Abstract

The urban heat island (UHI) effect is one of the most significant phenomena caused by urbanization. This study investigated the UHI effect in the Suzhou–Wuxi area, China, on 19–20 August 2010. Using a combination of meteorological station observations and Moderate Resolution Imaging Spectroradiometer (MODIS) surface skin temperature observations, this study demonstrated that an upwind UHI had an exacerbating influence on the downwind UHI during the study period. Numerical simulations using the Weather Research and Forecasting model also proved the importance of an upwind UHI influence on the leeward UHI in this area. For the near-surface UHI, the windward UHI effect is stronger at night than during the daytime because the background atmospheric stratification is more stable and the local lake breeze is weaker at night. However, in the daytime, a greater stability formed over the downwind city because of the warmer air heated by the windward urban area in the upper part of the planetary boundary layer and the cooler air transported from Tai Lake and the rural area in the lower part of the boundary layer. In comparison with the heating effect of a single city, the upwind UHI led to a decrease in the vertical wind speed of approximately 30% (from 0.15 to 0.10 m s−1) in the upper boundary layer over the downwind city and also reduced the near-surface turbulent movement by 25% (from 0.73 to 0.55 m2 s−2). These results improve the understanding of the overall influence of urban clusters on local synoptic/climate processes.

1. Introduction

Urbanization is one of the most extreme ways in which human activities change local land use and induce local land surface characteristics in urban areas that are substantially different from those in surrounding areas. These effects may alter the interactions between the land surface and the atmosphere (e.g., Grimmond et al. 2004) and may affect atmospheric processes ranging from the local scale, such as the urban climate/meteorological environment (e.g., C. L. Zhang et al. 2009), to the synoptic and global climate scales (e.g., Jin et al. 2005).

The urban heat island (UHI) phenomenon, characterized by a higher temperature difference between cities and the surrounding rural areas, is among the most significant urban phenomena. UHIs have been observed worldwide, not only in megacities (e.g., Childs and Raman 2005; Fast et al. 2005; Gaffin et al. 2008; Malevich and Klink 2011) but also in medium-sized and small cities (e.g., Giovannini et al. 2011; N. Zhang et al. 2011).

UHIs may trigger mesoscale circulation (UHI circulation). Using the large-eddy simulation method, Wang (2009) demonstrated that the UHI has a clear influence on wind flow and turbulence features, particularly under zero-wind conditions. Shepherd et al. (2002) documented the potential influence of urbanization on rainfall in Atlanta, Georgia, using Tropical Rainfall Measuring Mission (TRMM) satellite observations. Miao et al. (2011) described the urbanization effect on rainfall in Beijing City based on observations and numerical simulations.

In China, the urbanization process has accelerated significantly with the rapid economic development that has taken place in recent decades. Rapid urbanization in China has produced many city clusters and circles, such as the Yangtze River delta city cluster, the Pearl River delta city circle, and the Circum–Bohai Sea city zone. With the construction of such city clusters and circles, it is becoming increasingly important to study the effects of UHI clusters, rather than a UHI within a single city. An understanding of the influences of urbanization on regional meteorological–climate conditions is important to improve both urban and economic planning in rapidly developing countries such as China. Trusilova et al. (2008) investigated the regional influence of urbanization in Europe with a numerical model [the fifth-generation Pennsylvania State University–National Center for Atmospheric Research Mesoscale Model (MM5)] and found the conversion of rural to urban land results in significant changes to precipitation and near-surface temperature. Zhang et al. (2010) studied the urbanization impact in the Yangtze River delta and documented that in this area, the urbanization impact in summer is stronger and covers a larger area than that in winter because of the regional East Asian monsoon climate characterized by warm, wet summers and cool, dry winters. Wang et al. (2013) also found that over the Beijing–Tianjin–Hebei area of China the change caused by urbanization in near-surface-level temperature is most pronounced in winter, but the area influenced by urbanization is slightly larger in summer. These studies focused on the regional influence of urbanization, but little attention has been paid to the interactions between UHIs. D.-L. Zhang et al. (2009, 2011) reported an interaction and exacerbation of UHI effects between Baltimore, Maryland, and Washington, D.C., and demonstrated that by replacing Baltimore or its upstream area by natural land cover, the UHI effect might be reduced by 25%, and the windward UHI influence was strengthened by the effects of the local topography and sea-breeze circulation in the daytime. The present study focused on the interaction between two medium-sized cities, Suzhou and Wuxi, in the Yangtze River delta of China (Fig. 1). These two cities experienced a steady southeasterly background wind on 19–20 August 2010 and near-surface wind fields were not influenced by topography (the majority of the topography of the study area is less than 50 m above sea level); they are thus suitable for studying the upwind UHI influence caused only by urbanization.

Fig. 1.

(a) Terrain height (m) of the entire simulation domains; (b) Landsat 5 observed land use in the Suzhou–Wuxi area, which is the innermost domain of the WRF simulations conducted in this study.

Fig. 1.

(a) Terrain height (m) of the entire simulation domains; (b) Landsat 5 observed land use in the Suzhou–Wuxi area, which is the innermost domain of the WRF simulations conducted in this study.

2. Observed UHI effects

This study focused on a single day, that is, 0000 UTC 19 August–0000 UTC 20 August (or 0800 LST 19 August–0800 LST 20 August) 2010. On this day, a cyclone controlled the Siberian region, and the subtropical high pressure system persisted over the East China Sea. The Suzhou–Wuxi area was located at the western edge of the high pressure system. A steady southeasterly wind was present during this 24-h period. In this synoptic environment, the city of Wuxi was located downwind of the city of Suzhou. This study site and date constituted a suitable scenario for the study of the influence of an upwind UHI pattern on downwind UHI exacerbation and the diurnal variations in this influence.

Figure 2 shows the diurnal change of the UHI intensity at Suzhou and Wuxi. The UHI intensities were calculated using near-surface temperature differences between the observations at the urban stations of Suzhou and Wuxi and those at selected rural reference stations, as shown in Fig. 3. Only stations located upwind of these two cities were selected as the rural reference stations to avoid the influence of cities and the related upwind UHI effects. The observations show that discernible UHIs appeared in these two cities on 19 August, and the UHI intensities in the two cities shared the same diurnal profile. The UHI intensities were very weak (nearly 0°C) at sunrise. After sunrise, because of the solar heating of the earth’s surface, the UHI intensities in both Suzhou and Wuxi increased dramatically. After sunset, the UHI intensities continued to increase because of the release of heat stored in urban buildings and reached a maximum of 1.6°C in Suzhou and 2.1°C in Wuxi at 0200 LST 20 August. Strong UHIs were maintained until 0600 LST 20 August and then decreased dramatically prior to sunrise. In both cities, the nocturnal UHI intensities were much stronger than those in the daytime. From 0800 to 1600 LST, the UHI intensities were consistently less than 1.0°C and even less than 0.5°C before noon. Strong UHIs above 1.5°C occurred at 2300 LST and lasted until 0600 LST on the next day. In the daytime, the UHI intensities in both cities were very similar, despite the fact that Suzhou has a larger size and a greater population than Wuxi. At night, the averaged UHI intensity in Wuxi was approximately 0.3°C greater than that in Suzhou. In particular, at 2200 LST 19 August and 0200 LST 20 August, the UHI intensities of Wuxi were 0.7° and 0.6°C higher than those of Suzhou.

Fig. 2.

Observed and simulated UHIs and UHI difference (UHI-Wuxi − UHI-Suzhou) diurnal changes.

Fig. 2.

Observed and simulated UHIs and UHI difference (UHI-Wuxi − UHI-Suzhou) diurnal changes.

Fig. 3.

The 1-km-resolution MODIS-observed night UHI intensity (°C) on19 Aug 2010 in Suzhou and Wuxi. The meteorological observing stations used to calculate UHI (plus symbols) and the reference rural meteorological stations in Wuxi (circles) and in Suzhou (asterisks) are shown.

Fig. 3.

The 1-km-resolution MODIS-observed night UHI intensity (°C) on19 Aug 2010 in Suzhou and Wuxi. The meteorological observing stations used to calculate UHI (plus symbols) and the reference rural meteorological stations in Wuxi (circles) and in Suzhou (asterisks) are shown.

The Moderate Resolution Imaging Spectroradiometer (MODIS)/Aqua version 5 (V5) Land Surface Temperature and Emissivity (LST/E) L3 Global 1 km Grid product (MYD11A1; https://lpdaac.usgs.gov/products/modis_products_table/myd11a1) was used to analyze the surface skin temperature (TSK) distribution in this study. The V5 MYD11A1 products are projected in a sinusoidal grid by mapping the level 2 LST product (MYD11_L2) on 1-km (precisely, 0.928 km) grids, which comprise the following dataset layers for daytime and nighttime observations: TSK, quality control assessments, observation times, view zenith angles, clear-sky coverage, and bands 31 and 32 emissivities from land-cover types. Because the daytime TSK observations on 19 August are of poor observation quality, and TSK observations over most urban areas on this day are marked as missing values, only the night TSK observations are analyzed here, as shown in Fig. 3. These MODIS satellite observations of the nighttime TSK showed that the maximum TSK values occurred in the central areas of the two cities and that the minimum values occurred in rural areas (i.e., the surrounding cropland and water bodies). The maximum difference in TSK between the urban area of Suzhou and the reference rural area was greater than 4.6°C, and the difference in TSK between the central area of Wuxi and the same reference rural area was approximately 4.1°C (the urban areas and reference rural area were selected as the 1-km grids where urban–rural reference meteorological stations were located to compare with the station observations). The MODIS-observed TSK in Suzhou was 0.5°C higher than that in Wuxi, whereas the station-observed UHI in Wuxi was consistently higher than that in Suzhou during the night, as shown in Fig. 2. Both observations and numerical modeling proved that a strong relationship exists between TSK and T2m (air temperature at height of 2 m) and that T2m can be derived from satellite-observed TSK based on statistical models (Prihodko and Goward 1997; Gallo et al. 2011; Cui and de Foy 2012). Cui and de Foy (2012) also found that the land surface UHI is high at night and the near-surface air temperature UHI trend is similar to the land surface temperature UHI at night over Mexico City. On this day, the difference between meteorological station observations and MODIS observations indicates that the high nocturnal UHI in Wuxi occurred not only because of local surface heating, which was caused by the temperature difference between the earth’s surface and the overlying air, but also because of warm air advection from the upwind city caused by the southeasterly background wind field.

3. Numerical experiments

a. Model configuration

To investigate the windward UHI influence on 19 August 2010, the Advanced Research Weather Research and Forecasting numerical model (WRF) (ARW), version 3.3.0, was adopted. This model is based on a fully compressible and nonhydrostatic dynamic core (http://www.wrf-model.org). The model configuration consists of a parent domain and two nested domains centered at 31.3°N, 120.4°E, as shown in Fig. 1. A 1-km-resolution domain is nested inside a 3-km-resolution domain, and the 3-km-resolution domain is nested inside a 9-km-resolution domain. The simulations use a two-way interaction method (Zhang et al. 1986). The vertical grid system consists of 76 full-sigma levels, yielding approximately 30 layers below 3000 m. The model top is located at 50 hPa. The lowest full-sigma level above the ground is 0.998, and the lowest half-sigma level height is approximately 4 m above the ground.

The initial and boundary conditions were provided by National Centers for Environmental Prediction’s (NCEP) 1°-resolution Final Analyses (FNL) data, and the boundary conditions were forced every 6 h. The model integrations were conducted for 36 h, from 2000 LST 18 August to 0800 LST 20 August 2010. The results from the last 24 h are analyzed in this study.

The physics packages used in the simulations include the Rapid Radiative Transfer Model (RRTM; Mlawer et al. 1997) longwave radiation scheme and the Goddard (Chou and Suarez 1994) shortwave radiation scheme. The Noah land surface model (LSM), which is coupled with a single-layer urban canopy model (UCM; Chen et al. 2011), was used to represent the effect of urban land surface. For the domains of the three simulations, the fraction of built-up area was calculated using the 25-m-resolution land-cover database observed by the Landsat 5 satellite in the summer of 2006 (Xu et al. 2009). The average building height and street width were estimated using building information provided by local urban design and construction administrations. Based on these parameters, urban land cover was classified into three types: low-density residential, high-density residential, and commercial areas; the related parameters in the UCM were updated for the simulations. For other land-cover types, the land-cover data from the standard WRF database were used. The Mellor–Yamada–Janjić (MYJ) planetary boundary layer (PBL) scheme was used to parameterize the turbulence processes in the simulations. This PBL model has been shown to perform well (N. Zhang et al. 2011) in this area, and can output turbulent kinetic energy (TKE) for further boundary layer structural analysis. The Monin–Obukhov (Janjic) surface-layer parameterization scheme was used to calculate surface fluxes and to diagnose the near-surface meteorological fields, including air temperature at a height of 2 m and wind fields at a height of 10 m.

Four numerical experiments were designed to study the interaction of urban areas on 19–20 August under different urbanization scenarios, namely, experiments control (CTL), SUZHOU_ONLY, WUXI_ONLY, and NOURB. In the CTL experiment, the land-cover conditions from the Landsat 5 satellite observations were used to represent the current urbanization conditions of the Suzhou–Wuxi area. Two cities and three small towns were considered in this experiment, including Wuxi, Suzhou, and three small towns under the administration of the Suzhou political district that are located northeast of Suzhou. In the experiment WUXI_ONLY, only the urban cover of Wuxi was considered, and the land cover in the grids of the other cities was assumed to be cropland. In the experiment SUZHOU_ONLY, the urban cover in Suzhou and the three nearby small towns were considered. In the NOURB experiment, all urban land cover in the cities of Suzhou and Wuxi was replaced by cropland. Based on experiment CTL, experiment NOLAKE was designed to study the influence of Tai Lake; the Tai Lake was replaced by cropland in this experiment.

b. Simulation results

The simulation results of the CTL experiment were first compared with available observations to evaluate the model’s performance. The simulation error (E, defined as simulation results minus observation results), root-mean-square error (RMSE), and correlation coefficients (r) of T2m, Psfc (surface pressure), rh2m (relative humidity at 2 m), and U10m (wind speed at 10 m) were calculated using the hourly surface observations and the 1-km-resolution simulation results. They are given in Table 1, showing that the r values between observations and simulations were greater than 0.7, and demonstrating that the WRF model can generally simulate the meteorological fields well for this day. The model underestimated T2m and Psfc and overestimated rh2m and U10m. In our simulations, the release of anthropogenic waste heat is not considered, and this typically leads to an underestimation of near-surface air temperature. Another source of simulation error may be the PBL parameterization scheme. Zhang and Zheng (2004) found that surface wind and temperature were sensitive to the PBL parameterization scheme, and Hu et al. (2010) reported that MYJ parameterization produced cold and mist biases in the PBL simulations. The simulation error of U10m is only 0.15 m s−1, whereas the RMSE of U10m of 1.61 m s−1 is relatively high. This result is caused by the stronger effect of local underlying surface characteristics on near-surface wind speed; more detailed subscale information will improve the wind field simulation for complex urban areas.

Table 1.

Comparison of simulated and observed near-surface meteorological fields: E, RMSE, and r.

Comparison of simulated and observed near-surface meteorological fields: E, RMSE, and r.
Comparison of simulated and observed near-surface meteorological fields: E, RMSE, and r.

The UHIs of Suzhou and Wuxi were also calculated with the simulation results (shown in Fig. 2). The WRF model captures the UHIs in the two cities on 19–20 August. Both observations and simulations show that the UHI of Suzhou was stronger than that of Wuxi in the afternoon (from 1400 to 1700 LST), and the opposite at night (from 2100 to 0200 LST). From 0800 to 1300 LST, the simulated UHI intensity over Wuxi was close to that over Suzhou, while the observations showed that the UHI of Wuxi was slightly stronger (0.1°C higher) than that of Suzhou. For the period from 0300 to 0700 LST 20 August, the simulation overestimated the UHI in the two cities; the observations also show a slightly higher UHI in Wuxi, but the simulations showed a higher UHI in Suzhou.

The simulation results of 1400 LST 19 August and 0200 LST 20 August were selected to analyze the interaction between the UHIs of the two cities because the WRF model simulated the UHI intensities of the two cities well at these moment; at the same time, 1400 LST 19 August is when the maximum of T2m appeared, and 0200 LST 20 August is when the maximum UHI appeared in both observations and simulations. Figure 4 illustrates the horizontal distribution of T2m and TSK at 1400 LST 19 August and 0200 LST 20 August. Both T2m and TSK have high values in the urban area. At 1400 LST 19 August, TSK was above 43°C in both city centers, and T2m was nearly 36°C at the same time; at 0200 LST, TSK was about 29°C and T2m was 27°C in the city centers. The difference in TSK between Tai Lake, the largest water body in the inner simulation domain, and the land was about 8°C, and the difference in T2m was nearly 5°C at 1400 LST. The difference in thermal conditions between land and water might cause a land-breeze circulation, and the UHI could simultaneously enhance the strength of the circulation (N. Zhang et al. 2011). Figure 5 shows the difference in T2m and wind fields at the height of 10 m between experiments CTL and NOLAKE. The land–water thermal difference caused convergence of wind flow over the urban area, particularly over Wuxi. The convergence brought cool air from the lake surface and rural area, and reduced the daytime UHI intensity over Wuxi. In contrast, the land surface temperature difference at night was much lower than that in the daytime; Tai Lake only caused an increase of about 2.0°C in T2m. The convergence also disappeared at night, and wind flow only speeded up over the lake surface because of the decreasing of overlying surface roughness. UHIs will increase the lake breeze in the daytime by increasing the difference in T2m between water surface and land surface. At night, the lake breeze is weakened because the nocturnal UHI decreased the T2m contrast. The UHI effect on the lake breeze is relatively weak compared to the strong background wind and lake breeze, the changes in U10m were less than 1.0 m s−1, and there is no obvious modification of wind direction.

Fig. 4.

Simulated near-surface air temperature T2m (°C), skin temperature TSK (°C), and wind in experiment CTL: (a) T2m and (b) TSK at 1400 LST 19 Aug; (c),(d) as in (a),(b), respectively, but at at 0200 LST 20 Aug. Vectors indicate the horizontal wind field at a height of 10 m (m s−1).

Fig. 4.

Simulated near-surface air temperature T2m (°C), skin temperature TSK (°C), and wind in experiment CTL: (a) T2m and (b) TSK at 1400 LST 19 Aug; (c),(d) as in (a),(b), respectively, but at at 0200 LST 20 Aug. Vectors indicate the horizontal wind field at a height of 10 m (m s−1).

Fig. 5.

Differences in the near-surface air temperature (T2m, °C) and wind field at a height of 10 m (m s−1) between experiment CTL and NOLAKE at (a) 1400 LST 19 Aug and (b) 0200 LST 20 Aug.

Fig. 5.

Differences in the near-surface air temperature (T2m, °C) and wind field at a height of 10 m (m s−1) between experiment CTL and NOLAKE at (a) 1400 LST 19 Aug and (b) 0200 LST 20 Aug.

Figure 6 shows the difference in T2m between each of the experiments CTL, SUZHOU_ONLY, and WUXI_ONLY and experiment NOURB at 1400 LST 19 August and 0200 LST 20 August. The simulations show a stronger nocturnal UHI. Compared with the results from experiment NOURB, T2m in experiment CTL increased in the two cities. The increases in T2m in the two cities were approximately 0.9°C at 1400 LST and greater than 1.2°C at 0200 LST in the central area of the two cities. In experiment CTL, the windward UHI influence was clearly defined for Wuxi because of the stable and continuous background southeasterly wind, and a T2m increase of 1.5°C in Wuxi was therefore caused by the advection of warm air downwind from Suzhou. Compared with a T2m increase of 1.1°C in Wuxi in experiment WUXI_ONLY, the T2m increase of 1.5°C in Wuxi in experiment CTL indicates that the upwind influence contributed 27% of the temperature increase in Wuxi in this experiment. This percentage is similar to the results of D.-L. Zhang et al. (2009).

Fig. 6.

Differences in the near-surface air temperature (T2m, °C) between various experiments: (a) experiment CTL − experiment NOURB at 1400 LST 19 Aug; (b) experiment SUZHOU_ONLY − experiment NOURB at 1400 LST 19 Aug; (c) experiment WUXI_ONLY − experiment NOURB at 1400 LST 19 Aug; (d) experiment CTL − experiment NOURB at 0200 LST 20 Aug; (e) experiment SUZHOU_ONLY − experiment NOURB at 0200 LST 20 Aug; (f) experiment WUXI_ONLY − experiment NOURB at 0200 LST 20 Aug. Vectors indicate the horizontal wind field at a height of 10 m in experiment CTL (m s−1).

Fig. 6.

Differences in the near-surface air temperature (T2m, °C) between various experiments: (a) experiment CTL − experiment NOURB at 1400 LST 19 Aug; (b) experiment SUZHOU_ONLY − experiment NOURB at 1400 LST 19 Aug; (c) experiment WUXI_ONLY − experiment NOURB at 1400 LST 19 Aug; (d) experiment CTL − experiment NOURB at 0200 LST 20 Aug; (e) experiment SUZHOU_ONLY − experiment NOURB at 0200 LST 20 Aug; (f) experiment WUXI_ONLY − experiment NOURB at 0200 LST 20 Aug. Vectors indicate the horizontal wind field at a height of 10 m in experiment CTL (m s−1).

The windward UHI influence was not limited to the enhancement of the UHI downwind. The near-surface warm air from the upwind city had a widespread and distant effect in the downwind direction. The urbanization effect index (EI) of the near-surface meteorological parameter, defined as (Zhang et al. 2010), is used to quantify the area effect of urbanization on the near-surface air temperature. Here, x is the near-surface air temperature (T2m), Achange(x) is the total area where T2m changed between each of the experiments CTL, WUXI_ONLY, and SUZHOU_ONLY and experiment NOURB, and Aurban is the difference in the urban land-cover area. The EI (T2m) values at 1400 LST 19 August in experiments CTL, SUZHOU_ONLY, and WUXI_ONLY were 2.19, 2.49, and 2.31, respectively, and the EI (T2m) values at 0200 LST 20 August were 4.99, 5.87, and 5.58, respectively. The EI (T2m) values in experiment CTL were reduced by approximately 10% in the daytime and 15% at night, compared with the EI (T2m) values in experiments WUXI_ONLY and SUZHOU_ONLY, which considered the urbanization of only a single city. This decrease in the area of influence is attributed to the overlapping influence of the windward city (Suzhou) on the downwind city (Wuxi). This finding suggests that when several cities merge into a city cluster, the total influential area may decrease, and the UHIs over downwind cities may be enhanced.

UHIs change not only the near-surface air temperature field, but also the air temperature in the entire boundary layer. This influence may reach a height above 1 km when the PBL is unstable, particularly in the daytime. Figure 7 shows the differences in T1km (air temperature at the height of 1 km) between each of the experiments CTL, SUZHOU_ONLY, and WUXI_ONLY and experiment NOURB at 1400 LST 19 August. At a height of 1 km, the influence of Tai Lake was weak, and the convergence over the urban area disappeared at this time. Unlike T2m, the T1km increase shows a clear windward influence at 1400 LST. A T1km increase of more than 0.7°C appeared over Wuxi, whereas the T1km increase in Suzhou was approximately 0.5°C. In the experiment SUZHOU_ONLY, the T1km increase in Suzhou was close to that in experiment CTL, but in experiment WUXI_ONLY, the T1km increase over Wuxi was quite weak, only approximately 0.3°C. Compared with experiment WUXI_ONLY, the T1km over Wuxi city increased nearly 130% in experiment CTL. These results demonstrate that the windward influence of the UHI is more evident in the upper part of the boundary layer than near the surface.

Fig. 7.

Differences in air temperature at the height of 1 km (T1km, °C) between various experiments: (a) experiment CTL − experiment NOURB; (b) experiment SUZHOU_ONLY − experiment NOURB; (c) experiment WUXI_ONLY − experiment NOURB at 1400 LST 19 Aug. Vectors indicate the horizontal wind field at the height of 1 km in experiment CTL (m s−1).

Fig. 7.

Differences in air temperature at the height of 1 km (T1km, °C) between various experiments: (a) experiment CTL − experiment NOURB; (b) experiment SUZHOU_ONLY − experiment NOURB; (c) experiment WUXI_ONLY − experiment NOURB at 1400 LST 19 Aug. Vectors indicate the horizontal wind field at the height of 1 km in experiment CTL (m s−1).

Figure 8 shows the differences in the potential temperature and vertical wind speed between each of the experiments CTL, SUZHOU_ONLY, and WUXI_ONLY and NOURB at 1400 LST 19 August and 0200 LST 20 August in a cross section extending from the city center of Suzhou to the city center of Wuxi (the red line shown in Fig. 1b). At night, the warm air from Suzhou was transported downward near the surface; the influence was limited to a height of less than 300 m. In the daytime, because of the convective boundary layer structure, the warming influence reached up to 1500 m above the ground and expanded downwind to Wuxi. However, the warming influence in the low-level boundary layer was neutralized by the cool air converged to Wuxi. This process helps to explain why the upwind influence on the near-surface temperature was weaker in the daytime. The transport of warm air also changed the boundary layer structure over Wuxi in the daytime. Compared with experiment WUXI_ONLY, the potential temperature over Wuxi between a height of 800 m and 1.5 km increased by approximately 0.2°C, with a peak increase of 0.5°C in experiment CTL. Because of the increase in the potential temperature of the upper boundary layer, the PBL over Wuxi became more stable and less turbulent. Compared with the vertical air motion in experiment WUXI_ONLY, the vertical wind speed over Wuxi in experiment CTL decreased by nearly one-third, from 0.15 to 0.10 m s−1 at 1400 LST 19 August; at the same time, WRF-simulated near-surface turbulent energy decreased by 25%, from 0.73 to 0.55 m2 s−2.

Fig. 8.

Differences in the potential temperature (contours, °C) and vertical wind speed (color shading, m s−1) along a cross section extending between the cities of Wuxi and Suzhou (red line in Fig. 1c) between various experiments: (a) experiment CTL − experiment NOURB at 1400 LST 19 Aug; (b) experiment SUZHOU_ONLY − experiment NOURB at 1400 LST 19 Aug; (c) experiment WUXI_ONLY − experiment NOURB at 1400 LST 19 Aug; (d) experiment CTL − experiment NOURB at 0200 LST 20 Aug; (e) experiment SUZHOU_ONLY − experiment NOURB at 0200 LST 20 Aug; (f) experiment WUXI_ONLY − experiment NOURB at 0200 LST 20 Aug (the red bar under the x-coordinate axis indicates the location of Suzhou, and the black bar indicates the location of Wuxi). Gray vectors indicate the wind field in experiment CTL (m s−1).

Fig. 8.

Differences in the potential temperature (contours, °C) and vertical wind speed (color shading, m s−1) along a cross section extending between the cities of Wuxi and Suzhou (red line in Fig. 1c) between various experiments: (a) experiment CTL − experiment NOURB at 1400 LST 19 Aug; (b) experiment SUZHOU_ONLY − experiment NOURB at 1400 LST 19 Aug; (c) experiment WUXI_ONLY − experiment NOURB at 1400 LST 19 Aug; (d) experiment CTL − experiment NOURB at 0200 LST 20 Aug; (e) experiment SUZHOU_ONLY − experiment NOURB at 0200 LST 20 Aug; (f) experiment WUXI_ONLY − experiment NOURB at 0200 LST 20 Aug (the red bar under the x-coordinate axis indicates the location of Suzhou, and the black bar indicates the location of Wuxi). Gray vectors indicate the wind field in experiment CTL (m s−1).

4. Conclusions and discussions

In this study, the UHI effect occurring in the Suzhou–Wuxi area of China on 19 August 2010 was investigated using meteorological station observations, MODIS satellite observations and a coupled WRF–Noah–UCM model. The meteorological station data demonstrated that the UHI intensity in the downwind city (Wuxi) is similar to or stronger than that in the windward city (Suzhou), and the MODIS-observed skin temperature in the downwind city was lower than that in the upwind city. These results indicate that upwind urbanization enhances downwind UHI effects, as D.-L. Zhang et al. (2009, 2011) have documented.

A comparison of control and sensitivity simulations also demonstrated the importance of the upwind UHI influence on the downwind UHI. This study confirmed that the same upwind influence occurred over medium-sized cities, as well as over large cities (e.g., Washington, D.C.; D.-L. Zhang et al. 2009, 2011). Based on the simulation results, the nocturnal UHI in Wuxi may be reduced by 27% in the absence of an upwind urbanization influence. With regard to the near-surface UHI, the upwind exacerbating effect is stronger at night than in the daytime because the background atmospheric stratification is more stable at night. In the daytime, the upwind UHI effect on the lower PBL is reduced by the convergence over the downwind city. This effect increased the potential temperature and induced more stable conditions over the downwind city. Compared with the heating effect of a single city, this effect of the windward city led to a decrease in the vertical wind speed of approximately 30% (from 0.15 to 0.10 m s−1) in the upper boundary layer over the leeward city, as well as a reduction in turbulent air movement.

With rapid economic development, 18 city circles and city clusters have merged in China in recent decades, and the total number of these city circles/clusters may reach 50 by 2050 (from Research Center on Metropolitan Regions; http://www.rcmrc.sjtu.edu.cn/). Such rapid urbanization will likely impact environmental and climate processes from local scales to regional scales. Most new city circles/clusters will be composed of small or medium-sized cities, although urban climate research in these areas is just beginning. Study of the urbanization influence on weather and regional climate is fundamental in improving the urban residence comfort level and producing better urban plans to reduce hazard disasters. Additionally, the influence of urbanization on the local meteorological environment and regional climate is very complicated. Many effects are involved in this process, including radiation, precipitation, pollutant release, etc., but urban atmospheric boundary layer structure is one of the key processes to understanding other more complex processes. The results in this study will aid in the understanding of the influence of urbanization interactions in small to medium-sized cities in China.

Acknowledgments

This work is funded by the National Basic Research Program of China (2010CB428501), National Natural Science Foundation of China (41375014), and Jiangsu Collaborative Innovation Center for Climate Change, China. We are grateful to three anonymous reviewers for their valuable comments.

REFERENCES

REFERENCES
Chen
,
F.
, and et al
,
2011
:
The integrated WRF/urban modeling system: Development, evaluation, and applications to urban environmental problems
.
Int. J. Climatol.
,
31
,
273
288
.
Childs
,
P. P.
, and
S.
Raman
,
2005
:
Observations and numerical simulations of urban heat island and sea breeze circulations over New York City
.
Pure Appl. Geophys.
,
162
,
1955
1980
.
Chou
,
M.
, and
M.
Suarez
,
1994
: An efficient thermal infrared radiation parameterization for use in general circulation models. NASA Tech. Memo. 104606, 84 pp.
Cui
,
Y. Y.
, and
B.
de Foy
,
2012
:
Seasonal variations of the urban heat island at the surface and the near-surface and reductions due to urban vegetation in Mexico City
.
J. Appl. Meteor. Climatol.
,
51
,
855
868
.
Fast
,
J.
,
J.
Torcolini
, and
R.
Redman
,
2005
:
Pseudovertical temperature profiles and the urban heat island measured by a temperature datalogger network in Phoenix, Arizona
.
J. Appl. Meteor.
,
44
,
3
13
.
Gaffin
,
S. R.
, and et al
,
2008
:
Variations in New York City’s urban heat island strength over time and space
.
Theor. Appl. Climatol.
,
94
,
1
11
.
Gallo
,
K.
,
R.
Hale
,
D.
Tarpley
, and
Y.
Yu
,
2011
:
Evaluation of the relationship between air and land surface temperature under clear- and cloudy-sky conditions
.
J. Appl. Meteor. Climatol.
,
50
,
767
775
.
Giovannini
,
L.
,
D.
Zardi
, and
M.
de Franceschi
,
2011
:
Analysis of the urban thermal fingerprint of the city of Trento in the Alps
.
J. Appl. Meteor. Climatol.
,
50
,
1145
1162
.
Grimmond
,
C. S. B.
,
J. A.
Salmond
,
T. R.
Oke
,
B.
Offerle
, and
A.
Lemonsu
,
2004
: Flux and turbulence measurements at a densely built-up site in Marseille: Heat, mass (water and carbon dioxide), and momentum. J. Geophys. Res.,109, D24101,
doi:10.1029/2004JD004936
.
Hu
,
X.-M.
,
J. W.
Nielsen-Gammon
, and
F. Q.
Zhang
,
2010
:
Evaluation of three planetary boundary layer schemes in the WRF Model
.
J. Appl. Meteor. Climatol.
,
49
,
1831
1844
.
Jin
,
M.
,
R.
Dickinson
, and
D.
Zhang
,
2005
:
The footprint of urban areas on global climate as characterized by MODIS
.
J. Climate
,
18
,
1551
1565
.
Malevich
,
S. B.
, and
K.
Klink
,
2011
:
Relationships between snow and the wintertime Minneapolis urban heat island
.
J. Appl. Meteor. Climatol.
,
50
,
1884
1894
.
Miao
,
S.
,
F.
Chen
,
Q. C.
Li
, and
S. Y.
Fan
,
2011
:
Impacts of urban processes and urbanization on summer precipitation: An experiment study of heavy rainfall in Beijing on 1 August 2006
.
J. Appl. Meteor. Climatol.
,
50
,
806
825
.
Mlawer
,
E. J.
,
S. J.
Taubman
,
P.
Brown
,
M. J.
Iacono
, and
S. A.
Clough
,
1997
:
Radiative transfer for inhomogeneous atmosphere: RRTM, a validated correlated-k model for the longwave
.
J. Geophys. Res.
,
102
,
16 663
16 682
.
Prihodko
,
L.
, and
S. N.
Goward
,
1997
:
Estimation of air temperature from remotely sensed surface observations
.
Remote Sens. Environ.
,
60
,
335
346
.
Shepherd
,
J. M.
,
H.
Pierce
, and
A. J.
Negri
,
2002
:
Rainfall modification by major urban areas: Observations from spaceborne rain radar on the TRMM satellite
.
J. Appl. Meteor.
,
41
,
689
701
.
Trusilova
,
K.
,
M.
Jung
,
G.
Churkina
,
U.
Karstens
,
M.
Heimann
, and
M.
Claussen
,
2008
:
Urbanization impacts on the climate in Europe: Numerical experiments by the PSU–NCAR Mesoscale Model (MM5)
.
J. Appl. Meteor. Climatol
,
47
,
1442
1455
.
Wang
,
M. N.
,
X. Z.
Zhang
, and
X. D.
Yan
,
2013
:
Modeling the climatic effects of urbanization in the Beijing–Tianjin–Hebei metropolitan area
.
Theor. Appl. Climatol.
,
113
,
377
385
.
Wang
,
W.
,
2009
:
The influence of thermally induced mesoscale circulations on turbulence statistics over an idealized urban area under a zero background wind
.
Bound.-Layer Meteor.
,
131
,
403
423
.
Xu
,
Y.
,
Z.
Qin
, and
Y.
Zhu
,
2009
:
Spatial and temporal analysis of urban heat island in Suzhou city by remote sensing (in Chinese)
.
Sci. Geogr. Sin.
,
29
,
529
534
.
Zhang
,
C. L.
,
F.
Chen
,
S.
Miao
,
Q. C.
Li
,
X. A.
Xia
, and
C. Y.
Xuan
,
2009
:
Impacts of urban expansion and future green planting on summer precipitation in the Beijing metropolitan area
.
J. Geophys. Res.
,
114
, D02116, doi:10.1029/2008JD010328.
Zhang
,
D.-L.
, and
W.-Z.
Zheng
,
2004
:
Diurnal cycles of surface winds and temperatures as simulated by five boundary layer parameterizations
.
J. Appl. Meteor.
,
43
,
157
169
.
Zhang
,
D.-L.
,
H.
Chang
,
N.
Seaman
,
T. T.
Warner
, and
J. M.
Fritsch
,
1986
:
A two-way interactive nesting procedure with variable terrain resolution
.
Mon. Wea. Rev.
,
114
,
1330
1339
.
Zhang
,
D.-L.
,
Y.-X.
Shou
, and
R.
Dickerson
,
2009
:
Upstream urbanization exacerbates urban heat island effects
.
Geophys. Res. Lett.
,
36
, L24401,
doi:10.1029/2009GL041082
.
Zhang
,
D.-L.
,
Y.-X.
Shou
,
R.
Dickerson
, and
F.
Chen
,
2011
:
Impact of upstream urbanization on the urban heat island effects along the Washington–Baltimore corridor
.
J. Appl. Meteor. Climatol.
,
50
,
2012
2029
.
Zhang
,
N.
,
Z.
Gao
,
X.
Wang
, and
Y.
Chen
,
2010
:
Modeling the impact of urbanization on the local and regional climate in Yangtze River Delta, China
.
Theor. Appl. Climatol.
,
102
,
331
342
.
Zhang
,
N.
,
L.
Zhu
, and
Y.
Zhu
,
2011
:
Urban heat island and boundary layer structures under hot weather synoptic conditions: An experiment study of the city of Suzhou, China
.
Adv. Atmos. Sci.
,
28
,
855
865
.