Journal of Arid Environments 98 (2013) 113e122

Contents lists available at ScienceDirect

Journal of Arid Environments

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

Urban warming and global warming: Combined effect on thermal

discomfort in the desert city of Beer Sheva, Israel
Oded Potchter a, *, Hofit Itzhak Ben-Shalom b
a
Department of Geography and the Human Environment, Tel Aviv University and Dept. of Geography, Beit Berl Academic College, Israel
b
The Porter School of Environmental Studies, Tel Aviv University, Israel

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

Article history: The effect of climatic changes on human comfort levels was investigated through examination and
Received 20 March 2013 statistical analysis of long-term trends in human discomfort during summer months in the desert city of
Received in revised form Beer Sheva, Israel and in the adjacent rural area of Wadi Hatzerim. In an era of global warming, the urban
2 August 2013
warming effect is likely to be amplified and as a result increase human discomfort, especially during
Accepted 15 August 2013
Available online 13 September 2013
summer.
Climate data for the city of Beer Sheva over the last 40 years shows an increase of temperature and air
humidity in comparison to the surrounding rural area. Wind velocity data for Wadi Hatzerim show that
Keywords:
Desert city
changes are inconstant and not significant, while in Beer Sheva, wind velocity is significantly reduced.
DI Two indices e the Discomfort Index (DI) and Physiological Equivalent Air temperature (PET) e were used
Hot and arid climate to evaluate the effect of these climatic changes on human discomfort. Although the bio-meteorological
PET indices showed the same tendency of increasing heat stress values and duration, in Beer Sheva they were
Specific humidity more pronounced and more significant than in the desert environment. The study concludes that these
Thermal stress combined climatic effects negatively impact human comfort and are more noticeable in desert cities at
Urban heat island peak daytime hours during summer.
Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction and Gaffen, 2001; Willett et al., 2008). Between 1973 and 2003,
surface-specific humidity has increased significantly throughout
Urban populations are exposed to both urban-induced local the world, and warmer regions exhibit greater increases in specific
climate modification and to the effect of climate change known as humidity for a given air temperature change (Willett et al., 2008). In
global warming. According to the 2007 Intergovernmental Panel on the United States, Gaffen and Ross (1998) found an increase of several
Climate Change (IPCC) report, in the twentieth century the average percentage points per decade in specific humidity, and several tenths
global air temperature increased by 0.76
C, and the linear warming of a degree increase per decade in dew point. These specific humidity
trend over the past 50 years is nearly twice that for the last 100 increases are consistent with upward air temperature trends. Similar
years. This trend of rising global air temperature is likely to results were found in China (Wang and Gaffen, 2001).
continue (IPCC, 2007). Base on intensive literature review, These global climatic changes can potentially increase human
Alcoforado and Andrade (2008) concluded that the impact of global thermal stress. Hence, for example, the mean summertime heat
warming including its affect on human well-being and health may stress in the United States between 1949 and 1995 increased due
be exacerbated in urban areas. Global and regional warming can the combined effect of an increase in air temperature and humidity.
aggravate urban warming during summer months (Fujibe, 2009) These trends may be partly associated with increased urbanization
and may increase the frequency of air temperature extremes in (Gaffen and Ross, 1998).
urban areas and prolong the duration of these events, which are Climatic changes have also been observed in Israel, and are
more pronounced in arid regions (Golden, 2004). associated with global warming (Alpert, 2004). Ben-Gai et al.
Global warming is accompanied by several climatic changes, (1999) analyzed daily maximum and minimum temperatures
such as an increase of specific humidity (Gaffen and Ross,1999; Wang from 40 stations between the years 1964 and 1994 and found
minimum summer temperature increases of 0.26
C per decade and
maximum summer temperature increases of 0.21
C per decade. Ziv
* Corresponding author.
E-mail addresses: potchter@post.tau.ac.il, potchter@beitberl.ac.il (O. Potchter), and Saaroni (2011) analyzed temperature and humidity data from
hofitben@gmail.com (H. Itzhak Ben-Shalom). 16 stations in Israel between the years 1975 and 2008 and found

0140-1963/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.jaridenv.2013.08.006
114 O. Potchter, H. Itzhak Ben-Shalom / Journal of Arid Environments 98 (2013) 113e122

that the average temperature significantly increased by 0.4e1.1
C found to cause an increase in wind velocity during nighttime hours
per decade. Changes in humidity levels, however, were found to be when wind velocities were low, in New York City (Bornstein, 1975)
inconsistent (Shafir and Alpert, 2011). and in Phoenix, Arizona (Balling and Cerveny, 1987).

1.1. Urban climatic changes 1.2. Urban warming and human discomfort

The urban climate is characterized by amplification of air tem- One of the most important implications of urban warming is the
perature values, lower humidity values, and moderation of wind increase of heat stress during the day in the summer, which ag-
velocity (Landsberg, 1981; Oke, 1987). Increased air temperature gravates humans’ thermal discomfort (Baker et al., 2002). A study of
trends in urban sites have been found in many regions globally, and Phoenix, Arizona showed that rapid urbanization over a period of
are mostly due to urban growth (Choi et al., 2003; Karl et al., 1988). fifty years has caused the city to warm up by 3
C, and has doubled
Multiple studies have shown that the intensity of urban warming the number of hours with T > 38
C per day during the months of
increases with the size of the city (i.e., Oke, 1987; Roth, 2007). July and August (Baker et al., 2002). In contrast, in Kuwait City,
DeGaetano and Allen (2002) examined long-term air temperature Kuwait, no significant changes in the urban warming and heat
trends across the United States during the twentieth century and stress were observed between 1962 and 1989, mainly due to the
found that urbanization exerts a strong influence on extreme city’s proximity to the sea, lower overall building heights and the
trends found in recent air temperature. For minimum air temper- use of local building materials (Balling and Nasrallah, 1991).
atures, the composite slope at urban stations is nearly three times Only a few studies address the relationship between global
greater than at rural stations. warming and urban warming and their effect on human comfort
Mean air humidity values are lower for urban areas than rural (Alcoforado and Andrade, 2008). Thorsson et al. (2011) predicted
ones; occasionally, the absolute humidity can be higher at night in that according to global warming scenarios, heat stress is expected
urban sites (Holmer and Eliasson, 1999; Kuttler et al., 2007; to intensify in northern cities like Göteborg, Sweden, with the
Landsberg, 1981). Therefore, the urban effect is likely to decrease occurrence of hot extremes (strong and extreme heat stress)
long-term air humidity trend values due to both air temperature increasing by 100e300% in summer. Lam et al. (2010) found that
increase and modifications in land surface; such modifications can summer discomfort in Hong Kong has increased from 1968 to 2008
evoke a rapid runoff of precipitation and decreased vegetation. The (while winter discomfort decreased) due to warmer weather in
overall result of such modifications is reduced evapotranspiration recent years.
(Henry et al., 1985). Jauregui and Tejed (1997) reported that the The study of long-term urban warming trends and their effect
mean annual values of relative humidity and specific humidity in on human discomfort and heat stress is an important issue.
Mexico City decreased rapidly between 1974 and 1992. Liu et al. Moreover, the implications of heat stress during the hottest hours
(2009) examined seasonal trends of air humidity in Beijing, China of the day are especially important in urban areas located in hot and
and suggested that long-term urban effects can decrease both arid zones, since these areas are the frontier for the extension and
relative humidity and vapor pressure in the urban area. establishment of new desert settlements.
In desert cities, long-term humidity trends may differ from This research examines the combined effect of global warming
those in temperate climate zones, generally because a desert city and urban warming on human thermal stress, as is demonstrated in
acts as an oasis in the bare and dry environment of the surrounding the rapidly growing desert city of Beer Sheva. The aims of the
desert (Golden, 2004). Therefore, an increase in long-term hu- present study were to:
midity trends is expected, due to the development of irrigated ur-
ban greenery (i.e., parks, lawns, street trees and private gardens). a) Examine the long-term trend in climatic changes in air tem-
However, an examination of long-term monthly averages of dew perature, humidity and wind velocity in the city of Beer Sheva
point levels in Phoenix, Arizona showed that in between 1896 and and in comparison to its surrounding desert environment.
1984, dew points were generally unchanged. Changes in local land b) Demonstrate an urban effect that is superimposed on an
use and a developing urban heat island (UHI) are suggested causes environment of climatic change.
of the observed patterns (Brazel and Bailing, 1986). Cicek and c) Study the accumulated effect of climatic changes on heat stress
Türkog lu (2009) found a decreasing trend in the long-term values due to the combined effect of regional warming and
annual values of water vapor pressure in Turkey’s semi-arid urban warming.
climate, due to urbanization. Shafir and Alpert (2011) found that
both temperature and relative humidity have increased over the 2. Description of sites and methodology
last four decades in the Israeli city of Eilat, which is located in an
extremely hot and arid climate. 2.1. Study area
Wind velocity within the urban canopy is usually lower in
comparison with rural winds at the same altitude. However, when The primary location for this case study was the city of Beer
regional winds are very light or calm, with a cloudless sky at night Sheva, Israel, located north of the Negev desert at 31
15E/34
50N.
(ideal conditions for urban heat island development), the horizontal Beer Sheva is ideal for UHI examination due to its location on a large
air temperature gradient across the urban/rural boundary is suffi- plateau and its rapid growth: the Beer Sheva population grew from
cient to induce a breeze from the rural area into the city (Oke, 1987). 45,000 inhabitants in 1965 to 200,000 inhabitants in 2007
Bornstein (1975) and Bornstein and Johanson (1977) investigated (Statistical Abstract of Israel, 2007). The Beer Sheva climate is hot
the effect of New York City on regional wind flow and demonstrated and dry, BWh according to KÖppen Classification (Potchter and
that wind velocity over New York City was below the values Saaroni, 1998). Average annual precipitation is 204 mm; average
observed outside the city in a non-urban environment, due to the daily air temperature during July, the hottest month, is 25.7
C
higher roughness in the city. Examination of wind velocity in New (range of 18.5e32.1
C); average summer relative humidity is 60%
York City’s Central Park from 1929 to 2006 showed a drop in velocity (range of 31e88%). In August, a slight decrease in air temperature
over the course of these years (Gaffin et al., 2008). The same wind and a more pronounced increase in air humidity takes place. These
pattern was found in São Paulo (Oliveira et al., 2003) and in Shanghai summer climatic conditions are known to create severe heat stress
(Shun Djen, 1992). The development of an urban heat island was during daytime hours and slight cold stress during nighttime hours.
 O. Potchter, H. Itzhak Ben-Shalom / Journal of Arid Environments 98 (2013) 113e122 115

Table 1 and converted to 1.5 m above ground according to the following

The intensity of the urban heat island of Bear Sheva according to size of the con- equation (Anon., 1983):
structed area, population and population density between the years 1964 and 2005.

UHI intensity Size of built Populationa Population density

up area (km2)a per square km Vh ¼ V10 ½0:233 þ 0:656*logðh þ 4:75Þ:


3 C during winter 1964 12,310 45,000 3655 Radiation data for Wadi Hatzerim has never been recorded, and
(Ganor, 1965) in Beer Sheva it has only been recorded since 1992. Therefore, due
5
C during winter 1975 17,799 90,000 5056
to the small distance between Beer Sheva and Wadi Hatzerim, we
(Skibin, 1979)
7
C during winter 2005 30,985 185,300 5980 assumed that radiation differences between both stations were
4
C during summer 2005 insignificant.
(Potchter et al., 2006) Trend analyses were carried out on the hourly data from each
a
Statistical Abstract of Israel (2007). station, for each month, by regressing values with time. The slopes
of the fitted serial linear regression lines were then tested for sig-
nificance. In order to demonstrate the urban effect on air temper-
The wind regime, characterized as a local one, is influenced by ature, specific humidity and wind velocity values, the differences
synoptic conditions, local topography and the diurnal landesea between the long-term trend values observed in Beer Sheva and
breeze. From early afternoon hours, when the sea breeze develops, those observed in Wadi Hatzerim were calculated and analyzed
wind flows from north to west and its velocity rapidly increases, according to fitted linear regression lines, and were then tested for
peaking in the afternoon. During the night, wind velocity gradually significance.
decreases and its direction changes from north to east. Minimum 2. Second, we calculated thermal stress according to two
values of 2e4 m/s have been observed between midnight and early indices: the Discomfort Index (DI) and the Physiological Equivalent
morning. From late morning hours, wind velocity slowly increases Air Temperature (PET). The DI belongs to the category of “direct
until midday, at which point this cycle repeats itself (Bitan and indices” (Epstein and Moran, 2006; Parsons, 2003) e indices based
Rubin, 1994). on the measurement of basic environmental variables of air tem-
Previous climate studies that investigated Beer Sheva’s local UHI perature and air humidity. Such indices are simple and easy to use
by means of mobile traverses (Ganor, 1965; Potchter et al., 2006; and therefore have been widely used to characterize regional
Skibin, 1979) presented a rapid UHI increase from 3
C up to 7
C thermal stress (McGregor, 1993) and the influences of climatic
during the winter season between the years 1964 and 2005. Mea- changes on thermal stress (Gaffen and Ross, 1998; Lam et al., 2010;
surements during the summer were taken only in 2005 (Table 1). Ziv and Saaroni, 2011; Wan et al., 2010). The Discomfort Index (DI)
has been adopted by many international institutions, such as
2.1.1. Description of sites meteorological services and defense forces (Baker et al., 2002;
The Beer Sheva meteorological station is located 280 m above Epstein and Moran, 2006; Kolokotsa and Karapidakis, 2009).
sea level at the city center, in a small open space (a few hundred The Discomfort Index (DI) was originally proposed by Thom
m2) surround by building blocks of four to seven floors (the nears (1959), based on two parameters: Dry Bulb Air temperature (DBT)
building to station is at a distant of 150 m). The station has been in and Wet Bulb Air temperature (WBT), and was slightly modified
use for the last 50 years. Since the 1950s, however, the area sur- and adapted to the Israeli climate by Sohar et al., (1963) as follows:
rounding the station once comprised of small, low-density, gar- DI ¼ 0.5 WBT þ 0.5 DBT.
denless buildings has become densely constructed, with high-rise The following criteria were established in order to define envi-
buildings, gardens, parks and mature trees (Fig. 1). This study ronmental heat stress: DI values below 22 units represented no
compares data from Beer Sheva’s urban meteorological station with encounter with heat stress; DI values between 22 and 24 units
the Wadi Hatzerim meteorological station, located in an open represented light heat stress; DI values between 24 and 28 units
desert environment five kilometers west of Beer Sheva and represented medium heat stress; and DI values above 28 units
195 m above sea level. This latter area has remained largely un- represented severe heat stress.
changed during the last four decades, apart from the growth of wild The DI index is limited by the fact that it does not consider cli-
desert shrubs. matic parameters such as solar radiation and wind velocity and
ignores human physiology. However, application of the DI index in
2.2. Methodology this study enables comparison of its results to other desert cities
that have been investigated by means of this index (i.e., the study of
In order to investigate the combined effect of global warming Phoenix, Arizona by Baker et al., 2002).
and urban warming on heat stress values, the following two The Physiological Equivalent Air Temperature index (PET) be-
methods were used: longs to the category of “rational indices” and “empirical indices”
1. We examined and statistically analyzed air temperature, hu- that are based on the integration of environmental and physio-
midity and wind velocity trends in Beer Sheva and Wadi Hatzerim logical variables (Epstein and Moran, 2006). The PET, a universal
during summer months (MayeSeptember). The study includes data index used for characterizing thermal bioclimatic levels, is used
from the years 1967 to 2007, since data from Wadi Hatzerim only worldwide by urban and bio-climatologists (Cohen et al., 2013;
became available in 1967. Hourly temperature, relative humidity Hoppe, 1999; Lin and Matzarakis, 2008; Matzarakis and Mayer,
and wind velocity measurements were obtained from the two Is- 1996; Matzarakis et al., 2010; Thorsson et al., 2004; Ali-Toudert
raeli Meteorological Service stations, and average hourly degrees and Mayer, 2007).
for each month were calculated. The wet temperature bulb was Obtained from human energy balance, the PET is measured in
obtain from the metrological stations of Wadi Hatzerim and was
Celsius, bearing the advantage of making test results more

calculated for Beer Sheva. In general, since relative humidity values comprehensible. It also allows for the evaluation of thermal con-
represent the relationship between humidity and air temperature, ditions in a physiologically significant manner (Hoppe, 1999;
it is necessary to calculate the values of the specific humidity in Mayer and Hoppe, 1987). Originally applicable to Western/Mid-
order to estimate their influence on human comfort. Wind data was dle Europe, the PET has been adjusted in recent years to other
taken from meteorological masts at heights of 10 m above ground climatic zones. Cohen et al. (2013) have adapted the PET scale to
116 O. Potchter, H. Itzhak Ben-Shalom / Journal of Arid Environments 98 (2013) 113e122

Fig. 1. Map of KÖppen Climatic Classification of Israel and the location of the study area with aerial view of the meteorological station area of Beer Sheva in 1958 and 2006.

Israel as follows: measures between 15 and 19
C are defined for a sitting person; the insulation factor of clothing (Icl) has been
as slightly cool, 19e26
C is neutral, 26e28
C is slightly warm, standardized to 0.5 for light summer clothing (Jendritzky et al.,
28e34
C is warm, 34e40
C is hot and above 40
C is very hot. 1990).
The advantages of the PET index lie in its consideration of all As discussed above, the meteorological data from the city of
environmental variables such as radiation, air temperature, air Beer Sheva used in this research were obtained from a station sit-
humidity, wind velocity, topographical elevation and human uated in an urban open space and represent the surrounding urban
physiology. Its limitations are its complexity of calculation. effect. Therefore, the results presented in this paper only apply to
PET was calculated using the PC modeling program, RayMan exposed open areas within the urban area (such as a plaza, square
(Matzarakis et al., 2000). The RayMan model, developed according or courtyard). It should be noted that these results are not appli-
to Guideline 3787 of the German Engineering Society (VDI, 1998), cable to a specific complex urban environment, since no urban
calculates the radiation flux in simple and complex environments effects of the radiant exchange between a pedestrian and the
according to the StefaneBoltzmann Law. Radiant exchange is gov- micro-urban environment were calculated.
erned solely by surface temperatures and emissivities. The model
considers air temperature, air humidity and wind velocity, as well 3. Results
as human clothing and activity. Mean radiant temperature, which is
the most important factor in calculating PET scores, can be esti- 3.1. An examination of climatic variables between 1967 and 2007
mated by importing the observed value of global radiation or by
importing date of year, time, elevation, location and cloud cover. Climactic variables between 1967 and 2007 were investigated
The RayMan model’s simulation of radiation flux closely resembles for the months of May, June, July, August and September, which
the respective measured data (Lin et al., 2010; Matzarakis et al., together represent the hot season. Statistical analysis of air tem-
2010). perature, specific humidity and wind velocity trends in Beer Sheva
According to the model, the calculation of thermal sensation and Wadi Hatzerim revealed the following (Table 2a): all summer
requires adjustment of the following constants: body surface has months showed an increasing trend of air temperature values; the
been standardized to 1.9 m2, which represents a human with a trend in urban Beer Sheva was more significant and pronounced
height of 1.75 m and a bodyweight of 75 kg (Hoppe, 1984); the than its corresponding one in the rural area of Wadi Hatzerim; the
metabolic rate (Met) has been fixed at an average value of 65 W/m2 greatest temperature increase and the most significant (p < 0.05)
 O. Potchter, H. Itzhak Ben-Shalom / Journal of Arid Environments 98 (2013) 113e122 117

Table 2
The rates of linear slope and significant results of long-term trend between the Beer Sheva (urban) and Wadi Hatzerim (rural) meteorological stations during the years 1967e
2007 for: (a) air temperature (TD), specific humidity (SH) and wind velocity (WV) and (b) Values of long-term Discomfort Index (DI) and the PET Index trends. Statistically
significant slopes at 95% confidence level p < 0.05 are in bold.

Hour May June July August September

Slope p-value Slope p-value Slope p-value Slope p-value Slope p-value

(a)
Dry temperature (
C) Urban 06:00 0.038 0.016 0.040 0.000 0.064 0.000 0.072 0.000 0.055 0.000
Rural 06:00 0.023 0.098 0.029 0.001 0.029 0.001 0.047 0.000 0.027 0.006
Du-r 06:00 0.011 0.054 0.011 0.029 0.034 0.000 0.025 0.006 0.028 0.000
Urban 15:00 0.034 0.069 0.038 0.000 0.063 0.000 0.062 0.000 0.055 0.000
Rural 15:00 0.017 0.292 0.011 0.247 0.040 0.000 0.037 0.000 0.028 0.080
Du-r 15:00 0.002 0.481 0.027 0.000 0.023 0.000 0.025 0.000 0.026 0.047
Specific humidity (g/kg) Urban 06:00 0.027 0.013 0.036 0.000 0.045 0.002 0.071 0.000 0.027 0.029
Rural 06:00 0.018 0.048 0.030 0.000 0.023 0.022 0.068 0.000 0.022 0.058
Du-r 06:00 0.011 0.212 0.006 0.319 0.022 0.007 0.003 0.801 0.009 0.368
Urban 15:00 0.044 0.007 0.058 0.000 0.032 0.001 0.070 0.000 0.034 0.006
Rural 15:00 0.011 0.315 0.024 0.032 0.004 0.635 0.034 0.008 0.019 0.161
Du-r 15:00 0.035 0.618 0.034 0.011 0.028 0.001 0.035 0.001 0.056 0.001
Wind velocity (m/s) Urban 06:00 0.001 0.840 0.007 0.131 0.016 0.000 0.011 0.006 0.010 0.033
Rural 06:00 0.010 0.099 0.013 0.026 0.027 0.000 0.025 0.000 0.026 0.000
Du-r 06:00 0.009 0.271 0.006 0.446 0.011 0.035 0.014 0.015 0.015 0.004
Urban 15:00 0.027 0.030 0.025 0.004 0.028 0.000 0.025 0.000 0.017 0.018
Rural 15:00 0.010 0.170 0.009 0.138 0.008 0.893 0.008 0.141 0.006 0.199
Du-r 15:00 0.040 0.002 0.033 0.002 0.027 0.003 0.033 0.000 0.022 0.007
(b)
DI Urban 06:00 0.039 0.016 0.040 0.000 0.066 0.000 0.073 0.000 0.057 0.000
Rural 06:00 0.023 0.046 0.024 0.001 0.040 0.000 0.002 0.868 0.016 0.233
Du-r 06:00 0.022 0.351 0.016 0.002 0.024 0.001 0.070 0.000 0.073 0.000
Urban 15:00 0.022 0.209 0.038 0.000 0.054 0.000 0.070 0.000 0.047 0.000
Rural 15:00 0.023 0.040 0.023 0.003 0.032 0.000 0.052 0.000 0.037 0.000
Du-r 15:00 0.008 0.223 0.008 0.092 0.022 0.000 0.023 0.000 0.010 0.016
PET Urban 06:00 0.034 0.215 0.009 0.721 0.016 0.474 0.042 0.078 0.036 0.057
Rural 06:00 0.003 0.831 0.001 0.914 0.029 0.068 0.047 0.680 0.010 0.479
Du-r 06:00 0.230 0.004 0.010 0.715 0.047 0.021 0.005 0.813 0.026 0.162
Urban 15:00 0.090 0.048 0.084 0.000 0.120 0.000 0.116 0.000 0.096 0.000
Rural 15:00 0.015 0.562 0.008 0.590 0.058 0.000 0.061 0.001 0.035 0.046
Du-r 15:00 0.073 0.022 0.076 0.000 0.062 0.000 0.055 0.000 0.059 0.010

trend were found in the months of July and August. The long-term was 0.064 and R2 ¼ 0.56, while Wadi Hatzerim’s linear slope was
specific humidity changes showed increasing trend values during only 0.029 and R2 ¼ 0.29. During midday (15:00), when air tem-
the summer, and the long-term trend in the city of Beer Sheva was perature peaked, the linear slope in the Beer Sheva was 0.063 and
more intense than the trend in the rural area of Wadi Hatzerim. The R2 ¼ 0.449, while Wadi Hatzerim’s linear slope was 0.04 and
greatest humidity increase and the most significant (p < 0.05) trend R2 ¼ 0.29 (Fig. 2a). All results were significant (Table 2a).
was in the month of August. Wind velocity showed a different and In order to examine urban warming against a background of
more complicate trend. During late night and morning hours, the environmental warming, differences in long-term air temperature
wind long-term velocity trend increased; the maximum increase between the city of Beer Sheva and the rural area of Wadi Hatzerim
was in July, and the trend was more pronounced in Wadi Hatzerim. (DTu-r) were calculated for the daily times of maximum and mini-
During afternoon hours, long-term wind velocity significantly mum air temperature (06:00 and 15:00), in the month of July. In
decreased (p < 0.05) in the city of Beer Sheva, most pronounced in general, during early morning hours, the long-term trend increase
the month of July; no clear trend was observed in Wadi Hatzerim. was more pronounced (slope ¼ 0.0348 and R2 ¼ 0.429 at 06:00)
The following subsection focuses on hourly trend changes of air than in the afternoon (slope ¼ 0.0234 and R2 ¼ 0.3074 at 15:00).
temperature, specific humidity and wind velocity at 15:00 (the time Both results were significant (p ¼ 0.00) (Fig. 2b and Table 2a).
of maximum temperature) and 06:00 (the time of minimum tem-
perature) during the month of July e the hottest month. This month 3.1.2. Long-term specific humidity trend
reflects climatic trends similar but more pronounced than those An examination of long-term hourly-specific humidity changes
observed during the summer period as a whole. showed the following: during the early morning when air tem-
peratures were at the minimum, the long-term specific humidity
3.1.1. Long-term hourly air temperature trend increased at both stations. The increase in Beer Sheva was greater
Since Beer Sheva is located 85 m above Wadi Hatzerim, the and more significant (p < 0.05) than in rural Wadi Hatzerim
minimum air temperatures at the 60`th in Beer Sheva were lower (slope ¼ 0.045, R2 ¼ 0.227; slope ¼ 0.023, R2 ¼ 0.127, respectively),
than in Wadi Hatzerim, but at the 90`th, temperatures in the two as well as greater during midday than during the early morning.
sites corresponded, due to the urban warming of the city. Exami- During midday at 15:00, when air temperatures peaked, Beer
nation of the long-term hourly air temperature trend in the city of Sheva’s humidity showed an increased linear slope of 0.032,
Beer Sheva and its rural surroundings of Wadi Hatzerim empha- R2 ¼ 0.252, while in Wadi Hatzerim, the long-term changes were
sizes the fact that the increased temperature trend in urban area of negligible (Slope ¼ 0.004, R2 ¼ 0.005) (Table 2a and Fig. 3a).
Beer Sheva was more significant (p < 0.05) than in rural area of In order to examine the urban effect against a background of
Wadi Hatzerim. The linear slope at 06:00 in the city of Beer Sheva climatic changes in specific humidity, differences in long-term
118 O. Potchter, H. Itzhak Ben-Shalom / Journal of Arid Environments 98 (2013) 113e122

(a) (a) 19
38 y = 0.0453x + 14.005 y = 0.0232x + 14.844
y = 0.0635x + 31.638 y = 0.04x + 31.014 R = 0.2279 R = 0.1278
36 17
R = 0.4498 R = 0.2947

Specific Humidity
34

g/kg (dry air)

Temprature c

32 15
o

30
28 13
26
y = 0.0644x + 19.165 y = 0.0296x + 20.167
24 R = 0.5615 R = 0.2492 11
22 y = 0.0326x + 11.216 y = 0.0041x + 12.955
R = 0.252 R = 0.0058
20 9
18

1967

1971

1975

1979

1983

1987

1991

1995

1999

2003

2007
1967

1971

1975

1979

1983

1987

1991

1995

1999

2003

2007
Year
Year
Beer Sheva 15:00 Wadi Hatzerim 15:00
Beer Sheva 15:00 Wadi Hatzerim 15:00 Beer Sheva 06:00 Wadi Hatzerim 06:00
Beer Sheva 06:00 Wadi Hatzerim 06:00
(b)
(b) 2
y = 0.0286x - 1.7393 y = 0.0221x - 0.8387
3 R2 = 0.2464 R2 = 0.174

ΔSpecific Humidity u-r

y = 0.0234x + 0.6243 y = 0.0348x - 1.0014
2

g/kg (dry air)

R = 0.3074 R = 0.429
0
1
ΔTemprature

0 -2
c
o

-1
-4
-2

1967

1971

1975

1979

1983

1987

1991

1995

1999

2003

2007
-3
Year
1967

1971

1975

1979

1983

1987

1991

1995

1999

2003

2007

15:00 06:00
Year

15:00 06:00 Fig. 3. Long-term specific humidity trends, 1967e2007, in the month of July at 06:00
and 15:00. (a) In Beer Sheva and Wadi Hatzerim e (b) specific humidity differences
Fig. 2. Long-term air temperature trends, 1967e2007, in the month of July at 06:00 (DSHu-r) trend between Beer Sheva and Wadi Hatzerim.
and 15:00. (a) In Beer Sheva and Wadi Hatzerim e (b) Air temperature differences
(DTu-r) trend between Beer Sheva and Wadi Hatzerim.
3.2. Long-term human discomfort trend

specific humidity between the city of Beer Sheva and the rural area An examination of the long-term human discomfort trend be-
of Wadi Hatzerim (DSHu-r) were calculated in the early morning tween the years 1967 and 2007 was carried out for the summer
(06:00) and in the afternoon (15:00), in the month of July. The long- months of May, June, July, August and September. Statistical anal-
term increase was more pronounced and significant during the ysis of long-term human discomfort trends in Beer Sheva and Wadi
afternoon (slope ¼ 0.028 and R2 ¼ 0.246 at 15:00) than in the early Hatzerim applying the thermal stress indices DI and PET revealed a
morning (Slope ¼ 0.022 and R2 ¼ 0.174 at 06:00) (Table 2a and rise in the long-term human discomfort trend for both sites during
Fig. 3b). all summer months. However, the increasing human discomfort in
urban Beer Sheva was more pronounced than in rural Wadi
3.1.3. Long-term wind velocity trend Hatzerim.
An examination of the long-term hourly wind velocity trend
showed a complex diurnal pattern. During early morning hours, 3.2.1. Long-term DI trends
when wind velocity values were lower than 2 m/s, wind velocity The DI index (Table 3) shows that between 1967 and 1976,
increased at both stations, though the increase was more pro- 14 hours of heat stress were observed in Wadi Hatzerim and only
nounced in Wadi Hatzerim than in Beer Sheva (slope ¼ 0.0278, 13 hours in Beer Sheva, for two reasons. First, the topographical
R2 ¼ 0.46; slope ¼ 0.0165, R2 ¼ 0.307 respectively, at 06:00). All height of Beer Sheva is 100 m above Wadi Hatzerim’s, so the
results presented significant changes (p < 0.05). During daytime temperature in Beer Sheva was lower; and second, the population
hours, when wind velocity rose, the long-term changes were of the city in these years was small and the urban effect was very
negligible in Wadi Hatzerim, while in Beer Sheva, long-term limited. During the years 1997e2006, the number of heat stress
wind velocity decreased. This phenomenon was most pro- hours in both Wadi Hatzerim and Beer Sheva increased to 16
nounced at 15:00 (Wadi Hatzerim: slope ¼ 0.008, R2 ¼ 0.0005; (Wadi Hatzerim: þ2; Beer Sheva: þ3). An examination of heat
Beer Sheva: slope ¼ 0.028, R2 ¼ 0.29) (Table 2a and Fig. 4a). stress intensity showed that during the years 1967e1976, seven
These daily patterns of wind velocity can explain the long-term hours of light heat stress and seven hours of medium heat stress
trend of wind velocity differences (DWVu-r) between urban were observed in Wadi Hatzerim, while in Beer Sheva, four hours
Beer Sheva and rural Wadi Hatzerim. During the afternoon, of light heat stress and nine hours of medium heat stress were
the long-term decrease trend of wind velocity differences observed. During the years 1997e2006, the number of light heat
(DWVu-r) is more pronounced and significant (slope ¼ 0.0276, stress hours in Wadi Hatzerim decreased to five from seven (2),
R2 ¼ 0.209 and p ¼ 0.003 at 15:00) than during early morning while the number of medium heat stress hours increased to 11
(slope ¼ 0.011, R2 ¼ 0.109 and p ¼ 0.035 at 06:00) (Table 2a and (þ4). A comparison between the two areas of heat stress intensity
Fig. 4b). levels during the years 1997e2006 revealed the following: the
 O. Potchter, H. Itzhak Ben-Shalom / Journal of Arid Environments 98 (2013) 113e122 119

3
(a) 6 y = 0.0224x + 0.0531 y = 0.0242x - 1.3498
y = -0.0284x + 3.6663 y = -0.0008x + 4.1805
5 R = 0.2966 R = 0.0005 2 2
R = 0.3068
2
R = 0.2601
Wind Velocity (m/s)

4 1
3

ΔDIu-r
0
2
-1
1
0 y = 0.0165x + 0.1933 y = 0.0278x + 0.6904 -2
R = 0.3073 R = 0.4675
-1 -3
1967

1971

1975

1979

1983

1987

1991

1995

1999

2003

2007

1967

1971

1975

1979

1983

1987

1991

1995

1999

2003

2007
Year
Year
Beer Sheva 15:00 Wadi Hatzerim15:00
Beer Sheva 06:00 Wadi Hatzerim 06:00 15:00 06:00

(b) 3 Fig. 5. Long-term DI index differences (DDIu-r) trend between Beer Sheva and Wadi
Hatzerim e 1967e2007 in the month of July at 06:00 and 15:00.
y = -0.0276x - 0.5142 y = -0.0113x - 0.4971
2 R2 = 0.2098 R2 = 0.1095
Δ Wind Velocityu-r

1
number of light heat stress hours in Beer Sheva remained identical,
(m/s)

0 the number of medium heat stress hour increased to ten (þ1), and
two hours of severe heat stress (not previously observed between
-1
1967 and 1976) occurred.
-2 In order to examine the urban effect against a context of
increasing heat stress, long-term DI differences (DDIu-r) between
-3
urban Beer Sheva and rural Wadi Hatzerim were calculated. The
1967

1979

1995

2003
1971

1975

1983

1987

1991

1999

2007

results showed growing long-term difference during early morning

Year and afternoon hours in the month of July (Table 2b and Fig. 5).
15:00 06:00
Calculation of the long-term hourly DI differences (DDIu-r) uncov-
ered that in general, during early morning hours, the increase was
Fig. 4. Long-term wind velocity trends, 1967e2007, in the month of July at 06:00 and slightly more pronounced (slope ¼ 0.024; R2 ¼ 0.26 and p ¼ 0.001
15:00. (a) In Beer Sheva and Wadi Hatzerim e (b) Wind velocity differences (DWVu-r) at 06:00) than during the afternoon (slope ¼ 0.0224; R2 ¼ 0.3068
trend between Beer Sheva and Wadi Hatzerim.
and p ¼ 0.00 at 15:00).

3.2.2. Long-term PET trends

An examination of the PET index for the month of July during
Table 3 the years 1967e2006 showed that from the decade of 1967e1976
A comparison of the DI index between Beer Sheva (on the right) and Wadi Hatzerim to the decade of 1997e2006, the following changes occurred.
(on the left), in decades from 1967 to 2006, in the month of July.
(Table 4):
Wadi Hatzerim (Rural station) Beer Sheva (Urban station)
a) The “slightly cool” diurnal level duration in Wadi Hatzerim
1967- 1977- 1987- 1997- 1967- 1977- 1987- 1997-
Hour Hour decreased by three hours, while in Beer Sheva it decreased by
1976 1986 1996 2006 1976 1986 1996 2006
00:00 20.9 21.2 21.6 21.9 00:00 19.9 20.6 21.3 21.9 four hours.
01:00 20.3 20.8 21.3 21.8 01:00 19.4 20.1 20.7 21.2 b) The “neutral” diurnal level duration in Wadi Hatzerim
02:00 20.1 20.6 21.1 21.7 02:00 18.8 19.6 20.4 21.0
03:00 20.1 20.4 20.8 21.3 03:00 18.3 19.1 20.1 20.5 increased by three hours, while in Beer Sheva it increased by
04:00 19.8 20.2 20.6 21.1 04:00 18.2 19.0 20.0 20.3 two hours.
05:00 19.5 19.9 20.4 20.9 05:00 18.1 18.8 19.9 20.2 c) The total daily hours defined as slightly warm, warm, hot and
06:00 19.4 19.7 20.1 20.7 06:00 18.0 18.7 19.7 20.0
07:00 20.4 20.7 20.9 21.6 07:00 19.4 19.9 20.7 21.1 very hot in Wadi Hatzerim did not change, while in Beer Sheva
08:00 22.4 22.6 23.0 23.5 08:00 20.8 21.2 21.9 22.3 it increased by two hours.
09:00 23.2 23.5 23.8 24.4 09:00 22.2 22.4 23.0 23.5 d) The “hot” diurnal level duration in Wadi Hatzerim increased by
10:00 23.9 24.3 24.7 25.3 10:00 23.1 23.3 23.9 24.5
11:00 24.8 25.2 25.6 26.2 11:00 24.1 24.2 24.8 25.5 three hours, while in Beer Sheva it increased by two hours;
12:00 25.0 25.7 26.0 26.5 12:00 24.9 25.2 25.7 26.5 however, only urban Beer Sheva experienced the “very hot”
13:00 25.8 26.1 26.3 26.8 13:00 25.5 25.6 26.0 27.0 level, for two hours a day.
14:00 26.0 26.5 26.8 27.2 14:00 25.9 26.1 26.5 27.6
15:00 25.6 26.0 26.3 26.8 15:00 26.4 26.7 27.1 28.2
16:00 25.1 25.5 25.9 26.3 16:00 25.9 26.1 26.6 28.0 The intensity of the PET degree in the same stress level
17:00 24.7 25.0 25.4 25.9 17:00 25.4 25.6 26.6 27.3 increased in Beer Sheva more rapidly than in Wadi Hatzerim. For
18:00 23.9 24.3 24.6 25.0 18:00 24.9 25.0 25.7 26.8
19:00 23.3 23.6 23.9 24.4 19:00 24.0 24.3 24.8 25.8
example, at 15:00, the PET degree in Wadi Hatzerim increased by
20:00 22.6 22.8 23.1 23.6 20:00 23.2 23.5 24.0 24.9 2.1
C, while in Beer Sheva it increased by 4
C.
21:00 22.2 22.4 22.7 23.2 21:00 22.4 22.8 23.1 24.1 In order to examine the urban warming effect against a context
22:00 21.5 22.0 22.3 22.8 22:00 21.6 22.0 22.5 23.4
23:00 21.3 21.6 22.0 22.4 23:00 20.8 21.3 21.9 22.7
of an increasing PET value, the long-term trend of PET differences
between Beer Sheva and Wadi Hatzerim (DPETu-r) were calculated.
These differences increased during most of July (Table 2b and
Heat Stress
<22 22-24 24-28 28> Fig. 6). Long-term hourly PET differences (DPETu-r) revealed that
Medium heat Severe heat
Level No heat stress Light heat stress
stress stress during midday hours, the increase was more pronounced and sig-
nificant (slope ¼ 0.0627; R2 ¼ 0.3531 and p ¼ 0.00 at 15:00) than in
120 O. Potchter, H. Itzhak Ben-Shalom / Journal of Arid Environments 98 (2013) 113e122

Table 4 that background environmental warming and urban warming are

A comparison of the PET index between Beer Sheva (on the right) and Wadi Hat- likely to amplify heat stress values. Since 1958, Beer Sheva’s pop-
zerim (on the left), according to decades from 1967 to 2006, in the month of July.
ulation has tripled, causing significant changes in air temperature,
Wadi Hatzerim (Rural station) Beer Sheva (Urban station)
humidity and wind velocity in this still-growing urban area.
1967- 1977- 1987- 1997-
Hour
1967- 1977- 1987- 1997- The first stage in the study was to demonstrate an urban effect
Hour
1976 1986 1996 2006 1976 1986 1996 2006 that is superimposed on a background environment climatic
00:00 17.9 18.1 18.1 18.3 00:00 17.3 18.5 19.0 18.9
01:00 17.6 17.5 17.5 17.7 01:00 16.9 18.0 18.6 18.5 warming. It appears that the rapid growth of the desert city of Beer
02:00 17.3 17.1 17.0 17.1 02:00 16.5 17.6 18.2 17.9 Sheva has caused significant climatic changes in the urban area
03:00 16.5 16.4 16.6 16.8 03:00 16.1 17.2 17.8 17.5
during the summer months between the years 1967 and 2007. We
04:00 17.5 17.2 17.2 17.6 04:00 17.4 18.5 19.3 19.0
05:00 18.2 18.1 18.1 18.2 05:00 18.5 19.8 20.7 20.0 have exposed the significantly greater increase in air temperature
06:00 19.9 19.2 19.5 19.1 06:00 20.1 21.6 22.2 20.7 and specific humidity that took place in urban Beer Sheva relative
07:00 23.9 23.5 23.8 23.6 07:00 25.4 26.9 27.6 26.0
08:00 28.9 28.1 28.4 28.5 08:00 30.7 33.2 34.3 29.3
to that of nearby rural Wadi Hatzerim. There has also been a
09:00 33.82 32.7 32.6 33.0 09:00 35.9 37.6 39.7 35.0 decrease of wind velocity in the urban area compared to the sur-
10:00 34.3 33.5 33.7 34.6 10:00 36.0 37.4 39.5 35.5 rounding rural desert area in the afternoon hours, while during
11:00 34.9 34.6 34.8 36 11:00 36.3 37.5 39.3 36.5
12:00 35.4 35.5 36.0 37.4 12:00 36.6 37.7 39.1 39.0 early morning hours the wind velocity increase was higher in the
13:00 35.7 36.1 36.5 37.7 13:00 37.0 38.1 39.5 39.9 urban area compared to the rural area.
14:00 35.9 36.8 37.0 38 14:00 37.4 38.5 39.9 40.3 The long-term diurnal air temperature trend of the city of Beer
15:00 36.2 37.2 37.2 38.3 15:00 37.9 39.0 40.5 41.9
16:00 33.0 34.3 34.1 35.2 16:00 32.4 35 36.6 38.2 Sheva showed a grater increase in comparison to rural Wadi Hat-
17:00 30.0 31.2 31.3 32.1 17:00 29.0 31.7 33 34.8 zerim demonstrated an increased slope during early morning and
18:00 26.9 27.5 27.9 28.8 18:00 27.6 28.3 29.7 31.2
19:00
midday hours. This rise is likely due to the development of a
19:00 24.5 24.1 25 25.5 24.7 24.5 26.0 27.0
20:00 21.0 21.0 22.0 22.0 20:00 21.8 21.5 23.0 24.0 nighttime and daytime urban heat island (identified in previous
21:00 18.9 19.1 19.4 20.0 21:00 18.8 19.5 19.9 20.9 studies by Potchter et al., 2006). This urban effect on the slope of
22:00 18.6 18.7 19.0 19.5 22:00 18.2 19.0 19.6 20.2
23:00 18.3 18.4 18.6 19.0 23:00 17.8 18.5 19.4 19.5
the air temperature increase supports the results of DeGaetano and
Allen (2002) as well as those of Fujibe (2009), who found occasional
widespread temperature increases during nighttime hours and
15-19 19.1-26 26.1-28 28.1-34 34.1-40 40.1>
occasional widespread warming during daytime hours in the
Stress Level
Slightly cool Neutral
Slightly
Warm Hot Very hot summer season in urban areas. An examination of the air temper-
warm
ature difference (DTu-r) of long-term changes between Beer Sheva
and Wadi Hatzerim demonstrates the superimposed urban
warming in comparison to the rural environment.
the early morning (slope ¼ 0.047; R2 ¼ 0.128 and p ¼ 0.021 at The increase in specific humidity found in the rural area sup-
06:00). ports previous studies that found that long-term specific humidity
increased significantly worldwide during the last century (Gaffen
A comparison between the measures obtained through the DI
and PET indices (Table 2b) demonstrated that the linear slope of the and Ross, 1998), while warmer regions exhibit larger increases in
specific humidity for a given air temperature change (Willett et al.,
PET is higher than that of the DI index during afternoon hours, and
all the results were significant. During the early morning, the linear 2008). Moreover, the urban effect is likely to decrease long-term
humidity values (Henry et al., 1985; Jauregui and Tejed, 1997; Um
slope of the DI is higher than that of the PET. At these hours, the
linear slope of DI was significant, while the linear slope of the PET et al., 2007; Liu et al., 2009). Although Golden (2004) suggests
that in desert cities one can expect an increase in long-term hu-
was not.
midity trends due to the development of irrigated urban greenery,
evaluations of long-term humidity trends in arid and semi-arid
4. Discussion
cities, such as Phoenix, Arizona (Brazel and Bailing, 1986) and
Turkey (Cicek and Türkog lu, 2009) do not support this argument.
The aim of this study was to examine the combined effect of
Yet, the case of Beer Sheva does support Golden’s statement. The
global warming and urban warming on human thermal stress, as is
present study found a significantly increased long-term trend in
demonstrated in a rapidly growing desert city. An investigation of
specific humidity that was more pronounced than that found in the
climate trends and changes over the last several decades indicates
surrounding rural environment, apparently due to the substantial
development of irrigated urban parks, lawns, street trees and pri-
7 vate gardens over the past 40 years.
In general, urbanization has a moderating effect on wind ve-
5 locity. This effect is well represented in the city of Beer Sheva and its
outlying area, Wadi Hatzerim. During daytime hours, wind velocity
ΔPETu-r C

3 showed a significant long-term decrease, while in the rural envi-

o

ronment, wind velocity changes were inconsistent and insignifi-

1 cant. During nighttime hours, on the other hand, under calm
conditions or slow wind flow, wind velocity in both Beer Sheva and
-1 y = 0.0627x + 1.2948 y = 0.0476x + 0.6376
Wadi Hatzerim slightly increased over the long-term. Ultimately,
R2 = 0.3531 R2 = 0.1286
however, since wind velocity was very low, its effect on human
-3
comfort during these hours was quite limited.
1967

1971

1975

1979

1983

1987

1991

1995

1999

2003

2007

In the study’s last stage, we examined the accumulated effect of

Year urban warming and general climatic changes regarding air tem-
perature, humidity and wind velocity on human thermal stress
15:00 06:00
values. A comparison of heat stress trends between urban and rural
Fig. 6. Long-term PET index differences (DPETu-r) trend between Beer Sheva and Wadi desert areas showed that the increase of heat stress in the urban
Hatzerim e 1967e2007 in the month of July at 06:00 and 15:00. area is much more significant than that of the rural environment.
 O. Potchter, H. Itzhak Ben-Shalom / Journal of Arid Environments 98 (2013) 113e122 121

Climactic changes possibly caused by global warming in the rural  In the case of a desert city irrigation of urban vegetation can
area of Wadi Hatzerim e including air temperature increase and cause in the long-term significantly increased specific humid-
specific humidity increase e have had a negative impact on human ity, which aggravates heat stress values.
comfort by increasing the duration and intensity of heat stress. In  Universal or empirical indices such as PET, rather than rational
the urban area, these changes were accompanied by wind reduc- heat stress indices such as DI, are most appropriate for evalu-
tion and, as a result of the urban heat island, additional air tem- ating human thermal stress in desert cities that experience
perature increases. Such changes induced greater heat stress for a both heat stress and cold stress in the same day.
longer duration in comparison to the rural area.  Although currently according to PET index it seems that Beer
Although the two bio-meteorological indices applied in this Sheva has experienced an increase in nighttime human ther-
study showed parallel tendencies of increased heat stress values mal comfort hours, in view of predicted global warming and
and duration, significant differences emerged in their results: The future population growth, the total number of comfort hours
DI, based on dry bulb and wet bulb air temperatures, showed more will decrease and the intensity of heat stress is likely to
pronounced long-term increases in value during the late nighttime increase.
and dawn (although it did not reach the degree defined as heat
stress). This trend probably results from urban warming and the The results suggest that planners and architects should pay
increase in air humidity during these hours, when an urban heat special attention to climatic changes that can be caused by urban
island is developed (Oke, 1987). However, since the DI describes development. Moreover, efforts should be made to create sustain-
only heat stress values, its trends are limited to daytime hours. The able green areas that demand less water consumption and are
PET index showed a more pronounced long-term trend increase adapted to arid climates thus moderating the potential increase in
during daytime hours than during nighttime hours. This difference air humidity. It is also recommended that wind corridors be
can be explained by the fact the PET also takes wind velocity into designed to ventilate cities and that wind barriers be avoided.
account. Since the effect of urban structure on wind velocity It should be notice that the results presented in this paper only
moderation is more intense at midday and in the afternoon, the apply to open spaces with no urban effects upon the radiant ex-
increase in the urban effect on heat stress levels is dominated to a change between a pedestrian and the environment. Future study is
greater extent during these hours by the PET than by the DI. During needed in order to apply these results to the specific urban tissue of
early morning hours, wind e which slightly increased at both sta- the city of Beer Sheva.
tions over the years e moderates the combined effect of air tem-
perature and air humidity increases, and therefore the PET long-
term trend was lower and less significant than that of the DI, Acknowledgments
which did not consider the wind effect.
The PET index also calculates cold stress and neutral climate The authors wish to thank to Mr. Ido Mintz and Ms. Ilana
sensation, as well as demonstrating climate changes during night- Gelernter for their statistical analysis of data presented in this pa-
time hours. Data shows a reduction in the total number of hours per. The authors are also grateful to Mr. Yaron Yaakov for his
defined as “slightly cool” and an increase in the number of hours technical assistance and to Professor Haim Tzoar for his fruitful
defines as “neutral,” since nighttime temperature increased over the advice and remarks.
last 40 years. Such changes are more pronounced in Beer Sheva than
in rural Wadi Hatzerim. Although currently it seems that despite References
marked changes in climate, nighttime human thermal comfort
hours have increased, in view of the prospective warming trend, the Alcoforado, M.J., Andrade, H., 2008. Global warming and the urban heat island. In:
total number of comfort hours is likely to decrease in the future. Marzluff, J.M., Shulenberger, E., Endlicher, W., Alberti, M., Bradley, G., Ryan, C.,
ZumBrunnen, C., Simon, U. (Eds.), Urban Ecology: an International Perspective
The study also sought to identify which of the two types of
on the Interaction Between Humans and Nature. Springer, New York, New York,
indices, a “direct index” such as DI or an “empirical index” such as USA, pp. 249e262.
the universal PET index, is more suitable to describe the thermal Ali-Toudert, F., Mayer, H., 2007. Thermal comfort in an east–west oriented street
canyon in Freiburg (Germany) under hot summer conditions. Theor. Appl. Cli-
sensation in an urban area. It seems that in the case of a desert city
matol. 87, 223e237.
that experiences both heat stress and cold stress in the same day, Alpert, P., 2004. The water crisis in the Eastern Mediterranean e and relation to
using the PET index is advantageous, as it calculates both. More- global warming? In: Zereini, F.W., Jaeschke (Eds.), Water in the Middle-East and
over, in the case of an urban environment, the long-term wind North Africa. Springer, pp. 55e61.
Anon., 1983. Observers’ Handbook. Meteorological Office, HMSO, U.K.
velocity trend is affected by the growing constructed area, and Baker, L.A., Brazel, S.W., Selover, N., Martin, C., McIntyre, N., Steiner, F.R., Nelson, A.,
these effects can not be calculated using the DI. Therefore, it seems Musacchio, L., 2002. Urbanization and warming in Phoenix (Arizona, USA):
that for these kinds of urban studies in desert regions, the PET index impacts, feedbacks and mitigation. Urban Ecosyst. 6, 183e206.
Balling, R.C., Cerveny, R.S., 1987. Long-term associations between wind speeds and
might be more useful. the urban heat island of Phoenix, Arizona. J. Appl. Meteorol. 26, 712e716.
Ultimately, this type of thermal stress analysis is valid for Balling, J.R., Nasrallah, H.A., 1991. The impact of rapid urbanization on summer
exposed urban areas and can provide useful guidelines for adjust- whether stress in KUWAIT city. J. Arid Environ. 21, 261e265.
Ben Gai, T., Bitan, A., Manes, A., Alpert, P., Rubin, S., 1999. Temporal and spatial
ment to the specific urban tissue, street orientation and sky view trends of temperature patterns in Israel. Theor. Appl. Clim. 64, 163e177.
factors, among others. Therefore, a further study in needed to apply Bitan, A., Rubin, S., 1994. Climatic Atlas of Israel for Physical and Environmental
these findings to the specific urban tissue. Planning and Design. Ramot Publishing Co. Tel Aviv University, Tel Aviv.
Bornstein, R.D., 1975. The two dimensional URBMET urban boundary layer model.
J. Appl. Meteorol. 14, 1459e1477.
5. Conclusion Bornstein, R.D., Johanson, D.S., 1977. Urban-rural wind velocity differences. Atmos.
Environ. 11, 597e604.
Brazel, S.W., Bailing, J.R., 1986. Temporal analysis of long-term atmospheric mois-
The results obtained in the study suggest that:
ture levels in Phoenix, Arizona. J. Clim. Appl. Meteorol. 25 (2), 112e117.
Choi, Y., Jung, H.S., Nam, K.Y., Kwon, W.T., 2003. Adjusting urban bias in the regional
 The accumulated effect of urban climate modification and on a mean surface air temperature series of South Korea, 1968e1999. Int. J. Climatol.
larger scale the effect of environmental climate change and 23, 577e591. http://dx.doi.org/10.1002/joc.881.
lu, N., 2009. The effects of urbanization on water vapour pressure in
Cicek, I., Türkog
global warming may induce heat stress values in constructed a semi-arid climate. Theor. Appl. Clim. 95, 125e134. http://dx.doi.org/10.1007/
urban areas. s00704-007-0363-8.
122 O. Potchter, H. Itzhak Ben-Shalom / Journal of Arid Environments 98 (2013) 113e122

Cohen, P., potchter, O., Matzarakis, A., 2013. Human thermal per caption of Coastal Kalma, J.D., Oke, T.R., Aulicems, A. (Eds.), Biometeorology and Urban
Mediterranean outdoor urban environments. Appl. Geograph. 37, 1e10. Climatology at the Turn of the Millennium, pp. 273e278. Selected Papers
DeGaetano, A.T., Allen, R.J., 2002. Trends in twentieth-century air temperature ex- from Conference ICB-ICUC’99, Sydney, Australia. World Climate Application
tremes across the United States. J. Climatol. 15, 3188e3205. and Service Program. (WCASP-50), World Meteorological Organisation
Epstein, Y., Moran, D.S., 2006. Thermal comfort and heat stress indices. Ind. Health (WMO)/TD. 1026.
44, 388e398. Matzarakis, A., Rutz, F., Mayer, H., 2010. Modeling radiation fluxes in simple and
Fujibe, F., 2009. Detection of urban warming in recent air temperature trends in complex environments: basics of the Ray-Man model. Int. J. Biometeorol. 54,
Japan. Int. J. Clim. 29, 1811e1822. http://dx.doi.org/10.1002/joc.1822. 131e139. http://dx.doi.org/10.1007/s00484-009-0261-0.
Gaffen, D.J., Ross, R.G., 1998. Increased summertime heat stress in the US. Nature Mayer, H., Hoppe, P., 1987. Thermal comfort of man in different urban environ-
396, 529e530. http://dx.doi.org/10.1038/25030. ments. Theor. Appl. Climatol. 38, 43e49.
Gaffen, D.J., Ross, R.G., 1999. Climatology and trends in U.S. surface humidity and air McGregor, G.R., 1993. A preliminary assessment of the spatial and temporal char-
temperature. J. Clim. 12, 811e828. acteristics of human comfort in China. Int. J. Climatol. 13, 707e725.
Gaffin, S.R., Rosenzweig, C., Khanbilvardi, R., Parshall, L., Mahani, S., Glickman, H., Oke, T.R., 1987. Boundary Layer Climates, second ed. Methuen, New York, p. 435.
Goldberg, R., Blake, R., Slosberg, R.B., Hillel, D., 2008. Variations in New York City’s Oliveira, A.P., Bornstein, R.D., Soares, J., 2003. Annual and diurnal wind patterns in
urban heat island strength over time and space. Theor. Appl. Climatol. 94, 1e11. the city of São Paulo. Water Air Soil Pollut. 3, 3e15. http://dx.doi.org/10.1023/A:
Ganor, E., 1965. The Urban Climate of Beer Sheva: Differences Between the Centre 1026090103764.
and the Outskirts of a Town on the Desert Fringe (M.A. thesis in Geography). Parsons, K., 2003. Human Thermal Environment, second ed. Taylor and Francis,
The Hebrew University, Jerusalem (Hebrew). London, pp. 258e292.
Golden, J.S., 2004. The built environment induced urban heat island effect in rapidly Potchter, O., Saaroni, H., 1998. An examination of the map of climatic regions of
urbanizing arid regions e a sustainable urban engineering complexity. Environ. Israel, according to the KÖppen classification. Stud. Geogr. Isr. 15, 179e194
Sci. 1 (4), 321e349. (Hebrew).
Holmer, B., Eliasson, I., 1999. Urbanerural vapor pressure differences and their role Potchter, O., Yaakov, Y., Oren, G., 2006. The magnitude of the urban heat island
in the development of urban heat islands. Int. J. Climatol. 19 (9), 989e1009. of a city in an arid zone: the case of Beer Sheva, Israel. In: Proceeding of the
Hoppe, P., 1984. Die Energiebilanz des Menschen, p. 49. Wissenschaftliche Mittei- 6th International Conference on Urban Climate, Gothenburg, Sweden,
lungen des Meteorologischen Institutes der Universität München. pp. 450e453.
Hoppe, P., 1999. The physiological equivalent temperature e a universal index for Roth, M., 2007. Review of urban climate research in (sub) tropical regions. Int. J.
the biometeorological assessment of the thermal environment. Int. J. Bio- Climatol. 27, 1859e1873. http://dx.doi.org/10.1002/joc.1591.
meteorol. 43 (2), 71e75. Shafir, H., Alpert, P., 2011. Regional and local climatic effects on the Dead-Sea
Henry, J.A., Dirks, S.E., Marotz, G.A., 1985. Urban and rural humidity distribution: evaporation. Clim. Change 105, 455e468. http://dx.doi.org/10.1007/s10584-
relationships to surface materials and land use. J. Climatol. 5, 53e62. 010-9892-8.
IPCC, 2007. In: Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K., Shun Djen, C., 1992. The urban climate of Shanghai. Atmos. Environ. Part B. Urban
Tignor, M., Miller, H.L. (Eds.), Climate Change 2007: the Physical Science Basis. Atmos. 26 (1), 9e15.
Cambridge University Press, New York, p. 1056. Contribution of Working Group Skibin, D., 1979. Life quality in Beer Sheva: meteorological aspects. In: Gradus, Y.,
I to the Fourth Assessment Report of the IPCC. Stern, E. (Eds.), Beer Sheva. Keter, Jerusalem, pp. 323e331 (Hebrew).
Jendritzky, G., Menz, G., Schirmer, H., Schmidt-Kessen, W., 1990. Methodik zur Sohar, E., Adar, r., Kaly, J., 1963. Comparison of the environment heat load in various
räumlichen Bewertung der thermischen Komponente im Bioklima des Men- parts of Israel. Ball. Res. Counc. Isr. 10, 111e115.
schen. In: Beiträge der Akademie für Raumforschung und Landesplanung, Bd. Statistical Abstract of Israel, 2007. Statistical Abstract of Israel, vol. 58. Central Bu-
114. Hannover. reau of Statistics, Jerusalem.
Jauregui, A., Tejed, A., 1997. Urbanerural humidity contrasts in Mexico City. Int. J. Thom, E.C., 1959. The discomfort index. Weatherwise 12, 57e60.
Climatol. 17, 187e196. Thorsson, S., Lindberg, F., Björklund, J., Holmer, B., Rayner, D., 2011. Potential
Karl, T.R., Diaz, H.F., Kukla, G., 1988. Urbanization: its detection and effect in the changes in outdoor thermal comfort conditions in Gothenburg, Sweden due to
United States climate record. J. Clim. 1, 1099e1123. climate change: the influence of urban geometry. Int. J. Climatol. 31, 324e335.
Kolokotsa, D.A., Karapidakis, 2009. Heat island in southern Europe: the case study http://dx.doi.org/10.1002/joc.2231.
of Hania, Crete. Sol. Energy 83, 1871e1883. http://dx.doi.org/10.1016/ Thorsson, S., Lindqvist, M., Lindqvist, S., 2004. Thermal bioclimatic conditions and
j.solener.2009.06.018. patterns of behavior in an urban park in Göteborg, Sweden. Int. J. Biometeorol.
Kuttler, W., Weber, S., Schonnefeld, J., Hesselschwerdt, A., 2007. Urban/rural atmo- 48, 149e156.
spheric water vapour pressure differences and urban moisture excess in Krefeld, Um, H.H., Ha, K.J., Lee, S.S., 2007. Evaluation of the urban effect of long-term relative
Germany. Int. J. Climatol. 27, 2005e2015. http://dx.doi.org/10.1002/joc.1558. humidity and the separation of temperature and water vapor effects. Interna-
Lam, J.C., Wam, K.K.W., Wang, S.L., Lam, T.N.T., 2010. Long-term trends of heat stress tional Journal of Climatology 27, 1531e1542.
and energy use implications in subtropical climates. Appl. Energy 87, 608e612. VDI, 1998. Methods for the Human-biometeorological Evaluation of Climate and Air
http://dx.doi.org/10.1016/j.apenergy.2009.05.014. Quality for Urban and Regional Planning. Beuth, Berlin. Part I: Climate. VDI
Landsberg, H.E., 1981. The urban climate. Int. Geophys. Ser. 28, 275. Guideline 3787. Part 2.
Lin, T.P., Matzarakis, A., 2008. Tourism climate and thermal comfort in Sun Moon Wan, K.K.W., Wong, S.L., Yang, L., Lam, J.C., 2010. An analysis of the bioclimatic in
Lake, Taiwan. Int. J. Biometeorol. 52, 281e290. different climates and implications for the built environment in China. Build.
Lin, T.P., Matzarakis, A., Hawang, L.R., 2010. Shading effect on long-term outdoor Environ. 45, 1312e1318.
thermal comfort. Build. Environ. 45, 213e221. Wang, J.X.L., Gaffen, D.J., 2001. Late-twentieth-century climatology and trends of
Liu, W., You, H., Dou, J., 2009. Urbanerural humidity and air temperature differ- surface humidity and air temperature in China. J. Clim. 14, 2833e2845.
ences in the Beijing area. Theor. Appl. Clim. 96, 201e207. http://dx.doi.org/ Willett, K.M., Jones, P.D., Gllett, N.P., Thorne, P.W., 2008. Recent changes in surface
10.1007/s00704-008-0024-6. humidity: development of the HadCRUH dataset. J. Clim. 21, 5364e5383. http://
Matzarakis, A., Mayer, H., 1996. Another kind of environmental stress: thermal dx.doi.org/10.1175/2008JCLI2274.1.
stress. WHO News 18, 7e10. Ziv, B., Saaroni, H., 2011. The contribution of moisture to heat stress in a period of
Matzarakis, A., Rutz, F., Mayer, H., 2000. Estimation and calculation of the global warming: the case of the Mediterranean. Clim. Change 104, 305e315.
mean radiant air temperature within urban structures. In: de Dear, R.J., http://dx.doi.org/10.1007/s10584-009-9710-3.