See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/250221917

The urban heat island in Salamanca (Spain) and its relationship to

meteorological parameters

Article in Climate Research · June 2007

DOI: 10.3354/cr034039

CITATIONS READS

42 340

2 authors, including:

J. L. Labajo
Universidad de Salamanca
29 PUBLICATIONS 268 CITATIONS

SEE PROFILE

All content following this page was uploaded by J. L. Labajo on 22 April 2015.

The user has requested enhancement of the downloaded file.

 CLIMATE RESEARCH
Vol. 34: 39–46, 2007 Published June 14
Clim Res

The urban heat island in Salamanca (Spain) and its

relationship to meteorological parameters
M. S. Alonso*, M. R. Fidalgo, J. L. Labajo
Department of General and Atmospheric Physics, University of Salamanca, 37008 Salamanca, Spain

ABSTRACT: An urban heat island (UHI) was found to exist in Salamanca, Spain — a medium-sized
European city that has a continental climate and relatively little industrial pollution. We demonstrate
that urban heating can occur in cities with these characteristics, giving rise to a microclimate that may
alter the biological rhythm of the zone e.g. under these conditions, spring arrives earlier in the urban
zone then in the rural zone. The study was carried out using data from 2 recording stations (one in
and the other outside the city) for the years 1996–1998. The existence of a nocturnal UHI was
observed, with a highest annual mean value of 3.6°C and a lowest annual mean value (cool island) of
–0.9°C. The most intense nocturnal UHI was seen in autumn, while the strongest sinks occurred in
spring and summer. As in other types of city with different characteristics, the UHI was seen to vary
according to the atmospheric situation. The meteorological variables that most affected the UHI were
found to be: (1) wind, which at speeds of >~6 m s–1 prevented the development of UHIs; (2) cloudi-
ness, which altered the flux of incident solar radiation (the intensity of the nocturnal UHI was greater
with high clouds); and (3) atmospheric pressure, which characterised the days of atmospheric
stability or instability, leading to variations in the intensity of the UHI.

KEY WORDS: Urban heat island · Local climate · UHI sink · Wind · Cloudiness · Synoptic conditions ·
Salamanca
Resale or republication not permitted without written consent of the publisher

1. INTRODUCTION with frosts and increasing the distance of the snow limit
from the city. Repercussions on biological processes are
At small scales, human activity is of crucial impor- also important since UHIs may decrease the length of the
tance with regard to the behaviour of climate factors, cold period in cities; this means that the climatic charac-
generating artificial microclimates. The most impor- teristics of spring appear earlier, e.g. flowering occurs
tant manifestation of the impact of urbanisation on the earlier in city parks and gardens than for the same
climate is the difference in temperature between cities species in nearby rural areas, while the climatic charac-
and their surrounding rural zones, known as the ‘urban teristics of autumn are delayed. Additionally, a UHI in a
heat island’ (UHI). The effects of human activity on the city may lead to decreases in the use of heating systems
UHI range from air pollution to modifications of local in winter and increases in the use of air conditioning
climatic characteristics, and pose potentially serious systems in summer (López et al. 1993).
consequences for social welfare. The larger the city in Radiative processes play a very important role in the
question, the more important the UHI (López et al. generation of thermal contrasts between neighbouring
1993). Accordingly, most studies have been carried out zones (Hänel et al. 1990). Although there is often a
on fairly large cities throughout the world, especially local greenhouse effect caused by atmospheric pol-
those strongly affected by air pollution (Landsberg lution, which reduces the amount of solar radiation
1981, Saitoh & Hoshi 1992). reaching the ground, a combination of opposing fac-
Sometimes, the UHI phenomenon may alter the re- tors often create a temperature difference between the
gional climate, causing a decrease in the number of days city and the surrounding rural area. Urban heating

*Email: salgarci@terra.es © Inter-Research 2007 · www.int-res.com

40 Clim Res 34: 39–46, 2007

mainly derives from the absorption of the solar radia- 2. DATA AND METHODS
tion that reaches the ground and buildings from dawn
until the sun reaches its maximum height. When the The method used involved a comparison of the tem-
sun’s radiation reaches buildings, the ensuing succes- perature data series obtained from weather stations
sive reflections lead to more energy being confined located at urban and rural points, as has been used in a
within their sphere of influence than in rural settings number of other studies (García 1996, Sahashi & Hieda
(Wilby 2003). Materials forming the surface of cities 1996, Figuerola & Mazzeo 1998, Unger et al. 2001, Kim
usually have a greater heat absorption capacity than & Baik 2002, Alonso et al. 2003). The first step in a
natural soils, so that in urban zones the energy is study of this type is to select reliable weather stations
stored for longer than in rural zones (Hoyano et al. in zones with representative characteristics; e.g. rural
1999). In turn, long-wave radiation coming from the measuring stations should be situated outside the
ground is less able to cross the pollution layer, thus sphere of influence of the city.
further contributing to heating the urban zone. The intensity (I) of the UHI was calculated as the dif-
Meteorological factors also alter the energy balance ference between the urban mean temperature (Tu) and
existing between the ground and the top of the atmo- the rural mean temperature (Tr) (I = Tu – Tr) at the
sphere. For example, wind causes turbulence that climatological (i.e. synoptic) hours 00:00, 07:00, 13:00
homogenises the air temperature (Jáuregui 1988, and 18:00 h UTC. Following this, a statistical study was
Morris & Simmonds 2000), and clouds absorb or reflect performed, which first considered the mean and
solar radiation, thereby varying the amount of radia- extreme values of I in order to establish the range of
tion that reaches the ground (Labajo et al. 1988). values between which the intensity of the UHI fluctu-
Following the results of a pioneer study by Sund- ated. This was done considering the different hours of
borg (1950) into the temperature conditions of an the day or periods of the year separately. The annual
urban zone; and those of Morris & Simmons (2000) evolution of the UHI was established using a month-
and Morris et al. (2001) who related the intensity of by-month study, which in turn allowed the determina-
an UHI to meteorological factors, we were prompted tion of its range of intensity.
to determine the existence of a UHI in a medium- The relationship between the intensity of the UHI
sized European city and to analyse the temporal and meteorological factors was established using sim-
evolution of its intensity and its relationship to meteo- ple linear regressions, making an analysis of the corre-
rological variables. To conduct the study, we chose lation matrix. Plotting the joint evolutions of the inten-
the city of Salamanca (Spain), located on the northern sity of the UHI and the meteorological factors helped to
plateau of the Iberian Peninsula (40° 55’ N, 5° 28’ W), visualise the relationships found.
at a height of 800 m above sea level (Fig. 1). The The data corresponding to the rural setting were
population of the city is about 190 000; it is subject to those measured at the weather station of the observa-
a continental climate and receives little industrial tory network of the Instituto Nacional de Meterologìa
pollution, most air pollution being from traffic and (INM) located at Matacán (altitude 790 m). This station
central heating system emissions (Fidalgo et al. 1988). is considered to be ‘rural’ because it is sufficiently dis-
Throughout the year, the synoptic atmospheric situa-
tion is more stable than unstable since the city is
frequently under the influence of the Azores anti-
cyclone. Despite this, the climate of Salamanca can
be considered extreme, with annual thermal fluctua-
tions of up to 40°C, giving rise to very cold winters
and very hot summers. Most precipitation falls in
autumn and spring, while the summers are dry (La
Porte & Labajo 1983, Fidalgo 1994).
This is the first time that the UHI effect has been
studied in this part of Spain. Previous studies have
focused on Madrid (López et al. 1993), Barcelona
(Moreno-García 1994), and Granada (Montávez et al
2000). The data and methods used in the study are
described in Section 2. The evolution of the Salamanca
UHI over time, diurnally, seasonally and from year-
to-year is discussed in Section 3, while its relationships
to meteorological factors are explored in Section 4. Fig. 1. Location of the urban (Salamanca) and rural (Matacán)
Conclusions are presented in Section 5. study stations
 Alonso et al.: Urban heat island in Salamanca 41

tant from the city (16 km) and is in a green area. In the The nocturnal UHI is represented by the intensity
urban setting, the station ‘E1’ was used, which is calculated from the minimum temperatures (Imin), since
located in the heart of the city amid heavy traffic and a for these temperatures the highest mean positive val-
high density of buildings. The E1 recording station ues for the nocturnal UHI series are reached. The noc-
forms part of the atmospheric surveillance and control turnal UHI was found to have positive mean intensity
network of the regional community of Castilla y León. values. The lesser nocturnal cooling of the urban zone
For the urban station, the temperature data corre- compared to the rural zone is due to the slower rate of
sponding to those taken at 00:00, 07:00, 13:00 and emission of the energy accumulated during the day
18:00 h UTC were used in order to be able to make (Wilby 2003). Sometimes, the intensity of the nocturnal
direct comparisons with the rural station at Matacán, UHI reaches negative values. This can be attributed to
which makes measurements at these climatological the prevailing meteorological conditions, mainly the
hours. Data on the intensity of the heat island at daily arrival of warm air masses (Kim & Baik 2004).
maximum and minimum temperatures were also con-
sidered (maximum and minimum temperatures occur
at almost the same time at the rural and urban sta- 3.2. Seasonal evolution
tions). The study was carried out over a 3 yr period
(1996 to 1998). The most representative data sets of the daytime and
nocturnal UHI are, respectively, the intensities calcu-
lated using the maximum (Imax) and minimum (Imin)
3. EVOLUTION OVER TIME temperatures; thus these data series were considered
for the years 1996, 1997 and 1998. A temporal analysis
3.1. Daily evolution was conducted by season, these being: winter (De-
cember to February); spring (March to May); summer
The daily evolution of the intensity (I) of the UHI (June to August) and autumn (September to Novem-
over a 1 yr period (1996) was analysed first. The inten- ber). The mean values of the intensity of the seasonal
sity of the UHI was calculated for the daytime: 13:00 daytime and nighttime UHI (calculated using Imax and
and 18:00 h UTC, and at the hour of maximum temper- Imin) can be seen in Table 2.
ature (Imax); and for the nighttime: 00:00 and 07:00 h The most intense mean sinks of the daytime UHI are
UTC, and at the hour of minimum temperature (Imin), seen in spring. As considered for the daily evolution
because the initial conditions are dependent on the series, this is probably a result of the arrival of more
presence or absence of solar radiation. Table 1 shows solar radiation and more rapid heating of the ground
some of the statistical values for the intensity of the in the rural zone than in the urban zone. In turn, the
UHI, calculated for the day and night hours. highest mean values of the intensity of the nocturnal
The intensity calculated from the maximum temper- UHI occur in autumn (Alonso et al. 2003), probably due
ature (Imax) is representative of the daytime UHI. In to a slower nocturnal cooling of the urban ground com-
general, the daytime UHI displayed negative values, pared to the rural, together with the high incidence of
implying the existence of an ‘urban cool island’ (Mon- light winds frequent at this time of the year (La Porte &
távez et al. 2000). These negative values are a conse- Labajo 1983).
quence of the rapid daytime heating undergone by the
rural landmass as compared with the urban one.
3.3. Annual evolution
Table 1. Mean, maximum and minimum intensities I (°C)
of the Salamanca urban heat island (UHI) for daytime and The mean monthly values of daytime (Imax) and
nighttime hours in 1996. Imax: intensity at the maximum day- nighttime (Imin) intensities of the UHI for 1996, 1997
time temperature, Imin: intensity at the minimum nocturnal and 1998 are shown in Fig. 2.
temperature
During the cold months, the daytime UHI generally
shows the highest intensity, decreasing from January
Time (h) I (°C)
to May and increasing from September to December.
Mean Maximum Minimum
As in the case of the daily and seasonal evolutions, this
Day 13:00 –0.1 2.7 –3.7 may be due to the fact that in these months less solar
18:00 –0.3 5.7 –4.8 radiation reaches the ground, such that the daily heat-
Imax –0.9 6.5 –9.9 ing of the rural zone is less intense.
Night 00:00 1.3 6.5 –4.2 The annual mean and extreme values of the intensity
07:00 1.6 5.9 –3.3
of the daytime and nighttime UHI obtained over the
Imin 1.9 7.9 –4.6
period studied, 1996 to1998, are given in Table 3.
 42 Clim Res 34: 39–46, 2007

Table 2. Mean intensity I (°C) of the Salamanca UHI at day- Table 3. Annual mean, daytime (Imax) and nighttime ( Imin)
time and nighttime hours for each season in the years 1996 intensities (°C) of the Salamanca UHI at for the years 1996,
to 1998. Daytime values calculated using the intensity at 1997 and 1998
the maximum daytime temperature Imax; nocturnal values
calculated using the intensity at the minimum nocturnal
Year Mean Imax Imin
temperature Imin
Day 1996 –0.9 6.5 –9.9
Year I (°C) 1997 –0.9 1.4 –3.8
Winter Spring Summer Autumn 1998 –0.1 5.9 –5.5
Night 1996 1.9 7.9 –6.6
Day 1996 –0.6 –1.1 –1.0 –1.0
1997 2.9 7.8 –0.6
1997 –0.9 –1.6 –0.8 –0.4
1998 3.6 9.3 –3.4
1998 –0.0 0.0 –0.2 –0.2
Night 1996 1.1 1.4 1.9 2.5
1997 2.0 2.7 2.9 3.5
1998 3.7 3.3 3.4 3.6 4. RELATION TO METEOROLOGICAL FACTORS

The weather is the state of the atmosphere at a

Statistical study of the annual series corresponding given moment and place. In view of the definition of
to the nighttime UHI suggests the following defini- the UHI, it is logical to assume that the weather will
tions for Salamanca: a ‘weak’ island, when the inten- affect its intensity. The meteorological variables cho-
sity is < 2.0°C; a ‘moderate’ island, if the intensity lies sen to study the relationship between the intensity of
between (and including) 2.0°C and 4.0°C, and an the UHI and the synoptic situation were wind (both
‘intense’ island when the value is ≥ 4.0°C. speed and direction), cloudiness, and atmospheric
In Salamanca, maximum annual values of the night- pressure, using values measured at the station at
time UHI are in agreement with the value obtained Matacán. Since the most intense UHI was recorded
using the model proposed by Oke (1976) to calculate during nocturnal hours, only Imin was considered for
the maximum intensity reached by the nighttime UHI this aspect of the study.
as a function of the population of the city.
The differences observed in the intensities of the
UHI for the 3 yr studied may be due, in part, to the pre- 4.1. Wind
vailing synoptic conditions during these years, since
other factors, such as the morphology of the city or For wind conditions, both speed and direction were
sources of pollution, that might influence the intensity considered, although studies conducted by Oke for
of the UHI varied little from one year to another. In nighttime UHI in several cities have shown that the
comparing all 3 yr, it is also possible to identify inver- most influential meteorological variable affecting the
sions in the daytime heat UHI and positive values for nocturnal UHI is wind speed (Oke 1973; cited in
the nighttime UHI. Landsberg 1981).

Day Night
1.0 6
1996 1996
1997 1997
0.5 1998 5 1998

0.0 4
Imax (°C)

Imin (°C)

–0.5 3

Fig. 2. Mean monthly

–1.0 2 values of the intensity I
(°C) of the Salamanca
UHI measured at the
–1.5 1 maximum daytime tem-
perature Imax and mini-
mum nighttime temper-
–2.0 0 ature Imin for the years
Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 1996–1998
 Alonso et al.: Urban heat island in Salamanca 43

The evolution of the mean monthly intensities of the Regarding wind direction, the most frequent long-
nocturnal UHI (using Imin), through 1996 to 1998 com- term mean wind directions in Salamanca correspond
pared to the evolution of the mean monthly wind to the first (NE) and third (SW) quadrants, with a per-
speeds for the same period is shown in Fig. 3. It can centage of calm periods over the year of 21% (La Porte
be observed that the relative maximum of one of the & Labajo 1983). Table 5 shows mean values of the
variables tends to correlate chronologically with the intensities of the nighttime UHI, Imin, and wind speed
minimum of the other one; this applies to all months. at 07:00 h UTC (the time of occurrence of the minimum
October 1996 showed a maximum UHI intensity; temperatures with which the intensity of the nighttime
among other causes, this can be attributed to the fact UHI are normally recorded) for the most frequent wind
that the mean monthly wind speed during that month profiles (light wind, direction of wind in the first quad-
was only 0.6 m s–1. The opposite case was seen for rant, and wind direction in the third quadrant).
January 1996, which had a minimum UHI intensity, As expected, the most intense nighttime UHI were
attributable to the fact that the mean monthly wind associated with light wind situations. Regarding the
speed reached 3.7 m s–1. In general, the fact that the dominant wind directions, the highest intensity values
annual mean UHI intensity increased from 1996 to of the nocturnal UHI were reached when the wind was
1998 is consistent with changes in the number of in the first quadrant; lower values of nocturnal UHI in-
hours of light winds. tensity were recorded when the wind was coming from
According to Oke & Hannel (1970) and Oke (1976), the third quadrant. Moreover, the influence of wind
a wind speed — the ‘limiting wind speed’— exists for speed on the UHI is confirmed, since normally the wind
which the intensity of the UHI becomes null. In the coming from the first quadrant is less strong than that
case of Salamanca, this speed, which is a function of coming from the third quadrant (Fidalgo 1994).
the population, is close to 6 m s–1. From a simple linear According to the climatology of Salamanca for the
regression between the intensity of the nighttime UHI period of 1945–1974 (La Porte & Labajo 1983), the pre-
and the daily mean wind speed, it is possible to estab- dominant wind direction throughout the year is the
lish the correlation between both variables (Table 4). third quadrant, this direction also predominating
Additionally, we gave the intensity of the UHI a value monthly, with the exception of April and July, when it
of zero in the regression equation and so obtained a predominantly comes from the first quadrant. The low-
mean value for the limiting wind speed in Salamanca est wind speeds in Salamanca are recorded in autumn,
of 6.4 m s–1. The linear correlation coefficients (r) consistent with the finding that the intensity of the UHI
between both variables were, as expected, negative. is greatest in that season. March has the highest wind
The limiting speed obtained was close to that calcu- speeds, such that the corresponding intensity of the
lated theoretically with the model of Oke & Hannel UHI should in general be low. However, the intensities
(1970) and Oke (1976). of the UHI in March appear as relative maxima in the
annual evolution for the years we studied (Fig. 2). The
6 6
Imin Table 4. Regressions between the nighttime Salamanca UHI
–1 intensity (measured at the nocturnal minimum temperature
5 V (m s ) 5 Imin) and wind speed (V ). Vlim = limiting wind speed

4 4 Imin r Vlim (m s–1)

V (m s–1)
Imin (°C)

1996 –0.16 V + 3.63 –0.55 6.3

3 3 1997 –0.18 V + 4.57 –0.55 7.1
1998 –0.25 V + 5.81 –0.56 6.4
1996 to 1998 –0.21 V + 4.80 –0.55 6.4
2 2

Table 5. Mean values of the Salamanca UHI intensity at noctur-

1 1 nal minimum temperature Imin (°C) and mean values of wind
velocity at 07:00 h (I) V7 for the most frequent wind profiles
1996–1998
0 0
Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct
1996 1997 1998 Imin V7
Month
Light wind 3.7 0.0
Fig. 3. Mean values of the intensity of the Salamanca UHI 1st Quadrant 2.6 7.5
at the minimum nocturnal temperature Imin (°C) and avergage 3rd Quadrant 1.4 12.4
values of wind speed V (m s–1) for the years 1996–1998
44 Clim Res 34: 39–46, 2007

probable reason for this is that for these years the when the latter reaches 3/8 of clouded sky the UHI
March wind speed was low, with a large number of intensity increases again, but never up to the intensity
hours of light winds. obtained with cloudless sky. An explanation for the
increase in the intensity of the UHI when a value of 3/8
of cloudiness is reached can be found in the type of
4.2. Cloudiness cloud present. The mean intensity of the UHI was cal-
culated as a function of the type of cloud covering.
To study the relationship between cloudiness and During the period studied we considered only low
the intensity of the nighttime UHI, the cloudiness clouds and high clouds because, in principle, they can
recorded at 13:00 h UTC at Matacán for the 3 yr period be considered as the most influential as regards the
1996 to 1998 was used. This variable was employed energy received at the ground (Labajo et al. 1988). The
because the greatest amount of solar radiation reaches results are shown in Table 7.
the ground at midday. The joint monthly evolution of In all cases, the intensity of the nighttime UHI was
Imin with the number of clear days (0 eighths [0/8] of greater when the predominant clouds were high,
cloud cover) per month is shown in Fig. 4a. The evolu- suggesting that low clouds absorb more solar radia-
tion of the nocturnal UHI together with the number of tion than high ones. Moreover, high clouds, espe-
cloudy days (8/8 cloud cover) is shown in Fig. 4b. cially Cc, may allow up to 20% more radiation to
These plots indicate a positive relationship between reach the ground than low clouds (Labajo et al.
the intensity of the nighttime UHI and the number of 1988). This could explain why for a cloud cover of
clear days and a negative relationship with the number 3/8 formed by cirrus-like clouds, the intensity of the
of cloudy days, i.e. the maximum intensity generally UHI is greater than with a 2/8 cover formed by
coincides with the maximum number of clear days and clouds of other types.
in turn the minimum intensity generally coincides with To conclude the study, we simultaneously consid-
the minimum of cloudy days. ered the effect of wind speed and cloudiness. First, we
We next considered only the days on which the wind removed the cases with light winds and winds whose
was light, for the 07:00 h UTC observation, with a view speeds were greater than 6.1 m s–1 (the theoretical
to removing the influence of the wind from the cloud limiting speed in Salamanca — see Section 4.1). We
relationship. Clear days and those of 1/8, 2/8 or 3/8 then calculated the mean intensity of the nighttime
cloudiness were considered separately. Table 6 shows UHI for those days according to cloudiness. The
the mean values of the intensities of the nighttime UHI results are shown in Table 6b. If the wind is consid-
(Imin) according to cloudiness for 1997 and 1998; the ered, the mean value of the intensity of the nocturnal
number of days considered in each case is also shown. UHI decreases. These results are in agreement with
It can be seen that the mean intensity of the night- those reported by other authors for larger cities (Morris
time UHI decreases as cloudiness increases, although et al. 2001).

Clear days (0/8) Cloudy days (8/8)

25 25 20 20
Imin a b Imin
Days Days
20 20
15 15

15 15
Days (n)

Days (n)
Imin (°C)

Imin (°C)

10 10

10 10

5 5
5 5

0 0 0 0
Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct Jan Apr Jul Oct
1996 1997 1998 1996 1997 1998
Month Month
Fig. 4. Mean values of the intensity of the Salamanca UHI at the minimum nocturnal temperature Imin (°C) and the number (n) of:
(a) clear days (0/8 cloud cover); and (b) cloudy days (8/8 cloud cover) for the years 1996–1998
 Alonso et al.: Urban heat island in Salamanca 45

Table 6. Mean values of the Salamanca UHI intensity at Spanish city of Salamanca — characterised by a very
nocturnal minimum temperature Imin (°C) for the years large seasonal temperature cycle — is investigated for
1997 and 1998 according to cloudiness and wind speed.
the first time. Also novel is the particular emphasis
Low wind speed: 0 <V7 < 6.3 m s–1. n = no. of days
given to the relationship between UHI intensity and
cloudiness.
Cloud cover All wind speeds Low wind speeds
Imin n Imin n
The daily evolution of the UHI in this medium-sized
European city is shown to display 2 clearly differenti-
0/8 4.5 62 3.9 25 ated situations: on the one hand, the daytime UHI
1/8 4.2 52 3.0 22 shows negative mean values, which gives rise to the
2/8 3.5 61 2.4 20
3/8 4.4 38 2.6 18 so-called ‘urban cool island’, in particular caused by
the rapid heating undergone by the rural zone when
receiving solar radiation (López et al. 1993, Figuerola
Table 7. Mean values of the Salamanca UHI intensity at noc- & Mazzeo 1998). In the city of Salamanca, the strongest
turnal minimum temperature Imin (°C) according to the type of daytime urban cool island occurs in spring. On the
cloud (low or high) and the cloud cover (1/8 to 3/8) for the
other hand, there is a nighttime UHI with positive
years 1997 and 1998. n = number of days
mean annual values, and this is the result of the slower
nocturnal cooling undergone by the urban zone. The
Cloud cover Cloud type Imin n
highest mean intensities of the nighttime UHI in
1/8 Low 3.7 35 Salamanca are reached in autumn when, according to
High 5.3 17 the long-term climatology of the city, the wind speeds
2/8 Low 3.0 36 are lowest.
High 4.0 25
The following classification of the nighttime UHI
3/8 Low 3.7 13
intensity in Salamanca has been established: weak
High 4.8 25

4.3. Synoptic pressure conditions

An anticyclonic situation implies a settled atmo-

sphere with light wind conditions and cloudless skies.
However, a low atmospheric pressure situation implies
an unsettled atmosphere. To study the relationship
between the intensity of the nocturnal UHI and atmos-
pheric pressure, the synoptic conditions of a few days
in 1996, 1997 and 1998 were used.
It is evident that days with an intensity of the UHI
higher than its annual mean value are associated with
anticyclonic situations over the Iberian Peninsula. As
an example, we show the surface synoptic map for 29
July 1998 (Fig. 5), when wind was light and skies Fig. 5. Synoptic conditions on 29 July 1998 at 12:00 h UTC
cloudless. The intensity of the nighttime UHI was 7.8°C.
In contrast, if a day with an intensity of UHI lower
than annual mean value is considered, low atmospheric
pressure over the Iberian Peninsula is observed. This
occurred for example, on 4 March 1998 (Fig. 6), which
had a nighttime UHI intensity of only 1.1°C. The
weather conditions on this day were high wind speed
(8 m s–1 at 07:00 h UTC) and 3/8 cloud cover.

5. CONCLUSIONS

Although the UHI is a well-established pheno-

menon, previous studies have focused on large, non-
European cities. Understanding how the UHI operates
in different regions is important. Here, the UHI in the Fig. 6. Synoptic conditions on 4 March 1998 at 12:00 h UTC
 46 Clim Res 34: 39–46, 2007

island (lower than 2.0°C); moderate island (between elements contained in the aerosols of Salamanca (Spain).
2.0°C and 4.0°C), and intense island (greater than Atmos Environ 22:1405–1498
Figuerola P, Mazzeo N (1998) Urban–rural temperature diffe-
4.0°C).
rences in Buenos Aires. Int J Climatol 18:1709–1723
The intensity of the UHI shows temporal variations García F (1996) Manual de climatología aplicada. Clima,
that, in general, are due to the prevailing meteorologi- medioambiente y planificación. Síntesis, Madrid
cal situation. Those with the greatest influence in the Hänel G, Busen R, Weidert D (1990) Absorption of solar
UHI are considered to be wind, cloudiness and synop- radiation in an urban atmosphere. Atmos Environ B 24:
283–292
tic pressure conditions. Hoyano A, Lino A, Ono M, Tanighchi S (1999) Analysis of the
The wind speed negatively affects the intensity of influence of urban form and materials on sensible heat
the UHI and even abolishes it when a limiting value flux, a case study of Japan’s largest housing development,
is reached. In the case of Salamanca, the limiting Tama New Town. Atmos Environ 33:3931–3939
Jáuregui E (1988) Local wind and air pollution interaction in
speed is 6.4 m s–1, calculated from the mean daily
the Mexico basin. Atmósfera 1:131–140
wind speed data and consistent with theoretical esti- Kim YH, Baik JJ (2002) Maximum urban heat island intensity
mates. The direction of the wind, i.e. of the general in Seoul. J Appl Meteorol 41:651–659
flow, prevailing over Salamanca is from the first and Kim YH, Baik JJ (2004) Daily maximum urban heat island
third quadrants. The mean intensity of the UHI is intensity in large cities of Korea. Theor Appl Clim 79:
151–164
higher for the cases of the first quadrant since this is Labajo JL, Labajo A, Labajo S, Izquierdo MJ (1988) Influencia
associated with the lowest wind speeds. However, the de las nubes altas en el flujo de la radiación solar global.
most intense nighttime UHI are associated with light Rev Geofísica 44:61–68
wind, a conclusion that can be extrapolated to any Landsberg HE (1981) The urban climate. Academic, New
York
medium-sized city, whatever its predominant wind
La Porte P, Labajo JL (1983) El clima de Matacán (Sala-
direction. manca). INM A-74. Instituto Nacional de Meterologìa,
In general, cloudiness causes the intensity of the UHI Madrid
to decrease, unless there is a predominance of high López A, Lopez J, Fernandez F, Moreno A (1993) El clima
clouds, in which case the intensity of the nighttime urbano: teledetección de la isla de calor en Madrid.
MOPT, Madrid
UHI increases. Montávez JP, Rodríguez A, Jiménez JI (2000) A study of the
Regarding synoptic pressure conditions, anticyclonic urban heat island of Granada. Int J Climatol 20:899–911
situations imply atmospheric stability that favours the Moreno-García C (1994) Intensity and form of the urban heat
establishment of more intense UHI, whereas situations island in Barcelona. Int J Climatol 14:705–710
Morris CJG, Simmonds I (2000) Associations between varying
of atmospheric instability lead to decreases in the
magnitudes of the urban heat island and the synoptic
intensity of UHI. climatology in Melbourne, Australia. Int J Climatol 20:
The phenomenon of urban heating may by altered 1931–1954
by changing the structure of the city, in particular by Morris CJG, Simmonds I, Plummer N (2001) Quantification of
adding green zones in the city centre, where heating is the influences of wind and cloud on the nocturnal heat
island of a large city. J Appl Meteorol 40:169–182
usually accentuated (Upmanis et al. 1998). Under- Oke TR (1976) The distinction between canopy and boundary
standing the intensity and variability of the UHI and its layer urban heat islands. Atmosphere 14:268–277
relationships with meteorological variables, particu- Oke TR, Hannel FG (1970) The form of the urban heat island
larly with cloudiness, is an essential first step prior to in Hamilton, Canada. WMO Tech Note 108:113–126
Sahashi K, Hieda T (1996) Nitrogen-oxide layer over the
implementing and monitoring such remedial measures
urban heat island in Okayama City. Atmos Environ 30:
and also for assessing how the UHI might change in 531–535
the future, for example, in response to enhanced Saitoh TS, Hoshi H (1992) Urban warming in Tokyo metro-
greenhouse warming. olitan area. International Conference on Thermal Envi-
ronment (CUTEST’92), Fukuoka, p 11–17
Sundborg A (1950) Local climatological studies of the temper-
LITERATURE CITED ature conditions in an urban area. Tellus 2:222–232
Unger J, Sümeghy Z, Zoboki J (2001) Temperature cross-
Alonso MS, Labajo JL, Fidalgo MR (2003) Characteristics of section features in an urban area. Atmos Res 50:117–127
the urban heat island in the city of Salamanca, Spain. Upmanis H, Eliasson I, Lindsquist S (1998) The influence of
Atmósfera 16:137–148 green areas on nocturnal temperatures in a high latitude
Fidalgo MR (1994) El Clima: fichas de diagnóstico de salud. city (Göteborg, Sweden). Int J Climatol 18:681–691
Ayuntamiento de Salamanca, Salamanca Wilby RL (2003) Past and projected trends in London’s urban
Fidalgo MR, Mateos J, Garmendia J (1988) The origin of some heat island. Weather 58:251–260

Editorial responsibility: Clare Goodess, Submitted: December 18, 2002; Accepted: April 15, 2006
Norwich, UK Proofs received from author(s): May 13, 2007

View publication stats