• Alpert, P., , I. Osetinsky, , B. Ziv, , and H. Shafir, 2004: A new seasons definition based on classified daily synoptic systems: An example for the eastern Mediterranean. Int. J. Climatol., 24 , 10131021.

    • Search Google Scholar
    • Export Citation
  • Bitan, A., , and S. Rubin, 1994: Climatic Atlas of Israel for Physical and Environmental Planning and Design. Ramot Publishing, 358 pp.

  • Brazdil, R., , and M. Budıkova, 1999: An urban bias in air temperature fluctuations at the Klementinum, Prague, the Czech Republic. Atmos. Environ., 33 , 42114217.

    • Search Google Scholar
    • Export Citation
  • Comrie, A. C., 2000: Mapping a wind-modified urban heat island in Tucson, Arizona. Bull. Amer. Meteor. Soc., 81 , 24172431.

  • Dayan, U., , and J. Rodnizki, 1999: The temporal behavior of the atmospheric boundary layer in Israel. J. Appl. Meteor., 38 , 830836.

  • Dayan, U., , B. Lifshitz-Goldreich, , and K. Pick, 2002: Spatial and structural variation of the atmospheric boundary layer during summer in Israel—Profiler and rawinsonde measurements. J. Appl. Meteor., 41 , 447457.

    • Search Google Scholar
    • Export Citation
  • Goldreich, Y., 2003: The Climate of Israel: Observation, Research and Application. Kluwer Academic/Plenum, 298 pp.

  • Hafner, J., , and S. Q. Kidder, 1999: Urban heat island modeling in conjunction with satellite-derived surface/soil parameters. J. Appl. Meteor., 38 , 448465.

    • Search Google Scholar
    • Export Citation
  • Hawkins, T. W., , A. J. Brazel, , W. L. Stefanov, , W. Bigler, , and E. L. Saffell, 2004: The role of rural variability in urban heat island determination for Phoenix, Arizona. J. Appl. Meteor., 43 , 476486.

    • Search Google Scholar
    • Export Citation
  • Jonsson, P., 2004: Vegetation as an urban climate control in the subtropical city of Gaborone, Botswana. Int. J. Climatol., 24 , 13071322.

    • Search Google Scholar
    • Export Citation
  • Karl, T. R., , and R. W. Knight, 1997: The 1995 Chicago heat wave: How likely is a recurrence? Bull. Amer. Meteor. Soc., 78 , 11071119.

  • Kim, Y. H., , and J. J. Baik, 2002: Maximum urban heat island in Seoul. J. Appl. Meteor., 41 , 651659.

  • Landsberg, H. E., 1970: Micrometeorological temperature differentiation through urbanization. Urban Climates, WMO Publ. 254, TP 141, World Meteorological Organization, 129–136.

    • Search Google Scholar
    • Export Citation
  • Lowry, W. P., 1977: Empirical estimation of the urban effects on climate: A problem analysis. J. Appl. Meteor., 16 , 129135.

  • Martinez, A., , C. Yague, , and E. Zurita, 1991: Statistical analysis of the Madrid urban heat island. Atmos. Environ., 25 , 327332.

  • Meehl, G., , and C. Tebaldi, 2004: More intense, more frequent, and longer lasting heat waves in the 21st century. Science, 305 , 994997.

    • Search Google Scholar
    • Export Citation
  • Mendonca, F., 2009: Urban heat and urban cool islands: Influences of vegetation and soil surface in some cities, southern Brazil. Proc. Seventh Symp. on Urban Climate, Yokohama, Japan, International Association for Urban Climate, P1-39.

    • Search Google Scholar
    • Export Citation
  • Morris, C. J. G., , I. Simmonds, , and N. Plummer, 2001: Quantification of the influences of wind and cloud on the nocturnal urban heat island of a large city. J. Appl. Meteor., 40 , 169182.

    • Search Google Scholar
    • Export Citation
  • Nasrallah, H. A., , A. J. Brazel, , and R. C. Balling, 1990: Analysis of the Kuwait city urban heat island. Int. J. Climatol., 10 , 401405. doi:10.1002/joc.3370100407.

    • Search Google Scholar
    • Export Citation
  • Oke, T. R., 1976: The distinction between canopy and boundary-layer urban heat islands. Atmosphere, 14 , 269277.

  • Oke, T. R., 1981: Canyon geometry and the nocturnal urban heat island: Comparison of scale model and field observations. J. Climatol., 1 , 237254.

    • Search Google Scholar
    • Export Citation
  • Oke, T. R., 1982: The energetic basis of the urban heat island. Quart. J. Roy. Meteor. Soc., 108 , 128.

  • Oke, T. R., 1987: Boundary Layer Climates. Methuen Press, 435 pp.

  • Oke, T. R., 2004: Initial guidance to obtain representative meteorological observations at urban sites. IOM Rep. 81, World Meteorological Organization, 47 pp. [Available online at http://www.wmo.int/pages/prog/www/IMOP/publications/IOM-81/IOM-81-UrbanMetObs.pdf].

    • Search Google Scholar
    • Export Citation
  • Potchter, O., , Y. Yaacov, , and G. Oren, 2006: The magnitude of the urban heat island of a city in an arid zone: The case of Beer Sheva, Israel. Proc. Sixth Int. Conf. on Urban Climate, Gothenburg, Sweden, International Association for Urban Climate, 450–453.

    • Search Google Scholar
    • Export Citation
  • Roth, M., 2007: Review of urban climate research in (sub)tropical regions. Int. J. Climatol., 27 , 18591873.

  • Saaroni, H., , E. Ben-Dor, , A. Bitan, , and O. Potchter, 2000: Spatial distribution and microscale characteristics of the urban heat island in Tel-Aviv, Israel. Landsc. Urban Plan., 48 , 118.

    • Search Google Scholar
    • Export Citation
  • Saaroni, H., , B. Ziv, , J. Edelson, , and P. Alpert, 2003: Long-term variations in summer temperatures over the eastern Mediterranean. Geophys. Res. Lett., 30 , 1946. doi:10.1029/2003GL017742.

    • Search Google Scholar
    • Export Citation
  • Saaroni, H., , A. Bitan, , E. Ben-Dor, , and N. Feller, 2004: The mixed results concerning the ‘oasis effect’ in a rural settlement in the Negev Desert, Israel. J. Arid Environ., 58 , 235248. doi:10.1016/j.jaridenv.2004.08.010.

    • Search Google Scholar
    • Export Citation
  • Saaroni, H., , B. Ziv, , I. Osetinski, , and P. Alpert, 2010: Factors governing the interannual variation and the long-term trend of the 850 hPa temperature over Israel. Quart. J. Roy. Meteor. Soc., 136 , 305318. doi:10.1002/qj.580.

    • Search Google Scholar
    • Export Citation
  • Shigeta, Y., , Y. Ohashi, , and O. Tsukamoto, 2009: Urban cool island in daytime—Analysis by using thermal image and air temperature measurements. Proc. Seventh Symp. on Urban Climate, Yokohama, Japan, International Association for Urban Climate, P5-7.

    • Search Google Scholar
    • Export Citation
  • Sofer, M., , and O. Potchter, 2006: The urban heat island of a city in an arid zone: The case of Eilat, Israel. Theor. Appl. Climatol., 85 , 8188.

    • Search Google Scholar
    • Export Citation
  • Stewart, I. D., , and T. R. Oke, 2009a: A new classification system for urban climate sites. Bull. Amer. Meteor. Soc., 90 , 922923.

  • Stewart, I. D., , and T. R. Oke, 2009b: Newly developed “thermal climate zones” for defining and measuring urban heat island magnitude in the canopy layer. Preprints, T. R. Oke Symp. and Eighth Symp. on Urban Environment, Phoenix, AZ, Amer. Meteor. Soc., J8.A. [Available online at http://ams.confex.com/ams/pdfpapers/150476.pdf].

    • Search Google Scholar
    • Export Citation
  • Unger, J., 1996: Heat island intensity with different meteorological conditions in a medium-sized town: Szeged, Hungary. Theor. Appl. Climatol., 54 , 147151.

    • Search Google Scholar
    • Export Citation
  • Unger, J., , Z. Sümeghy, , A. Gulyas, , Z. Bottyan, , and L. Mucsi, 2001: Land-use and meteorological aspects of the urban heat island. Meteor. Appl., 8 , 189194.

    • Search Google Scholar
    • Export Citation
  • Ziv, B., , H. Saaroni, , and P. Alpert, 2004: The factors governing the summer regime of the eastern Mediterranean. Int. J. Climatol., 24 , 18591871.

    • Search Google Scholar
    • Export Citation
  • Ziv, B., , H. Saaroni, , A. Baharad, , D. Yekutieli, , and P. Alpert, 2005: The indications for aggravation in summer heat conditions over the Mediterranean Basin. Geophys. Res. Lett., 32 , L12706. doi:10.1029/2005GL022796.

    • Search Google Scholar
    • Export Citation
  • View in gallery

    (a) The location of the study region. (b) Aerial photograph of the study region with notations of the rural and urban stations. (c) Topographical cross sections from north to south across the rural (black line) and the urban (gray line) stations.

  • View in gallery

    Interannual variation of ΔTur (°C) for the period 1967–2004 and the long-term linear trend for the minimum and maximum temperatures during the (a) summer and (b) winter.

  • View in gallery

    Variation of ΔTur (°C) for the period 1967–2004 as a function of Beer Sheba population and the corresponding linear relationship, for the minimum and maximum temperatures, for the (a) summer and (b) winter. The dashed extensions of the linear lines are extrapolated to zero population.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 64 64 14
PDF Downloads 41 41 8

Estimating the Urban Heat Island Contribution to Urban and Rural Air Temperature Differences over Complex Terrain: Application to an Arid City

View More View Less
  • 1 Department of Geography and the Human Environment, Tel Aviv University, Tel Aviv, Israel
  • | 2 Department of Natural Sciences, The Open University of Israel, Raanana, Israel
© Get Permissions
Full access

Abstract

This study proposes a method for estimating the canopy-layer net urban heat island (UHI) in regions with complex terrain that lack preurban observations. The approach is based on a linear relationship between the urban–rural temperature difference (ΔTur), measured via screen-level air temperature, and the population of the city, which was found to have the highest correlation with observations. The linear relation is extrapolated to zero population to yield the desired preurban value. The difference between the zero population ΔTur and the current one is proposed to represent the net UHI. Given the uncertainties of the population method, the relatively short time period of the temperature record, and possible inhomogeneity in the data, the results should be regarded as a first-order approximation of the net UHI contribution. The UHI was evaluated for an arid city, Beer Sheba, Israel, for the minimum and maximum air temperatures for the summer and the winter. The study region resembles the combined effect of complex terrain (i.e., the concave topography of the city in contrast with the plateau landscape surrounding it), the UHI, and the regional warming trend. The study assumes that the regional warming does not affect the ΔTur. The concave topography of the city dominates over the UHI contribution during nighttime, resulting in an average lower minimum temperature in the city relative to the rural area. This difference has decreased considerably during the study period and has even reversed for the summer nights toward the end of the period. The estimated net UHI contribution in Beer Sheba varies between +0.8° and +3.1°C, with the highest values during the night hours. The high positive UHI during the night is in line with previous studies. The positive UHI in the summer implies further aggravation of heat stress beyond that occurring, and that predicted to increase, over the region.

Corresponding author address: Dr. Hadas Sooroni, Tel Aviv University, POB 39040, Tel Aviv 69978, Israel. Email: saaroni@post.tau.ac.il

Abstract

This study proposes a method for estimating the canopy-layer net urban heat island (UHI) in regions with complex terrain that lack preurban observations. The approach is based on a linear relationship between the urban–rural temperature difference (ΔTur), measured via screen-level air temperature, and the population of the city, which was found to have the highest correlation with observations. The linear relation is extrapolated to zero population to yield the desired preurban value. The difference between the zero population ΔTur and the current one is proposed to represent the net UHI. Given the uncertainties of the population method, the relatively short time period of the temperature record, and possible inhomogeneity in the data, the results should be regarded as a first-order approximation of the net UHI contribution. The UHI was evaluated for an arid city, Beer Sheba, Israel, for the minimum and maximum air temperatures for the summer and the winter. The study region resembles the combined effect of complex terrain (i.e., the concave topography of the city in contrast with the plateau landscape surrounding it), the UHI, and the regional warming trend. The study assumes that the regional warming does not affect the ΔTur. The concave topography of the city dominates over the UHI contribution during nighttime, resulting in an average lower minimum temperature in the city relative to the rural area. This difference has decreased considerably during the study period and has even reversed for the summer nights toward the end of the period. The estimated net UHI contribution in Beer Sheba varies between +0.8° and +3.1°C, with the highest values during the night hours. The high positive UHI during the night is in line with previous studies. The positive UHI in the summer implies further aggravation of heat stress beyond that occurring, and that predicted to increase, over the region.

Corresponding author address: Dr. Hadas Sooroni, Tel Aviv University, POB 39040, Tel Aviv 69978, Israel. Email: saaroni@post.tau.ac.il

1. Introduction

The urban heat island (UHI) has been studied mostly through the temperature differences between urban and neighboring rural stations (ΔTur). These studies refer to the “canopy-layer heat island” (Oke 1976, 1987, 252–302) via screen-level air temperatures (∼2 m AGL). Such an approach may be found to be problematic because ΔTur may be influenced by differences in topographic posting between the urban and the rural regions and by spatial and temporal variations in the characteristics of the rural region itself. “In selecting station pairs it is particularly important to try to eliminate extraneous effects due to topography, water bodies and the downwind effects of the urban area itself” (Oke 1987, p. 273). According to Lowry (1977), Oke (1987, 252–302), and Stewart and Oke (2009b), the ideal database needed for evaluating the UHI is an extensive series of preurban measurements that can be compared with present-day observations. However, only in rare cases are such data available, and ΔTur is commonly used to represent the magnitude of the UHI. Stewart and Oke (2009b) claim that “one-third of empirical UHI literature gives no quantitative or qualitative description of the measurement sites defining UHI magnitude … ”.

The intensity of UHI depends on several factors. It is most intense under stable weather, calm air, and cloudless skies, defined by Oke (1982) as ideal conditions for the development of UHI. Clouds and strong winds have been shown to suppress it (e.g., Morris et al. 2001; Unger et al. 2001). UHI magnitude reaches nighttime maxima because of differential radiative cooling above urban and rural surfaces, which in turn is a response to surface geometry (urban canyon vs open rural) and heat storage [urban concrete vs rural soils; e.g., Oke (1981, 1982)]. Thus, the UHI is most clearly seen in daily minimum temperatures (e.g., Martinez et al. 1991; Unger 1996). Several studies have demonstrated the intensification of the UHI in time together with the increase of the urbanization, for example, in Madrid, Spain (Martinez et al. 1991); Tucson, Arizona (Comrie 2000); Seoul, Korea (Kim and Baik 2002); and Prague, Czech Republic (Brazdil and Budıkova 1999).

Most studies of the UHI have focused on mid- and high latitudes. UHI intensity in subtropical regions is regarded as lower relative to that in temperate cities with similar population and shows a seasonal variation, with higher intensity during the dry season (Roth 2007). Arid cities, such as Tucson and Phoenix, Arizona (Comrie 2000; Hawkins et al. 2004), showed significant UHI, on the order of 3°C. On the contrary, in Kuwait City, Kuwait, the UHI is negligible because of the similarities in the urban and rural landscapes and because the city is in close proximity to a large water body (Nasrallah et al. 1990). Saaroni et al. (2000) found significant UHI in Tel Aviv, Israel (subjected to Mediterranean climate), reaching 5°C in the canopy layer of a stable winter night. For desert cities in Israel, Sofer and Potchter (2006) studied the UHI of the superarid city of Eilat for several case studies in the summer and winter seasons. Their measurements indicate the development of moderate UHI, 0.5°–1.5°C in magnitude, that is more significant at midday, during the summer. Potchter et al. (2006) studied the magnitude of the UHI in Beer Sheba, Israel, through several case studies and found a maximum magnitude, of 6.9°C, in the early-morning hours.

The time evolution of the UHI can be measured by comparing the long-term temperature trend in the urban and rural regions. Nasrallah et al. (1990) showed that the long-term temperature trend in Kuwait City was 0.12°C (10 yr)−1, larger than the 0.07°C (10 yr)−1 of the near desert area. Comrie (2000), using a similar method, showed that in Tucson the UHI increased by 3°C over the twentieth century, with >2°C of this warming within the last 30 yr. Because the eastern Mediterranean region is subjected to a more intense warming trend during the summer season relative to the global rate (Saaroni et al. 2003; Ziv et al. 2005), any signal of positive UHI implies further aggravation of summer heat conditions within urban regions there.

The complexity of the factors affecting the ΔTur implies that the UHI cannot be deduced directly from the temperature difference. This study proposes a method according to which the time variation of ΔTur against population is used to estimate the net contribution of the UHI. The method is applied for the arid city of Beer Sheba.

2. Geographical and climatological background

The city of Beer Sheba (31°14′N, 34°47′E, 260 m MSL) is the largest city in the Negev Desert of Israel (Figs. 1a and 1b). The city covers 54 km2, and its population in 2005 was 185 000 [from the Statistical Abstract of Israel 2008 (No. 59) of the Central Bureau of Statistics; obtained online at http://www1.cbs.gov.il/reader/shnatonenew.htm (URL displays current version)], as compared with 70 000 in 1967, the beginning of the study period. The meteorological station (the urban station, hereinafter, belonging to the Israeli Meteorological Service) is located in one of the residential neighborhoods, in the commercial center of the city (Fig. 1b). The city of Beer Sheba is dominated by a “block”-type landscape [according to Oke (2004) and Stewart and Oke (2009a)] with 47% built area, a typical building height of four floors, and a sky-view factor (SVF) of ∼0.8 due to relatively large roads.

The rural station [World Meteorological Organization (WMO) Climate Station 401910] is located 10 km west of Beer Sheba, at an elevation of 195 m MSL, 65 m lower than the urban station (Fig. 1b). The topographic posting of this station is a plateau, unlike the urban area, which has a concave landscape (Fig. 1c). The area is a “hot desert” [according to Stewart and Oke (2009a)], dominated by loess soil. The rural station is surrounded by isolated low buildings of one–two floors with spacing of ∼100 m between them and an SVF of >0.95. These features have not been changed along the study period.

Beer Sheba is located near the border between the semiarid (BS) and the arid (BW) climate regions according to the Köppen classification (Goldreich 2003, 12–14), having an average annual rainfall of 200 mm, confined to the winter season (Bitan and Rubin 1994). Table 1 presents the average daily minimum and maximum temperatures for the summer and the winter in the urban and the rural stations. The diurnal range is 13.2°C in the summer and 10.6°C in the winter, and the seasonal range is 12.8°C.

According to Lowry (1977) the temperature differences between urban and neighboring rural locations are affected by three factors: the regional “background” climate, the local landscape, and the effects of urbanization. For our study region, the regional factor is the arid climatic conditions described above. The regional effect is subjected to the warming trend, mainly in the summer (Saaroni et al. 2010), but we assume that it is equal for both stations and so does not affect ΔTur. The landscape factor is composed of two elements. One is the higher elevation of the urban station with respect to the rural, implying lower temperatures in the former, on the order of 0.5°C, especially in the day hours, when the planetary boundary layer is well mixed. The second is the concave topography of the city, in contrast to the flat landscape of the rural region (Fig. 1c), implying enhanced nocturnal cooling (and daytime warming) in the urban location. These two topographic effects are expected to be constant during the study period. The cooling imparted by the concave topography of the city opposes the heating imparted by the urban effect during the night hours, and both intensify when the winds are weak.

3. Materials and method

The study period starts in 1967, when the rural station began to perform measurements, and ends in 2004. The location of the urban station has not been changed during the study period. The rural station was moved about 500 m southeast in 1977 but remained at the same altitude and posting. Both stations have been operated according to the WMO regulations for climatological stations (although they shifted from manual to automatic instrumentation in 2000) and are controlled by the Israeli Meteorological Service.

The analysis is divided into the two main seasons: the summer, June–August (JJA), and the winter, December–March [DJFM, as defined by Alpert et al. (2004)]. The observational data for the two stations include the daily minimum and maximum temperatures and the wind at 0000 and 1200 UTC, corresponding to 0200 and 1400 LST, which represent the approximate times of minimum and maximum temperature, respectively.

Because the magnitude of the UHI cannot be resolved directly from the ΔTur because of topographical differences between the two stations, we used the temporal variation of ΔTur (ΔTT, hereinafter) for deriving the evolution rate of the UHI and for reconstructing its magnitude. The ΔTT was derived separately for the minimum and maximum temperatures and for the summer and winter seasons. The population of cities is commonly used as a measure for the magnitude of UHI (Oke 1987, 252–302), although UHI has no direct physical relationship with the number of inhabitants in a city. Nevertheless, the population serves as a substitute for urban form [e.g., SVF and city geometry; Oke (1981)] and magnitude of anthropogenic activity. Following the above arguments, the ΔTT is analyzed here with respect to the population growth. To reconstruct the preurban temperature difference between the urban and the rural locations, we extrapolated the dependence of ΔTur back to zero population (ΔTp=0, hereinafter). Then, the difference between the present day and ΔTp=0 yields the estimated net UHI. Our approach for estimating the UHI is based on the assumption that the regional climate changes do not alter ΔTur for any population size (and, therefore, ΔTT) and therefore it depends on the urban features alone.

4. Results

The interannual variation of ΔTur, calculated for the minimum and maximum temperature, is shown in Figs. 2a and 2b for the summer and the winter, respectively. In both seasons an increasing trend was found for the minimum and the maximum temperatures, with the largest ΔTT for the minimum temperatures in the summer and the winter, 0.46° and 0.26°C (10 yr)−1, respectively. For the maximum temperatures the values are also positive but lower, 0.17° and 0.20°C (10 yr)−1 for the summer and the winter, respectively. The interannual variations show several irregularities, especially in the early 1990s, when ΔTur for the minimum temperature rose sharply in both seasons and dropped for the maximum temperature, especially in the summer. This could be explained by a change in the location or immediate surrounding of one of the stations, but such a change did not occur in any of the stations during that period. We, therefore, suggest that these changes result from a considerable population growth in the city of Beer Sheba of 35% within 5 yr (1990–95) due to the emigration wave from Russia [from the Statistical Abstract of Israel 2008 (No. 59) of the Central Bureau of Statistics; obtained online at http://www1.cbs.gov.il/reader/shnatonenew.htm (URL displays current version)], accompanied by intensive construction activity there. The construction activity most likely caused increased dust loading over the city (no evidence is available for that), producing a “blanket effect,” that is, blocking of both solar and terrestrial radiation, which decreases the maximum temperatures and increases the minimum temperatures in the urban station. The larger decrease in ΔTur for the maximum temperature in the summer supports this hypothesis because in the summer the solar radiation is stronger and more persistent than in the winter as a result of the absence of cloudiness in the daytime during that season.

The use of ΔTT as a function of time does not allow reconstruction of ΔTur in the preurban period because the rural station records do not go farther back than 1967. Moreover, the urbanization rate may vary irregularly with time, making simple functional dependence on time problematic. This problem is circumvented by the use of population as a surrogate for urbanity and as an argument for the UHI magnitude. Oke (1987, 252–302) proposed a logarithmic dependence for calm conditions and a quadruple dependence (population0.25) when the wind appears as an additional independent variable. We plotted ΔTur as a function of population and examined various functional relationships between them (quadruple, logarithmic, linear, polynomial, and exponential). The highest correlation between observed and calculated ΔTur was found for the linear relationship and thus is used for our analysis. Note that convex functional relationships (such as logarithmic and quadruple) yield a considerable UHI even for small populations, in agreement with Landsberg (1970) and Oke (1987, 252–302). Therefore, our linear assumption may underestimate this effect.

The UHI is best exhibited under calm conditions (Oke 1987, 252–302). Therefore, the analysis was done separately for reduced samples that exclude observations in which the wind speed was >3 and >2 m s−1. For a 3 m s−1 threshold, the smallest size was obtained for the daytime in the summer (271; 8% of the full sample), but for the 2 m s−1 threshold only 91 observations (<3%) remained for the summer days. Nevertheless, no considerable change was found in the derived parameters while lowering the thresholds. The results for the full sample and the reduced sample of ≤3 m s−1 are specified in Table 2. The ΔTT was found to be positive for both times of the day and seasons. It is considerably higher during the nighttime hours: about 4 times that in the day hours in the summer and 1.4 times in the winter. The smallest rate (for the reduced sample) is 0.42°C per 100 000 population for the summer maximum, and the largest rate is 1.66°C per 100 000 population for the summer minimum. The results for the full sample are similar (Table 2). The pronounced positive UHI in the night hours is consistent with previous studies, but the significant positive UHI in the day hours in both seasons is not trivial. Daytime “cool islands” have been found in several cities, resulting from the urban geometry, that is, street canyons or vegetation (i.e., parks and green space effects; e.g., Hafner and Kidder 1999; Shigeta et al. 2009; Mendonca 2009). Moreover, settlements in arid regions tend to develop “oasis effects” expressed in cooling during daytime (Saaroni et al. 2004), which has also been noted in several arid cities, such as in Botswana (Jonsson 2004).

The temperature difference for zero population, ΔTp=0, when subtracted from the present-day ΔTur, is considered here as the estimated net UHI. It is demonstrated in Fig. 3 by the temperature difference (seen on the y axis) between the intersections of the straight line approximating ΔTur with population = 0 and with population = 185 000 (x axis). The values (reduced sample) for the nights are +3.07°C for the summer and +2.11°C for the winter, and for the days they are +0.78°C for the summer and +1.54°C for the winter (Table 2). The values of ΔTp=0, implied by Figs. 3a and 3b and specified in Table 2, indicate that during the nighttime hours the temperatures at the urban location in its preurban state were lower than at the rural location by over 2°C in both summer and winter, presumably due to its concave landscape. Note that for the study period (when the average population was on the order of 100 000) the city was still cooler on average than the rural location, by 0.8°C in the summer and 1.0°C in the winter (see Table 1). For the maximum temperatures, ΔTp=0 is +0.29°C for the summer and −0.91°C for the winter. The average ΔTur for the study period (Table 1) shows that the city was hotter than the rural location by 0.9°C in the summer, whereas in the winter it had the same average temperature. This reflects a delicate balance between two opposing effects: the concavity of Beer Sheba, implying higher temperatures there, and its higher elevation, implying its being cooler. This implies that during the day hours the concavity of Beer Sheba dominates in the summer whereas the higher elevation of Beer Sheba dominates in the winter. The negligible contribution of the concavity of Beer Sheba for the winter days may be attributed to the weak insolation in that season.

Note that when the linear relationship is adopted, still, the magnitude of the net UHI is sensitive to the derived slope of ΔTur. However, the correlations between the observed and the estimated ΔTur were found to be high and significant (at the 0.05 level; see Table 2) for each time and season. To assess further the uncertainty in the slope of ΔTur the estimated slope was recalculated for four subsamples for each time and season. In each subsample, five observations were subtracted randomly. The resulting deviations in the estimated net UHI did not exceed 0.13° and 0.11°C for the summer and winter nights, respectively, and 0.39° and 0.26°C for the summer and winter days, respectively.

5. Summary and conclusions

This study proposes a method for estimating the canopy-layer net UHI in a region with complex terrain in which observations from the “preurban” period are not available. The approach is based on a linear relationship between ΔTur and city population, which was found to have the highest correlation with the observations. First, the linear relationship is extracted for the study period for which observations are available. Then, it is extrapolated to zero population to yield the desired preurban value, ΔTp=0. The difference between ΔTp=0 and the current ΔTur is proposed to reflect the net UHI.

The UHI was evaluated for an arid city (Beer Sheba) for 1967–2004, during which period the population grew from 70 000 to 185 000. It was derived for the minimum and the maximum temperatures for the summer and the winter months separately. Because the UHI is best exhibited under calm conditions (Oke 1987, 252–302), the analysis was done also for reduced samples, excluding observations for which the wind speed was >3 and >2 m s−1. The net UHI found for the minimum temperature are on the order of +3° and +2°C for the summer and the winter, respectively. These values are comparable with those found by Comrie (2000) and Hawkins et al. (2004) for the arid cities of Arizona. The high positive UHI in the night hours is in line with previous studies. The large values in the summer nights relative to those in the winter can be attributed to the stable conditions that prevail in Israel during that season, when a marine inversion predominates (Dayan and Rodnizki 1999; Dayan et al. 2002; Ziv et al. 2004). For the maximum temperatures, the values are also positive: +0.8°C for the summer and +1.5°C for the winter. The positive UHI in the day hours is suggested to express the growth of the urban area and the increase in anthropogenic activity (e.g., energy consumption, transportation, and air conditioning), which sets off daytime cooling effects such as the oasis effect imparted by municipal parks (Oke 1987, 252–302; Saaroni et al. 2004) or the cool islands often found in large cities (e.g., Saaroni et al. 2000).

Given the uncertainties of the population method, the relatively short time period of the temperature record, and the possible inhomogeneity in the data, the results should be regarded as a first-order approximation of the net UHI contribution and as the best estimates that can be offered in view of the data and method restrictions. Note also that the method proposed here cannot capture abrupt changes (e.g., massive construction projects) in the cities’ development, such as that which happened in the early 1990s when a wave of immigrants came to Beer Sheba.

The positive UHI in the summer implies further aggravation of heat stress in this arid, hot region beyond that occurring, and that expected to proceed, over the region. The linear temperature trends in the summer season are significantly positive. For the maximum temperature, it is +0.31°C (10 yr)−1 at the rural station and +0.48°C (10 yr)−1 at the urban station. For the minimum temperature the warming trends are even higher, +0.33°C (10 yr)−1 and +0.80°C (10 yr)−1 for the rural and the urban stations, respectively. Further rapid increase in the minimum temperature of the summer in the urban area will bring about heat-stress conditions even during the nighttime hours and will increase the probability for continuous periods with warm nights. Meehl and Tebaldi (2004) incorporated in their heat-wave definition the nighttime temperatures, indicating its crucial contribution for uncomfortable conditions. In addition, Karl and Knight (1997) stated that sequences of consecutive nights with no relief that result from very warm nighttime minimum temperatures may be most important in terms of health impacts. It is imperative that any discussion of global warming and its regional and environmental implications should not discard the further warming imparted by the urban heat island.

Acknowledgments

The authors thank the Israel Meteorological Service for providing the data. Special thanks are given to Na’ama Eisenbruch, Eran Beja, and Dr. Orna Zafrir-Reuven.

REFERENCES

  • Alpert, P., , I. Osetinsky, , B. Ziv, , and H. Shafir, 2004: A new seasons definition based on classified daily synoptic systems: An example for the eastern Mediterranean. Int. J. Climatol., 24 , 10131021.

    • Search Google Scholar
    • Export Citation
  • Bitan, A., , and S. Rubin, 1994: Climatic Atlas of Israel for Physical and Environmental Planning and Design. Ramot Publishing, 358 pp.

  • Brazdil, R., , and M. Budıkova, 1999: An urban bias in air temperature fluctuations at the Klementinum, Prague, the Czech Republic. Atmos. Environ., 33 , 42114217.

    • Search Google Scholar
    • Export Citation
  • Comrie, A. C., 2000: Mapping a wind-modified urban heat island in Tucson, Arizona. Bull. Amer. Meteor. Soc., 81 , 24172431.

  • Dayan, U., , and J. Rodnizki, 1999: The temporal behavior of the atmospheric boundary layer in Israel. J. Appl. Meteor., 38 , 830836.

  • Dayan, U., , B. Lifshitz-Goldreich, , and K. Pick, 2002: Spatial and structural variation of the atmospheric boundary layer during summer in Israel—Profiler and rawinsonde measurements. J. Appl. Meteor., 41 , 447457.

    • Search Google Scholar
    • Export Citation
  • Goldreich, Y., 2003: The Climate of Israel: Observation, Research and Application. Kluwer Academic/Plenum, 298 pp.

  • Hafner, J., , and S. Q. Kidder, 1999: Urban heat island modeling in conjunction with satellite-derived surface/soil parameters. J. Appl. Meteor., 38 , 448465.

    • Search Google Scholar
    • Export Citation
  • Hawkins, T. W., , A. J. Brazel, , W. L. Stefanov, , W. Bigler, , and E. L. Saffell, 2004: The role of rural variability in urban heat island determination for Phoenix, Arizona. J. Appl. Meteor., 43 , 476486.

    • Search Google Scholar
    • Export Citation
  • Jonsson, P., 2004: Vegetation as an urban climate control in the subtropical city of Gaborone, Botswana. Int. J. Climatol., 24 , 13071322.

    • Search Google Scholar
    • Export Citation
  • Karl, T. R., , and R. W. Knight, 1997: The 1995 Chicago heat wave: How likely is a recurrence? Bull. Amer. Meteor. Soc., 78 , 11071119.

  • Kim, Y. H., , and J. J. Baik, 2002: Maximum urban heat island in Seoul. J. Appl. Meteor., 41 , 651659.

  • Landsberg, H. E., 1970: Micrometeorological temperature differentiation through urbanization. Urban Climates, WMO Publ. 254, TP 141, World Meteorological Organization, 129–136.

    • Search Google Scholar
    • Export Citation
  • Lowry, W. P., 1977: Empirical estimation of the urban effects on climate: A problem analysis. J. Appl. Meteor., 16 , 129135.

  • Martinez, A., , C. Yague, , and E. Zurita, 1991: Statistical analysis of the Madrid urban heat island. Atmos. Environ., 25 , 327332.

  • Meehl, G., , and C. Tebaldi, 2004: More intense, more frequent, and longer lasting heat waves in the 21st century. Science, 305 , 994997.

    • Search Google Scholar
    • Export Citation
  • Mendonca, F., 2009: Urban heat and urban cool islands: Influences of vegetation and soil surface in some cities, southern Brazil. Proc. Seventh Symp. on Urban Climate, Yokohama, Japan, International Association for Urban Climate, P1-39.

    • Search Google Scholar
    • Export Citation
  • Morris, C. J. G., , I. Simmonds, , and N. Plummer, 2001: Quantification of the influences of wind and cloud on the nocturnal urban heat island of a large city. J. Appl. Meteor., 40 , 169182.

    • Search Google Scholar
    • Export Citation
  • Nasrallah, H. A., , A. J. Brazel, , and R. C. Balling, 1990: Analysis of the Kuwait city urban heat island. Int. J. Climatol., 10 , 401405. doi:10.1002/joc.3370100407.

    • Search Google Scholar
    • Export Citation
  • Oke, T. R., 1976: The distinction between canopy and boundary-layer urban heat islands. Atmosphere, 14 , 269277.

  • Oke, T. R., 1981: Canyon geometry and the nocturnal urban heat island: Comparison of scale model and field observations. J. Climatol., 1 , 237254.

    • Search Google Scholar
    • Export Citation
  • Oke, T. R., 1982: The energetic basis of the urban heat island. Quart. J. Roy. Meteor. Soc., 108 , 128.

  • Oke, T. R., 1987: Boundary Layer Climates. Methuen Press, 435 pp.

  • Oke, T. R., 2004: Initial guidance to obtain representative meteorological observations at urban sites. IOM Rep. 81, World Meteorological Organization, 47 pp. [Available online at http://www.wmo.int/pages/prog/www/IMOP/publications/IOM-81/IOM-81-UrbanMetObs.pdf].

    • Search Google Scholar
    • Export Citation
  • Potchter, O., , Y. Yaacov, , and G. Oren, 2006: The magnitude of the urban heat island of a city in an arid zone: The case of Beer Sheva, Israel. Proc. Sixth Int. Conf. on Urban Climate, Gothenburg, Sweden, International Association for Urban Climate, 450–453.

    • Search Google Scholar
    • Export Citation
  • Roth, M., 2007: Review of urban climate research in (sub)tropical regions. Int. J. Climatol., 27 , 18591873.

  • Saaroni, H., , E. Ben-Dor, , A. Bitan, , and O. Potchter, 2000: Spatial distribution and microscale characteristics of the urban heat island in Tel-Aviv, Israel. Landsc. Urban Plan., 48 , 118.

    • Search Google Scholar
    • Export Citation
  • Saaroni, H., , B. Ziv, , J. Edelson, , and P. Alpert, 2003: Long-term variations in summer temperatures over the eastern Mediterranean. Geophys. Res. Lett., 30 , 1946. doi:10.1029/2003GL017742.

    • Search Google Scholar
    • Export Citation
  • Saaroni, H., , A. Bitan, , E. Ben-Dor, , and N. Feller, 2004: The mixed results concerning the ‘oasis effect’ in a rural settlement in the Negev Desert, Israel. J. Arid Environ., 58 , 235248. doi:10.1016/j.jaridenv.2004.08.010.

    • Search Google Scholar
    • Export Citation
  • Saaroni, H., , B. Ziv, , I. Osetinski, , and P. Alpert, 2010: Factors governing the interannual variation and the long-term trend of the 850 hPa temperature over Israel. Quart. J. Roy. Meteor. Soc., 136 , 305318. doi:10.1002/qj.580.

    • Search Google Scholar
    • Export Citation
  • Shigeta, Y., , Y. Ohashi, , and O. Tsukamoto, 2009: Urban cool island in daytime—Analysis by using thermal image and air temperature measurements. Proc. Seventh Symp. on Urban Climate, Yokohama, Japan, International Association for Urban Climate, P5-7.

    • Search Google Scholar
    • Export Citation
  • Sofer, M., , and O. Potchter, 2006: The urban heat island of a city in an arid zone: The case of Eilat, Israel. Theor. Appl. Climatol., 85 , 8188.

    • Search Google Scholar
    • Export Citation
  • Stewart, I. D., , and T. R. Oke, 2009a: A new classification system for urban climate sites. Bull. Amer. Meteor. Soc., 90 , 922923.

  • Stewart, I. D., , and T. R. Oke, 2009b: Newly developed “thermal climate zones” for defining and measuring urban heat island magnitude in the canopy layer. Preprints, T. R. Oke Symp. and Eighth Symp. on Urban Environment, Phoenix, AZ, Amer. Meteor. Soc., J8.A. [Available online at http://ams.confex.com/ams/pdfpapers/150476.pdf].

    • Search Google Scholar
    • Export Citation
  • Unger, J., 1996: Heat island intensity with different meteorological conditions in a medium-sized town: Szeged, Hungary. Theor. Appl. Climatol., 54 , 147151.

    • Search Google Scholar
    • Export Citation
  • Unger, J., , Z. Sümeghy, , A. Gulyas, , Z. Bottyan, , and L. Mucsi, 2001: Land-use and meteorological aspects of the urban heat island. Meteor. Appl., 8 , 189194.

    • Search Google Scholar
    • Export Citation
  • Ziv, B., , H. Saaroni, , and P. Alpert, 2004: The factors governing the summer regime of the eastern Mediterranean. Int. J. Climatol., 24 , 18591871.

    • Search Google Scholar
    • Export Citation
  • Ziv, B., , H. Saaroni, , A. Baharad, , D. Yekutieli, , and P. Alpert, 2005: The indications for aggravation in summer heat conditions over the Mediterranean Basin. Geophys. Res. Lett., 32 , L12706. doi:10.1029/2005GL022796.

    • Search Google Scholar
    • Export Citation

Fig. 1.
Fig. 1.

(a) The location of the study region. (b) Aerial photograph of the study region with notations of the rural and urban stations. (c) Topographical cross sections from north to south across the rural (black line) and the urban (gray line) stations.

Citation: Journal of Applied Meteorology and Climatology 49, 10; 10.1175/2010JAMC2473.1

Fig. 2.
Fig. 2.

Interannual variation of ΔTur (°C) for the period 1967–2004 and the long-term linear trend for the minimum and maximum temperatures during the (a) summer and (b) winter.

Citation: Journal of Applied Meteorology and Climatology 49, 10; 10.1175/2010JAMC2473.1

Fig. 3.
Fig. 3.

Variation of ΔTur (°C) for the period 1967–2004 as a function of Beer Sheba population and the corresponding linear relationship, for the minimum and maximum temperatures, for the (a) summer and (b) winter. The dashed extensions of the linear lines are extrapolated to zero population.

Citation: Journal of Applied Meteorology and Climatology 49, 10; 10.1175/2010JAMC2473.1

Table 1.

Average minimum and maximum seasonal temperatures (°C) in the urban (Beer Sheba) and the rural stations for the summer and the winter (1967–2004).

Table 1.
Table 2.

Measures of the UHI in Beer Sheba (as a function of population) for the minimum and maximum temperature for summer and winter seasons: the slope of ΔTur (degrees Celsius per 100 000 residents) calculated by linear regression, the correlation between the observed and calculated ΔTur, the extrapolated ΔTur to zero population p, and the calculated net UHI. The measures and the number of days included are shown separately for the full sample and for the days with wind speed ≤3 m s−1 (boldface).

Table 2.
Save