Characteristics of urban heat island effect in Lhasa City

Zhuo Ga , YunDan NiMa , Jian Jun , PuBu CiRen

1. Tibet Institute of Plateau Atmospheric and Environmental Sciences, Lhasa, Tibet 850000, China

2. Meteorological Bureau of Nagqu, Nagqu, Tibet 852000, China

3. Department of Geography, University of Colorado, Boulder, Colorado 80309, U.S.A

4. Meteorological Bureau of Shannan, Tsetang, Tibet 856000, China

*Correspondence to: Dr. Zhuo Ga, Research fellow of Tibet Institute of Plateau Atmospheric and Environmental Sciences, Tibet Meteorological Bureau. No. 2, North LinKuo Road, Lhasa, Tibet 850000, China. Tel: 86-891-6336344; Email:zhuoga2000@yahoo.com.cn

Characteristics of urban heat island effect in Lhasa City

Zhuo Ga1*, YunDan NiMa2,3, Jian Jun4, PuBu CiRen1

1. Tibet Institute of Plateau Atmospheric and Environmental Sciences, Lhasa, Tibet 850000, China

2. Meteorological Bureau of Nagqu, Nagqu, Tibet 852000, China

3. Department of Geography, University of Colorado, Boulder, Colorado 80309, U.S.A

4. Meteorological Bureau of Shannan, Tsetang, Tibet 856000, China

*Correspondence to: Dr. Zhuo Ga, Research fellow of Tibet Institute of Plateau Atmospheric and Environmental Sciences, Tibet Meteorological Bureau. No. 2, North LinKuo Road, Lhasa, Tibet 850000, China. Tel: 86-891-6336344; Email:zhuoga2000@yahoo.com.cn

This paper analyzes the Urban Heat Island (UHI) effect in Lhasa City of Tibet using meteorological observations, the Normalized Difference Vegetation Index (NDVI), Land Surface Temperature (LST) data obtained from Earth Observing System/Moderate resolution Imaging Spectroradiometer (EOS/MODIS) information, and correlation and composite analyses. The results show: (1)Areas with high temperature are primarily located in the center of the city or nearby counties, while low temperature areas are in the suburbs of counties. The area with high temperature has expanded in recent years and some high-temperature centers have even migrated to certain other regions. (2) The UHI intensity tends to be stronger both in annual and seasonal variations, especially in winter. Also, LST is somewhat positively related to mean air temperature. (3) A negative correlation exists between the changes of LST and NDVI with the increase of vegetation from urban to rural areas in different seasons. (4) The UHI intensity is negatively correlated with precipitation while positively correlated with wind speed, and the relation between the UHI intensity and evaporation varies with the seasons, namely, the intensity is positively correlated with summer evaporation but negatively correlated with winter evaporation. (5) UHI intensity might be enhanced by intensified urbanization, wherein built-up areas expand,there is increased heat from human activity, and there is more artificial heat input to the atmosphere.

Lhasa; urban heat island effect; characteristics; meteorological conditions

1. Introduction

It is well known that the Urban Heat Island (UHI) effect refers to the phenomenon of air temperature of urban areas being higher than that of their surrounding areas, which is caused by human activities (Zhou and Shu, 1994; Sun and Lu, 2002). In 1833, British chemist Lake Howard first found this effect after he compared the temperature of downtown London to that of its suburbs (Howard, 1833). Since then,environmental scientists from various countries have conducted extensive research to explore the impact of the UHI effect and have obtained many insightful findings.

Studies on UHI in China mainly focus on its temporal and spatial distribution, and its formation from both observational and numerical simulation aspects. Specifically, most of the previous investigations have primarily focused on the temporal and spatial distribution of UHI in Beijing. Many researches (Zhanget al., 2002a; Lin and Yu, 2005; Yuet al.,2005; Liet al., 2006, 2008; Wanget al., 2006b; Wanget al.,2006c) have been conducted on the spatial characteristics of UHI, the relationship between its intensity and urbanization index, variation of scale range, meteorological variables, and so forth. Furthermore, Zhenget al. (2006) examined the multi-center and multi-scale structure of UHI in Beijing. In addition, quite a few studies have been conducted to investigate the UHI effect in Shanghai (Qi, 2004; Zhuet al., 2006;Sunet al., 2007; Tanet al., 2008; Zhouet al., 2008a) and other cities (Zhanget al., 2002b; Heet al., 2003; Baiet al.,2005; Liuet al., 2005; Hanet al., 2007; Haoet al., 2007;Jianget al., 2007; Dinget al., 2008; Liuet al., 2008; Qiuetal., 2008; Liet al., 2009; Ronget al., 2009).

With the development of satellite and remote sensing technologies, many works (Zhanget al., 2003; Yan and Deng, 2004; Gonget al., 2005; Liet al., 2005; Zhanget al.,2005; Jianget al., 2006; Wanget al., 2006a; Zhanget al.,2006; Zhaoet al., 2006; Lv and Gong, 2007; Penget al.,2007; Wanget al., 2007; Caoet al., 2008; Wanget al., 2009;Zhu and Zhu, 2009) have aimed to analyze the general distribution of UHI; quantitatively evaluate its effects using various satellite data; identify its influencing factors; and examine its parameters based on satellite images to efficiently reveal its variation in different seasons.

In recent years, numerical simulation has gradually become an important technology in UHI research. A number of studies (Tong and Sang, 2002; Yanget al., 2003; Chenet al.,2004; Qi, 2004; Liet al., 2007; Lin and Yang, 2007; Sunet al., 2007; Tanet al., 2008; Zhouet al., 2008b) simulated the structure of wind and temperature, and pollutant concentration and its effect on the atmospheric boundary layer in Beijing. Regarding the UHI effect, they analyzed its influencing factors, its formation mechanism, and possible improvement approaches. Zhouet al. (2008b) coupled a single-layer canopy model with a forecast model of an urban boundary layer to indicate the common function of the urban canopy structure and manmade heating sources for UHI. Some research works (Heet al., 2008; Hanet al., 2009) have presented the impacts of the UHI effect on thermal structure and pollutants diffusion.

This paper aims to investigate the inter-annual and seasonal variations of UHI, and the temporal and spatial distribution of the thermal field in Lhasa City. Section 2 presents the data, methods, and study area. In Section 3, we analyze the distribution of UHI intensity and LST on the basis of meteorological observations and satellite data, and then compare LST with air temperature. The influencing factors of UHI are discussed in Section 4. Conclusions and future prospects are summarized in Section 5.

2. Data sources, methods, and investigation region

2.1. Data sources

Both LST and NDVI datasets used in the current research were obtained from the Land Process Distributed Active Archive Center (LP DAAC, https://lpdaac.usgs.gov/main.asp), derived from 1-km MODIS data. The LST data are available at eight-day intervals, while NDVI data are produced monthly. These datasets not only have higher temporal and spatial resolution, but also are widely used in land ecosystem and global climate change research. In this study, we chose monthly distributions of the thermal field to represent the characteristics of the different seasons in 2001,2004, and 2007 (i.e., January, April, June, and October respectively represent spring, summer, fall, and winter). In addition, meteorological variables used here include mean air temperature, wind speed, precipitation, and evaporation data acquired on a monthly time scale at three stations(Lhasa, MoZhugongka, and Nimu) from 1978 to 2007. Socioeconomic data are from the statistical almanac of the Tibet Autonomous Region.

2.2. Methods

The composition and reprojection of LST and NDVI were realized by the MODIS Reprojection Tool (MRT) and the Environment for Visualizing Images (ENVI), which is provided by LP DAAC. First, we matched the downloaded data to the Lhasa area, saving the MODIS data in a GEOTIF format. We then reprojected the images onto standard latitude/longitude coordinates in ENVI, overlaying county boundaries in the Lhasa area. Finally, we redeposited the above data in an ASCII format for further calculations.

2.3. Brief introduction to the study area

The study area, shown in figure 1, is situated in the northern part of middle reaches of the Yarlung Zangbo River and its tributary stream, the Lhasa River (Chu, 2007). The geographic location is 29°14′26′′–31°03′45′′N,89°45′9′′–92°37′16′′E, with a total land area of 29,528.91 km2, including one district (Chengguanqu) and seven counties (Nimu, Dangxiong, Duilongdeqing, Qushui, Linzhou,Dazi, and Mozhugongka). The topographic character is diversified by hills and mountains, with the western and northern parts being higher than the eastern and southern parts. The main climate in Lhasa are characterized by strong radiation, long sunshine duration, large daily range of temperature, and high probability of night rain since it belongs to the plateau temperate and semi-arid monsoon climate region.

Figure 1 Spatial distribution of different counties over the Lhasa area

3. Distribution of UHI and LST

3.1. UHI Intensity

UHI intensity is defined by Oke (1995) in his observational study of land surface meteorology, as the difference between air temperature in urban and rural areas at the same observational time and height (usually 1.5 m above the surface). Earlier, Chandler (1965) had proposed that UHI intensity was the difference between the mean air temperature of urban and rural meteorological stations. In recent years, due to the lack of observations in urban areas, scientists have also used the difference between air temperature from one typical station in an urban area and one in a rural area to represent the variation of UHI intensity (Lee, 1979).

In this paper, because the altitude of Dangxiong is much higher than that of Lhasa City, its air temperature was not suitable to use in UHI study. Therefore, we chose observations in Chengguanqu to represent urban temperatures, and the mean values of air temperature at Mozhugongka and Nimu as rural temperatures. We considered their difference to be UHI intensity in Lhasa in order to calculate its annual,seasonal and inter-decadal variations (first decade:1978–1987, second decade: 1988–1997, third decade:1998–2007) (Figure 2) in different seasons, together with air temperature data from 1978 to 2007.

Figure 2a shows that the annual UHI intensity gradually increases with evident variations from 1978 to 1990; it appears stable after 1990, then rapidly intensifies after 1997.The mean multi-year UHI intensity is 1.74 °C and the maximum value appeared in 2007 (2.22 °C), while the minimum occurred in 1984 (1.31 °C). Seasonal variation shows a maximum value of 2.48 °C in winter followed by that of spring, while the effects in fall and summer are weaker, especially given that summer’s minimum value is only 0.98 °C. The mean multi-year UHI intensities for the above seasons are 1.98 °C, 1.75 °C, 1.62 °C, and 1.60 °C,respectively.

As shown in figure 2b, we calculated the mean value of each decade (10 years) from 1978 to 2007 in various seasons,along with the difference of adjacent two decades to understand the inter-decadal changes in UHI intensity. Our results show that the UHI intensity increases especially rapidly in winter. The differences in spring and summer between the second decade and the first decade are larger than those in other seasons, with values of 0.139 °C and 0.372 °C, respectively. However, the differences in annual mean, autumn and winter, between the third decade and the second decade are larger than those in other seasons, with values of 0.294 °C,0.502 °C, and 0.399 °C, respectively. In general, autumn’s increasing ratio of UHI intensity has been the largest during the most recent 10 years.

Figure 2 Temporal variation of UHI intensity in different seasons over the Lhasa area (unit: °C);

Because of cold winters in Lhasa, people there have to use firewood and coal for heat. The combustion of coal not only increases manmade heat input to atmosphere, but also produces a great quantity of greenhouse gases, including CO2, which leads to higher temperatures in winter and causes the UHI effect to be more obvious than in other seasons.

3.2. Land surface temperature

LST images (Figure 3) show that, in general, the highest temperature region is essentially located in the center of the city or nearby counties, and its maximum temperature is above 30 °C. The lowest temperatures mainly occur in the rural areas, which are sparsely populated and often have water as their underlying surface; temperatures there are lower than 10 °C or even sometimes 0 °C. Thermal field intensity has basically been getting stronger with the region’s economic development from 2001 to 2007. However,the temperature is observed to be lower for some parts of certain counties when comparing LST in 2004 with that in 2001. The mainly increasing trend of LST is not only reflected by rising temperatures per se, but is also evidenced by the expanding areas of high temperatures. In addition,LST images in different months demonstrate that some high-temperature centers have migrated to other regions.

On the whole, the areas with high temperature have gradually expanded over Lhasa in the past few years, which is closely related to the development of the local economy and an increase in the number of high-rise buildings in this region (underlying surfaces with complicated configurations are favorable to the formation of UHI). As the economic activity of some regions has intensified, high-rise buildings and human activities increased, which may have led to the shift of high-temperature centers in Lhasa.

Figure 3 Spatial distribution of LST in year 2001, 2004, and 2007 (unit: °C);

3.3. Comparison of LST to air temperature

UHI is usually indicated as the outcome of the interaction between urban underlying surfaces and weather conditions, which in turn has a close relation with the thermal field of urban surfaces. Nevertheless, there is still an obvious difference between thermal field and mean air temperature.In general, the LST from satellite data correlates well to air temperature. Quite a few studies have shown that the major factors influencing LST include received radiation and the thermal characteristics of ground features, whereas air temperature is determined by the sensible heat of underlying surfaces, atmospheric turbulence, and manmade heat. In order to determine the relationship between LST and air temperature in Lhasa City, we used monthly LST and mean temperature of Lhasa, Nimu, and Mozhugongka stations to respectively represent thermal field and meteorological conditions. As shown in figure 4, a somewhat positive relationship appears between LST and air temperature, indicating a wide application potential to use LST in studying the UHI effect over cities of the Tibetan Plateau, where there is a lack of meteorological observations.

Besides weather conditions, Liu (2008, personal communication) found that the main factors affecting the relationship between LST and air temperature include: (1) LST obtained from satellite data is the average luminance temperature for each mixed pixel, whereas the air temperature from meteorological stations is the temperature of one point.(2) Geometric correction error of images might be within one pixel, but the spatial difference of luminance temperatures is quite large. (3) Different ground features contribute differently to the increase of temperature. In short, LST is well correlated to air temperature under stable weather conditions.

Figure 4 Comparison of LST with air temperature

4. Influencing factors

UHI is the result of interactions between underlying surfaces and local weather conditions, as described previously.In cities, the decrease of vegetation and water leads to the increase of sensible heat and the decrease of latent heat. Although the underlying surface is relatively stable in cities, its impact on the formation of UHI varies due to the variation of weather conditions. Specifically, any given city at any given time has a fairly uniform underlying surface factor,manmade heat, greenhouse gases, emission sources, and concentration of atmospheric pollutants, but its UHI may be quite unsteady and its intensity may change with meteorological conditions. Also, the emission of high air temperatures and water vapor to the atmosphere affect the intensity of UHI, from the result of many combustions and industrial activities. In the following sections we investigate factors that influence UHI, including vegetation, meteorological conditions, and social factors.

4.1. Vegetation index

As an index to reflect the existence and density of vegetatio n, NDVI can be used to monitor seasonal variations and changes in land use/land cover. With the index obtained from EOS/MODIS satellite data in April, 2007 as an example, the relationship between LST and NDVI in the Lhasa area has been analyzed as follows.

The distributions of LST (Figure 3a) and NDVI (Figure 5)suggest that the air temperature is lower in areas with dense vegetation, while higher in areas with sparse vegetation; this indicates a negative relationship between LST and vegetation. This suggests that vegetation plays an important role in the distributional proportion of latent heat and sensible heat (i.e., green plants are beneficial to the decrease of temperature). Also, the UHI effect is closely related to the range of vegetation areas, density, species, and growth rates.Furthermore, based on the central location of Chengguanqu,we examined vegetation distribution of cross sections along 29.6 °N and 91.0 °E, and found that the urban area has higher LST and small NDVI, while the rural area has lower LST and large NDVI.

Figure 5 Distribution of NDVI in the Lhasa area

Since the maximum NDVI could better express the coverage extent of seasonal vegetation growth, in order to investigate its variations in different regions we calculated the annual composite maximum NDVI in 2001, 2004, and 2007,and analyzed their different images among these years. Our results show that the vegetation cover increased remarkably from 2001 to 2004 in most of the Lhasa area except Nimu,especially in the eastern part of Lhasa, but decreased after 2004 in each county. With the negative relation between UHI intensity and NDVI, UHI intensity in 2004 was weaker than in 2001 (Figure 3), indicating that the distribution of vegetation has a great effect on the variations of UHI intensity.

4.2. Meteorological conditions

Since the UHI effect reflects the difference of air temperature between urban areas and rural areas, we used the difference between the Lhasa station data and the mean temperature of Mozhugongka, Nimu stations as analysis dataset to study the relationship between meteorological conditions and UHI intensity. The inter-annual variation of meteorological factors shows that wind speed in urban areas is slightly less than that in rural areas, while rural precipitation is greater than urban precipitation, and evaporation in urban areas is evidently greater than in rural areas. In general,evaporation is determined by many factors such as temperature, wind speed, and radiation, and temperature is the major influencing factor among them. Higher temperature and evaporation are observed in urban areas than in rural areas under the influence of UHI.

In a comparison of monthly data and the correlation coefficients between UHI intensity and meteorological factors(Figure 6), UHI intensity demonstrates a positive relationship with wind speed and negative one with precipitation(except in July). The relationship between the UHI intensity and evaporation is more complicated (i.e., evaporation has a positive relationship with the UHI intensity from April to October, and then it has negative relationship with the evaporation in the other months). Similar analysis indicated that UHI intensity has the same relationship with these meteorological factors in different seasons. However, some coefficients did not pass significance testing, such as those between UHI intensity and wind speed in spring and winter,those between the UHI intensity and precipitation for all seasons, and those between the UHI intensity and evaporation in spring and autumn.

UHI intensity is negatively related to precipitation in general. Increasing precipitation is beneficial to the decreasing of temperature in urban areas, which results in smaller differences between city and rural areas (i.e., decreases of UHI intensity). However, wind speed has the opposite effect on UHI intensity in terms of humidity reduction and rising temperatures in urban areas. Increase of evaporation has been found to contribute to the decrease of humidity and to the increase of temperature (i.e., increases of UHI intensity).However, since the difference of evaporation between urban and rural areas is the smallest in winter, it may have an opposite effect on UHI intensity. Therefore, the relationship between UHI intensity and evaporation varies by season,with positive and negative relationships in summer and in winter, respectively. This also indirectly confirms our conclusion that UHI intensity is strongest in winter and weakest in summer.

4.3. Social factors

Population is a significant index to indicate the growth of a city. With socioeconomic development, the number of permanent residents has increased rapidly in Lhasa City during the past few years; it increased from 4.03752 million in 2000 to 4.64736 million in 2007. In the cases of Chengguanqu and other counties, the numbers have also increased in various degrees. The population of Chengguanqu, which is located in the center of Lhasa City, increased from 1.4136 million in 2000 to 1.81991 million in 2007. Meanwhile, the number of tourists escalated quickly after 1995, reaching 6.08335 million in 2000 and 40.29438 million in 2007. Another important factor causing rapid population enhancement in Lhasa is that more people are coming to work in the city. Still another factor is that built-up areas in the city, including complete buildings and buildings under construction,and residence housing, have increased rapidly since 2000.

To sum up, population density and human activity are significantly increasing and built-up areas are expanding gradually along with the intensification of urbanization. It should be noted that discharges of smoke and industrial mill dust are decreasing with the rise of citizens’ desire for environmental protection, and treatment of industrial waste materials by relevant organizations in a timely manner. Gas emissions are held within limits but are increasing slightly.The increase of population and enhancement of living standards have also led to the multiplication of energy consumption. All of these factors, and the increasing number of cars and the increasing emission of manmade heat input to the atmosphere might cause the enhancement of UHI intensity.

Figure 6 Correlation coefficients between UHI intensity and meteorological factors in different months and seasons

5. Conclusions and future prospects

Using meteorological observations and LST data obtained from satellite EOS/MODIS, this paper discusses the UHI effect and its influencing factors in Lhasa City. Our primary conclusions can be summarized as follows:

(1) Areas with high temperature are primarily located in thecenter of the city or in nearby counties, while low-temperature areas are in the suburbs of counties. The area with high temperature has expanded in recent years and some high-temperature centers have even migrated to certain other regions.

(2) The UHI intensity tends to be stronger both in annual andseasonal variations, especially in winter. LST is somewhat positively related to mean air temperature.

(3) A negative correlation exists between the changes of LST and NDVI with the increasing of vegetationfrom urban to rural areas in different seasons.

(4) The UHI intensity is negatively correlated with precipitation, but positively correlated with wind speed. The relationship between UHI intensity and evaporation varies with the seasons, namely, UHI intensity is positively correlated with summer evaporation but negatively correlated with winter evaporation.

(5) UHI intensity might be enhanced by intensified urbanization, wherein built-up areas expand, there is increased heat from human activity, and there is more artificial heat input to the atmosphere.

This study did not examine diurnal variation of UHI, and the analysis here is not meticulous due to the limitations of the LST satellite dataset and meteorological observations. In addition, some research shows that increasing wind speed is beneficial to UHI formation to some extent, but if it exceeds certain magnitude it may cause UHI to weaken or even disappear. Therefore, there are still many questions needing further study. With the improvement of observational approaches, remote sensing, and numerical simulation technology, the study of the UHI effect in plateau cities will be more fruitful and will be based on diverse kinds of information, such as automatic weather station data, satellite products, land-use status and so forth.

The authors would like to thank the Land Process Distributed Active ArchiveCenter and the Information Center of the Tibet Meteorological Bureau for providing the data used in this study. The authors would like to express thanks to Dr.Liu Weidong for his insightful comments and suggestions.This work was jointly supported by programs of the Institute of Urban Meteorology (No. UMRF200705), the China Meteorological Administration (No. CMATG2010M24), and the Institute of Plateau Meteorology (No. BROP200705).

Bai HZ, Ren GY, Fang F, 2005. Characteristics of urban heat island effect and its influencing factors in Lanzhou. Meteorological Science and Technology, 33(6): 492–495.

Cao LQ, Zhang LP, Li PX, Huang W, 2008. Numerical simulation of variation of urban underlying surface types to heat island effect. Geomatics and Information Science of Wuhan University, 33(12): 1229–1232.

Chandler TJ, 1965. The Climate of London. Hutchinson, London.

Chen Y, Jiang WM, Wu J, Gu JQ, Xu JY, 2004. A numerical simulation on Hangzhou city heat island using regional boundary model. Plateau Meteorology, 23(4): 519–528.

Chu D, 2007. Study on Landuse/Land Cover over Mountainous Region—Using Lhasa Area as An Example. Meteorological Press, Beijing.

Ding SJ, Zhang JQ, Liu XP, Dong FL, 2008. Quantitative relationship between urban development and heat island effect for Harbin city. Advances in Climate Change, 4(4): 230–234.

Gong AD, Li J, Wang XD, Chen YH, Hu HL, 2005. Study on temporal and spatial distribution characteristics of the urban heat island in Beijing.Geography and Geo-information Science, 21(6): 15–18.

Han SQ, Guo J, Huang SL, Bian H, 2007. Study on the evolution of urban heat island in Tianjin city. Ecology and Environment, 16(2): 280–284.

Han SQ, Meng DM, Tong H, Li XJ, Li YH, 2009. Numerical simulation of urban heat island and its influence on pollutant diffusion. Ecology and Environment, 18(2): 403–407.

Hao LP, Fang ZF, Li ZL, Liu ZQ, He JH, 2007. The inter-annual climate change and heat island effect of Chengdu during the recent fifty years.Scientia Meteorologica Sinica, 27(6): 648–654.

He P, Li HB, Shu J, Lu XQ, Sun J, 2003. The analysis of the urban climate in Chuxiong of Yunnan: comparison of urban heat island effect in China.Acta Geographica Sinica, 58(5): 712–720.

He XF, Jiang WM, Liu HN, Zhou RW, 2008. A study of thermal effect of urban canopy structure on urban heat island with town energy balance scheme. Chinese Journal of Atmospheric Sciences, 32(6): 1445–1457.

Howard L, 1833. Climate of London deduced from meteorological observation. Harvey and Darton, 1(3): 1–24.

Jiang XD, Xia BC, Guo L, Li L, 2007. Characteristics of multi-scale temporal/spatial distribution of urban heat island in Guangzhou. Chinese Journal of Applied Ecology, 18(1): 133–139.

Jiang ZY, Chen YH, Li J, 2006. Heat island effect of Beijng based on Landsat TM data. Geomatics and Information Science of Wuhan University,31(2): 120–123.

Lee D, 1979. Contrast in warming and cooling rates at an urban and rural site.Weather, 24: 60–66.

Li JJ, He LH, Chen J, 2005. A study on urban thermal field of Nanjing based on satellite remote sensing. Resources and Environment in the Yangtze Basin, 14(6): 760–763.

LiLG, Jin W, Wang HB, Ma YJ, 2009. Analysis of urban heat island characteristics of city groups inmiddle Liaoning Province and their influencing factors. Journal of Anhui Agricultural Sciences, 37(5): 2113–2116.

LiXR, Hu F, Shu WJ, 2006. Characteristics and variations of urban heat island effect in Beijing for last 15 years. Meteorological monthly, 32(8):42–46.

LiXR, Hu F, Shu WJ, 2007. Analysis and numerical simulation of summer strong urban heat island in Beijing. Meteorological Monthly, 33(6):25–31.

LiXR, Hu F, Shu WJ, Liang BL, 2008. Characteristics of urban heat island effect and its meteorological influencing factors over Beijing in autumn.Climatic and Environmental Research, 13(3): 291–299.

LinBH, Yang DW, 2007. Numerical simulating of the heat island in Beijing.Advances in Water Science, 18(2): 258–263.

Lin XC, Yu SQ, 2005. Interdecadal changes of temperature in the Beijing region and its heat island effect. Chinese Journal of Geophysics, 48(1):39–45.

LinXC, Yu SQ, Tang GL, 2005. Study on relation of urbanization process and heatisland intensity. Advances in Natural Science, 15(7): 882–886.

Liu XF, Ruan X, Gu YL, 2005. Study of temperature change and heat island effect in Shijiazhuang area. Research of Environmental Sciences, 18(5):11–14.

LiuZN, Yin XJ, 2008. Analysis of urban heat island effect with its meteorological factors in Xian city. Journal of arid land Resources and environment, 22(2): 87–90.

LvGH, Gong L, 2007. Application of MODIS thermal infrared data in researching urban heat island in Urumqi. Arid Zone Research, 24(5):698–703.

Oke, TR, 1995. The heat island of the urban boundary layer: Characteristics,causes and effects. In: Cermak JE, Davenport AG, Plate EJ, Viegas DX(eds.). Wind climate in cities. Kluwer Academic. Dordrecht, 81–107.

Peng J, Liu WD, Long BJ, Hu JF, 2007. A study of the urban heat island of Beijing based on remote sensing images. Progress in Geophysics, 22(6):1942–1947.

QiLB, 2004. Trend analysis of midsummer high temperature and urban heat island effect in Shanghai over past 10 years. Meteorological Science and Technology, 32(6): 433–437.

QiuXF, Gu LH, Zeng Y, Jiang AJ, He YJ, 2008. Study on urban heat island effect of Nanjing. Climaticand Environmental Research, 13(6):807–814.

Rong CB, Liu HN, Zhu Y, 2009. Study on urban heat island effect and its influencingfactors over Suzhou in summer. Scientia Meteorologica Sinica, 29(1): 84–87.

SunJ, Shu J, Yue Q, Lu XQ, 2007. Multi-scale characteristics of the urban heat asland intensity inShanghai. Journal of East China Normal University (Natural Science), 2: 36–43.

Sun SM, Lu CY, 2002. Study on Monitoring Intensity of urban heat island and taking it as an indicator forurban ecosystem by remote sensing.Journal of Xiamen University (Natural Science), 41(1): 66–70.

TanJG, Zhen YF, Peng L, Gu SQ, Shi J, 2008. Effect of urban heat island to summer heat wave in Shanghai city. Plateau Meteorology, 27(special issue): 144–149.

Tong H, Sang JG, 2002. Numerical studies of atmospheric boundary layer in HaiDian districtof Beijing. Journal of Applied Meteorological Science,13(special issue): 51–60.

Wang GL, Jiang WM, Wei M, 2007. An Assessment of urban heat island effect using remote sensing data. Journal of Nanjing Institute of Meteorology, 30(3): 298–304.

Wang WJ, Shen WM, Liu XM, Zhang F, Pan YZ, Luo HJ, 2006a. Research on the relation of the urbanization and urban heat island effect changes in Beijing based on remote sensing. Research of Environmental Sciences,19(2): 44–48.

Wang XQ, Wang ZF, Guo H, 2006b. The study of the urban heat island in Beijing city. Climatic and Environmental Research, 11(5): 627–636.

Wang Y, Hu F, 2006. Variations of the urban heat island in summer of the recent 10 years over Beijing and its environment effect. Chinese Journal of Geophysics, 49(1): 61–68.

Wang YJ, Yan F, Zhang PQ, Ren FM, 2009. Variation of urban heat island in Beijing based on the vegetationindex and land surface albedo. Research of Environment Sciences, 22(2): 215–220.

Yan H, Deng LT, 2004. Character analysis of heat island in Shanghai city using remote sensing parameters. Journal ofTropical Meteorology, 20(5):579–585.

Yang YH, Xu XD, Weng YH, 2003. Simulation of daily cycle of boundary layer heat island in Beijing. Journal of Applied Meteorological Science,14(1): 61–68.

YuSQ, Bian LG, Lin XC, 2005. Changes in the spatial scale of Beijing urban heat island andurban development. Science in China D series: Earth Sciences, 35(special issue I): 97–106.

Zhang GZ, Xu XD, Wang JZ, Yang YQ, 2002a. A study of characteristics and evolution of urban heat island over Beijing and its surrounding area.Journal of Applied Meteorological Science, 13(special issue): 43–50.

Zhang JH, Hou YY, Li GC, Yan H, Yang LM, 2005. The diurnal and seasonal characteristics of urban heat island variation in Beijing city and surrounding areas and impact factors based on remote sensing satellite data. Science in China D series: Earth Sciences, 35(special issue I):187–194.

Zhang S, He BY, Du Y, 2003. Investigation on heat-island effect in Wuhan city zone by remote sensing. Resources and Environment in the Yangtze Basin, 12(5): 445–449.

Zhang Y, Yu T, Gu XF, ZhangYX, Chen LF, Yu SS, Zhang WJ, Li XW, 2006.Land surface temperature retrieval from CBERS-02 IRMSS thermal infrared data and its applications in quantitative analysis of urban heat island effect. Journal of Remote Sensing, 10(5): 789–797.

Zhang YP, He YL, Ma YX, Liu YH, Li YR, Dou JX, Guo P,2002b. Characteristics of vertical distribution of urban heat island effect in Kunming city. Plateau Meteorology, 21(6): 604–609.

Zhao LL, Zhao YS, Dong GH, 2006. Study on heat island effect in Guangzhou based on ETM+. Environmental Monitoring in China, 22(3):59–62.

Zheng ZF, Liu WD, Wang YC, 2006. Distributive character of urban heat island effect in the Beijing region. Journal of Nanjing Institute of Meteorology, 29(5): 694–699.

Zhou HM, Gao Y, Ge WQ, Li TT, 2008a. The Research on the relationship between the urban expansion and the change of the urban heat island distribution in Shanghai area. Ecology and Environment, 17(1): 163–168.

Zhou RW, Jiang WM, He XF, 2008b. Numerical Simulation of thermal effect of urban canopy structure on formation and intensity of urban heat island. Chinese Journal of Geophysics, 51(3): 715–726.

Zhou SZ, Shu J, 1994. City Climatology. Meteorological Press, Beijing.244–334.

Zhu JQ, Tang X, JiangH, 2006. The air temperature variation and heat island effect in urban area of Shanghai city. Plateau Meteorology, 25(6):1154–1160.

Zhu Y, Zhu LF, 2009. An analysis on summer urban heat island in Suzhou using satellite data. Scientia Meteorologica Sinica, 29(1): 77–83.

10.3724/SP.J.1226.2011.00070

11 April 2010 Accepted: 23 July 2010