station at
day by day. GCU are mostly positive (20 out of 23 days), irrespective of atmospheric
Correlations found between relative differences to the reference patterns. For the open-air sites, correlations are normally negative
weather station at GCU and each point's SVF were varied, ranging − (again, 20 out of 23 days) exhibiting an inverse relationship between
0.56 to 0.37 (with a mean of −0.01), with inverse and direct relation- SVF and air temperature variability. An explanation is that in more
ships between urban density and local air temperature depending on constrained locations, solar radiation gains in the canyon will increase
daytime weather conditions on a given day. heat trapping and enhance sensible heat, whereas open-air locations

Table 3
Array of transect days, climate descriptors and corresponding PGT classes.

Date Minimum temperature (°C) Maximum temperature (°C) Wind speed (m/s) Solar radiation (W/m2) Atmospheric stability Class (PGT)

May 21th 18.2 19.3 1.8 591 A

July 18th 28.0 28.6 1.2 816 A
July 19th 27.5 28.5 1.1 779 A
May 22th 14.5 15.9 2.4 656 A–B
May 30th 19.3 19.9 1.9 359 A–B
June 5th 19.7 20.8 1.1 330 A–B
June 6th 21.5 23.3 0.4 549 A–B
June 7th 19.7 20.3 2.5 816 A–B
June 17th 19.6 20.6 0.5 398 A–B
June 18th 20.8 22.3 1.8 539 A–B
June 26th 19.3 20.4 2.6 830 A–B
August 13th 19.0 23.8 2.1 699 A–B
May 20th 22.9 23.8 2.1 514 B
May 29th 18.7 19.5 1.0 207 B
June 19th 19.2 20.0 2.9 589 B
June 25th 17.9 18.6 2.6 363 B
June 27th 12.8 13.1 1.5 130 B
July 1st 15.2 16.7 3.0 491 B
May 9th 10.3 11.5 3.2 110 C
May 10th 10.0 10.7 2.6 147 C
May 28th 12.6 13.3 2.2 169 C
July 3rd 15.8 16.6 2.3 224 C
August 14th 19.3 21.0 2.3 117 C
 P. Drach et al. / Science of the Total Environment 627 (2018) 782–791 787

Fig. 2. Relative air-temperature differences to the reference station located at GCU, day by day.

promote more heat losses due to ventilation; in the first case, radiation all 32 points) for the canyon situations and are virtually nonexistent in
is the driver, in the latter, air movement. the case of the open-air locations, which shows that such factor (atmo-
Patterns of relationship between relative difference variations and spheric stability) loses its explanatory power entirely in less constrained
atmospheric classes are not as strong as for the whole dataset (i.e., for locations (Fig. 6).

Fig. 3. Relative air-temperature differences to the reference station located at GCU, accounting for atmospheric stability classes (PGT scheme).
788 P. Drach et al. / Science of the Total Environment 627 (2018) 782–791

Fig. 4. Relative air-temperature difference variations to the reference weather station at GCU for varying atmospheric conditions – determined from the range of air-temperature
differences for all 32 point locations.

Fig. 5. Correlations between relative air-temperature difference variations to the reference weather station at GCU and SVF attributes.
 P. Drach et al. / Science of the Total Environment 627 (2018) 782–791 789

Fig. 6. Relative air-temperature difference variations to the reference weather station at GCU for varying atmospheric conditions – a) for E-W street canyons, b) for open-air locations.

4.2. Thermal comfort within −0.5 and +0.5 thermal sensation vote, normally assumed to
be a feasible comfort range (ISO 7730, 2005).
The thermal votes obtained in the outdoor comfort campaigns over Using the regression formula, the maximum MTS increase relative to
late winter, spring and summer 2011 were binned for each one- the reference weather station at GCU for the 23 traverse days resembles
degree centigrade variation in ambient temperature. Fig. 7 shows the the drop of intra-urban temperature variations shown in Fig. 4 with ris-
obtained graph for the range of air temperatures monitored, with re- ing atmospheric stability. The first two classes (A and A–B) show a max-
spective (binned) thermal sensation votes. imum rise in thermal sensation corresponding to a change in MTS
The regression line and formula give us a comfort band between 11 category, e.g. moving from ‘neutral’ to “slightly warm” or from ‘slightly
and 16 °C (Ta) where the reported mean thermal sensation (MTS) lies warm’ to ‘warm’. In terms of heat stress situations assuming a minimum

Fig. 7. Binned thermal sensation versus air temperature Ta (from outdoor comfort campaigns).
790 P. Drach et al. / Science of the Total Environment 627 (2018) 782–791

threshold of 16 °C for Glaswegians, 17 out of 23 days were above that References

threshold at the reference station. The average increase in the transect
ANSI/ASHRAE Standard 90.2, 2004. American Society Of Heating, Refrigerating, and Air-
points resulted in 19 out of 23 days with heat stress. Conditioning Engineers (Energy-Efficient Design of Low-Rise Residential Buildings).
Beniston, M., 2004. The 2003 heat wave in Europe: a shape of things to come? An analysis
based on Swiss climatological data and model simulations. Geophys. Res. Lett. 31 (2).
Carter, J.G., 2011. Climate change adaptation in European cities. Curr. Opin. Environ. Sus-
4.3. Dominant atmospheric class in summer for Glasgow tain. 3 (3), 193–198.
Chandler, A.J., 1965. The Climate of London. Hutchinson and Co., London Available at.
Continuous data for 2011 (solar radiation and wind speed), mea- http://urban-climate.com/wp3/wp-content/uploads/2011/04/TonyChandler_
ClimateOfLondon.pdfN (accessed September 2017).
sured at the reference weather station at GCU, allow us to assess the Chen, L., Ng, E., 2012. Outdoor thermal comfort and outdoor activities: a review of re-
most frequent atmospheric stability classes during daytime in summer search in the past decade. Cities 29 (2), 118–125.
in Glasgow, thereby confirming the trend obtained during the monitor- Davoudi, S., Crawford, J., Mehmood, A., 2009. Planning for Climate Change: Strategies for
Mitigation and Adaptation for Spatial Planners. Routledge.
ing days. From Table 3, the dominant class is A-B (9 out of 23 days or 40%
Eliasson, I., 1996. Urban nocturnal temperatures, street geometry and land use. Atmos.
of the total traverse days). Hourly data for 2011 were interpreted in Environ. 30 (3), 379–392.
terms of daytime periods (when solar radiation was present) for a Eliasson, I., Svensson, M.K., 2003. Spatial air temperature variations and urban land use - a
total of 85 days between summer solstice and fall equinox, disregarding statistical approach. Meteorol. Appl. 10, 135–149.
Emmanuel, R., Krüger, E., 2012. Urban Heat Island and its impact on climate change resil-
gaps in the dataset. From that total, 47% of the days exhibited an A-B at- ience in a shrinking city: the case of Glasgow, UK. Build. Environ. 53, 137–149.
mospheric condition according to the PGT classification scheme. Emmanuel, R., Loconsole, A., 2015. Green infrastructure as an adaptation approach to
tackling urban overheating in the Glasgow Clyde Valley Region, UK. Landsc. Urban
Plan. 138, 71–86.
Gronlund, C.J., Zanobetti, A., Schwartz, J.D., Wellenius, G.A., O'Neill, M.S., 2014. Heat, heat
5. Summary and conclusions waves, and hospital admissions among the elderly in the United States, 1992–2006.
Environ. Health Perspect. 122 (11), 1187.
Hebbert, M., Jankovic, V., 2013. Cities and climate change: the precedents and why they
In summary, results from the traverse study lead us to conclude that matter. Urban Stud. 50, 1332–1347.
accounting for background atmospheric conditions is advisable to better Holmer, B., Thorsson, S., Lindén, J., 2012. Evening evapotranspirative cooling in relation to
explain intra-urban temperature differences during daytime. Tempera- vegetation and urban geometry in the city of Ouagadougou, Burkina Faso. Int.
J. Climatol. 33, 3089–3105.
ture variations will tend to be more accentuated in less stable atmo-
Höppe, P., 1999. The physiological equivalent temperature—a universal index for the biomete-
spheric classes, under sunny conditions with light wind, with more orological assessment of the thermal environment. Int. J. Biometeorol. 43 (2), 71–75.
remarkable effects on pedestrian thermal sensation. Furthermore, dif- Intergovernmental Panel on Climate Change, 2007. Climate change 2007: The physical
science basis. In: Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt,
ferences in air temperature are noticeable in urban canyons, with a di-
K.B., et al. (Eds.), Contribution of Working Group I to the Fourth Assessment Report
rect correlation to the site's SVF (or sky openness) and with an inverse of the Intergovernmental Panel on Climate Change. Cambridge University Press, Cam-
trend under open-air conditions. Finally, it has been found that the bridge, UK.
dominant PGT class for summer time in Glasgow is among the classes ISO 7726, 1998. Ergonomics of the thermal environment - instruments for measuring
physical quantities. International Organisation for Standardisation, Geneva.
which yielded the highest intra-urban variability. Such results comple- ISO 7730, 2005. Moderate Thermal Environments - Determination of the PMV and PPD
ment findings related to Glasgow's Urban Heat Island intensity Indices and Specification of the Conditions for Thermal Comfort. International Orga-
(Krüger & Emmanuel, 2013). In that study, it had been found that, nisation for Standardisation, Geneva.
Kassomenos, P., 2003. Anatomy of the synoptic conditions occurring over southern
from a set of urban and suburban temperature/relative humidity (T/ Greece during the second half of the 20th century. Part I. Winter and summer.
RH) stations, intra-urban air temperature differences during nighttime Theor. Appl. Climatol. 75, 65–77.
were accentuated when accounting for Pasquill-Gifford-Turner (PGT) Kershaw, T., Sanderson, M., Coley, D., Eames, M., 2010. Estimation of the urban heat island
for UK climate change projections. Build. Serv. Eng. Res. Technol. 31, 251–263.
stability classes. Kleerekoper, L., van Esch, M., Salcedo, T.B., 2012. How to make a city climate-proof, ad-
As pointed out by Eliasson & Svensson (2003), the effect of different dressing the urban heat island effect. Resour. Conserv. Recycl. 64, 30–38.
surface coverings on air temperature patterns is more pronounced dur- Kolokotroni, M., Giridharan, R., 2008. Urban heat island intensity in London: an investiga-
tion of the impact of physical characteristics on changes in outdoor air temperature
ing clear and calm weather with high air stability. Windy and cloudy during summer. Sol. Energy 82, 986–998.
conditions smooth out these differences. In this sense, our results con- Kolokotroni, M., Giannitsaris, I., Watkins, R., 2006. The effect of the London urban heat is-
firm such expected behavior (cf. Fig. 3). land on building summer cooling demand and night ventilation strategies. Sol. En-
ergy 80, 383–392.
Considering predicted increases in frequency, intensity and duration
Kolokotsa, D., Psomas, A., Karapidakis, E., 2009. Urban heat island in southern Europe: the
of heat-wave episodes (Beniston, 2004; Intergovernmental Panel on case study of Hania, Crete. Sol. Energy 83, 1871–1883.
Climate Change, 2007; Russo et al., 2014) and consequences of outdoor Konarska, J., Holmer, B., Lindberg, F., Thorsson, S., 2016. Influence of vegetation and build-
exposure to extreme heat or less so to warmer conditions in summer, ing geometry on the spatial variations of air temperature and cooling rates in a high-
latitude city. Int. J. Climatol. 36 (5), 2379–2395.
pedestrian areas can be seriously affected in terms of a drop in the num- Koppe, C., Kovats, S., Jendritzky, G., Menne, B., Breuer, D.J., 2004. Heat Waves: Risks and
ber of visitors and as regards their attractiveness (Chen & Ng, 2012). Responses. World Health Organization, Regional Office for Europe.
Furthermore, cities like Glasgow could be seriously impacted in terms Krüger, E., Emmanuel, R., 2013. Accounting for atmospheric stability conditions in urban
heat island studies: the case of Glasgow, UK. Landsc. Urban Plan. 117, 112–121.
of mortality. As pointed out by Gronlund et al. (2014), heat-related mor- Krüger, E., Drach, P., Emmanuel, R., Corbella, O., 2013. Assessment of daytime outdoor
tality was found to be stronger particularly in North-American cities comfort levels in and outside the urban area of Glasgow, UK. Int. J. Biometeorol. 57
with milder summers and lower air conditioning prevalence. Thus, im- (4), 521–533.
Lai, L.-W., Cheng, W.-L., 2009. Air quality influenced by urban heat island coupled with
provements in outdoor thermal conditions to be met by sound urban synoptic weather patterns. Sci. Total Environ. 407, 2724–2733.
design measures should be sought in those areas. Lee, S., Lee, K., Jin, W., Song, H., 2009. Effect of an urban park on air temperature differ-
From the trends shown in this paper, further work should focus on ences in a central business district area. Landsc. Ecol. Eng. 5, 183–191.
Matzarakis, A., Rutz, F., Mayer, H., 2010. Modelling radiation fluxes in simple and complex
urban density transitions (here expressed as the site's SVF, though environments: basics of the RayMan model. Int. J. Biometeorol. 54, 131–139.
other morphology indicators could and should additionally be tested Mavrogianni, A., Davies, M., Batty, M., et al., 2011. The comfort, energy and health impli-
in the future) that affect the pattern of relationship between local air cations of London's urban heat island. Build. Serv. Eng. Res. Technol. 32, 35–52.
Mirzaei, P.A., Haghighat, F., 2010. Approaches to study urban heat islands - abilities and
temperature and built density, as observed in Fig. 5. Parametric analysis
limitations. Build. Environ. 45, 2192–2201.
could be performed by means of computer simulation in order to draw Mohan, M., Siddiqui, T.A., 1998. Analysis of various schemes for the estimation of atmo-
additional conclusions on the effects of density and local air spheric stability classification. Atmos. Environ. 32 (21), 3775–3781.
temperature when accounting for given orientations, seasons of the Oikonomou, E., Davies, M., Mavrogianni, A., Biddulph, P., Wilkinson, P., Kolokotroni, M.,
2012. Modelling the relative importance of the urban heat island and the thermal
year and day types. Finally, it should be stressed that the sample ob- quality of dwellings for overheating in London. Build. Environ. 57, 223–238.
tained shows limitations in terms of the characteristics of the points Oke, T.R., 1981. Canyon geometry and the nocturnal urban heat island: comparison of
surveyed. scale model and field observations. J. Climatol. 1, 237–254.
 P. Drach et al. / Science of the Total Environment 627 (2018) 782–791 791

Russo, S., Dosio, A., Graversen, R.G., Sillmann, J., Carrao, H., Dunbar, M.B., Vogt, J.V., 2014. Turner, D.B., 1970. Workbook of atmospheric dispersion estimates. Office of Air Program
Magnitude of extreme heat waves in present climate and their projection in a Pub. No. AP-26. Environmental Protection Agency, USA.
warming world. J. Geophys. Res.-Atmos. 119 (22). UK MET Office, d. Western Scotland: Climate. Available at. http://www.metoffice.gov.uk/
Scott, A.A., Zaitchik, B., Waugh, D.W., O'Meara, K., 2017. Intraurban temperature variabil- climate/uk/regional-climates/ws, Accessed date: August 2017.
ity in Baltimore. J. Appl. Meteorol. Climatol. 56 (1), 159–171. Unger, J., 2004. Intra-urban relationship between surface geometry and urban heat island:
Shimoda, Y., 2003. Adaptation measures for climate change and the urban heat island in review and new approach. Clim. Res. 27, 253–264.
Japan's built environment. Build. Res. Inf. 31, 222–230. Unger, J., 2009. Connection between urban heat island and sky view factor approximated
Souza, L.C.L., 2007. Thermal environment as a parameter for urban planning. Energy Sus- by a software tool on a 3D urban database. Int. J. Environ. Pollut. 36, 59–80.
tain. Dev. 11 (4), 44–53. Unger, J., Sümeghy, Z., Gulyás, Á., Bottyán, Z., Mucsi, L., 2001. Land-use and meteorological
Stewart, I.D., Oke, T.R., 2012. Local climate zones for urban temperature studies. Bull. Am. aspects of the urban heat island. Meteorol. Appl. 8, 189–194.
Meteorol. Soc. 93 (12), 1879–1900. Upmanis, H., Chen, D., 1999. Influence of geographical factors and meteorological vari-
Stone, B., Vargo, J., Habeeb, D., 2012. Managing climate change in cities: will climate ac- ables on nocturnal urban-park temperature differences—a case study of summer
tion plans work? Landsc. Urban Plan. 107, 263–271. 1995 in Göteborg, Sweden. Clim. Res. 13 (2), 125–139.
Svensson, M.K., 2004. Sky view factor analysis - implications for urban air temperature
differences. Meteorol. Appl. 11, 201–211.
Tomlinson, C.J., Chapman, L., Thornes, J.E., Baker, C.J., 2012. Derivation of Birmingham's
summer surface urban heat island from MODIS satellite images. Int. J. Climatol. 32,
214–224.