1. Introduction
Land use and land cover (LULC) change can affect regional climate by altering energy and water exchange between the land and atmosphere (Gibbard et al. 2005; Pielke et al. 1998; Weaver and Avissar 2001). Contributions from LULC changes, including urban heat islands, to the globally averaged land surface air temperature at 2 m (SAT) change are unlikely to exceed 10%; however, their impact at the regional or local scales in an area with rapid economic development and human activities may be greater (Findell et al. 2007; IPCC 2013).
Urbanization in China has occurred over a 30-yr period from the 1980s to 2010s, in contrast with the urbanization that has occurred over 100 yr in developed western countries, leading to different impacts. The rapid and concentrated urbanization in China, especially in the eastern part (Liu et al. 2005), is sure to have had a significant influence on the local and even regional energy budget, and so studies of the impacts on temperature have attracted a great deal of attention (Li et al. 2010; J. Liu et al. 2014; Ren et al. 2008; Yan et al. 2014). Meanwhile, as East Asia experiences a typical monsoon climate (Fu et al. 2000), another key question of how to uncover the impact of the rapid urbanization in the last 30 yr on the monsoon circulation and precipitation requires further study. There are clear seasonal variations in the effect of urbanization in China on SAT, which are more significant for the maximum in summer (W. D. Liu et al. 2014) and have influence on spatial and temporal patterns of precipitation (Dong et al. 2008; Ren et al. 2010).
Several methods have been used to detect the urbanization effect on SAT due to increases in both population and urbanized areas. The most popular methods involve urban minus rural meteorological observation (UMR; Gallo et al. 1999; Yang et al. 2011) and observation minus reanalysis (OMR; Hu et al. 2010; Kalnay and Cai 2003; Zhou et al. 2004). However, because of the varying impact of urbanization on the identified rural sites and the meteorological station relocations (Yan et al. 2010; Yang et al. 2013; Zhou et al. 2013), rural site selection is under much dispute (Hansen et al. 2010; Portman 1993; Ren et al. 2008). Meanwhile, the heterogeneity of observation site distributions compared with the reanalysis data and both the diversity and scale differences of different reanalysis data require care and attention when using the OMR method. Thus, the contributions to local or regional warming estimated by these two methods can differ substantially. Satellite observations (Ren and Ren 2011; Wang and Ge 2012) and sea surface temperature datasets have also been used in urbanization studies (Jones et al. 2008). However, it can be difficult to reveal the impact of the urban surface on the precipitation with such methods because of the spatiotemporal inhomogeneity of the precipitation itself. Therefore, using real urban surface data with a model with nested downscaling at finer spatial resolution than that of the standard observed data is a valuable approach to detect the impact on SAT and precipitation (Barth et al. 2005; Bukovsky and Karoly 2009; Z. X. Liu et al. 2011), which is the key object of the present work. It is challenging to evaluate the impact on regional climate from urban surface expansion at different spatial and temporal scales.
In reality, many previous studies of the influence of urban surface expansion have been performed; however, most of these concentrate on a single city or city cluster (Han et al. 2014; Lin et al. 2008; Mote et al. 2007; A. Y. Zhang et al. 2010). For example, some studies have focused on urban heat island and the increased precipitation in the regions downstream of a city or city cluster (Changnon et al. 1991; Shepherd et al. 2002). Some studies have been performed that depend on virtual or semivirtual urban surface data, including such changes as from cropland, grassland, or forest to urban, which may lead to overestimated or underestimated land parameter changes and thus result in errors in evaluation of the urbanization-induced climatic effect (Chen and Zhang 2013; Miao et al. 2010; Shao et al. 2013).
The LULC data in the commonly adopted regional climate models, such as the Fifth-generation Pennsylvania State University–NCAR Mesoscale Model version 3.7 (MM5V3.7; Grell et al. 1994) and the Weather Research and Forecasting (WRF) Model (Skamarock et al. 2008), are from two datasets: the U.S. Geological Survey (USGS) land use data (U92; Guo and Chen 1994; Loveland et al. 2000) and the International Geosphere–Biosphere Programme (IGBP) Moderate Resolution Imaging Spectroradiometer (MODIS) data (U01). These two datasets are usually adopted in numerical experiments with unchanged LULC. Meanwhile, a significant difference can be detected between the two datasets in urban surface areas (Li et al. 2014; Ye et al. 2014).
With the rapid economic development of the last 30 yr, especially in eastern China, an increasing number of people are moving to the cities, which induced rapid urban surface expansion (J. Liu et al. 2014). The roughness length is changed because of the varied land surface characteristics, as well as the building density and height (Shem and Shepherd 2009; Thielen et al. 2000). The increased impervious surfaces due to the urban surface expansion induce rapid runoff of rainwater into drains (Z. H. Liu et al. 2011). Shortwave–longwave radiation trapping in urban canyons serves to reduce albedo compared to rural surfaces (Jiang et al. 2007). Meanwhile, the building materials that are widely used in urban areas generally have greater heat capacity than rural areas (Zhao et al. 2008). The urban-associated pollutants (Zhuang et al. 2014) and anthropogenic heat releases (Ichinose et al. 1999) are also enhanced, which might have different (positive/negative) impacts on regional climate. These changes represent the key research questions, which have attracted much interest and should be studied in detail to understand the climatic effect of urbanization. However, there are difficulties associated with obtaining these data as a result of their spatiotemporal inhomogeneity, in which there exist large uncertainties. Consequently, our study is based on a regional climate model with urban effect considered. However, with the development of remote sensing technology, the accuracy of satellite-based urban surface data has been improved. It is possible to more accurately represent the urban surface distribution and expansion (Jia et al. 2014). Therefore, realistic data can be attained with which to perform studies into the impact of urban surface expansion on regional climate.
To reproduce the effect of urbanization on regional climate in rapidly developing regions, especially eastern China, satellite-derived urban surface data (Hu et al. 2015; Jia et al. 2014), which should reflect the dynamic variations of urban surface expansion, are adopted. A 10-yr simulation with fine resolution over East Asia using nested MM5V3.7 with urban effect considered is performed to explore the impact of urban surface expansion on the SAT and precipitation. Meanwhile, the impacts on regional climate from different urban surface backgrounds were also differentiated.
2. Numerical experiment design and data
a. Driving data
Initial conditions and time-varying boundary conditions for the integrations with nested MM5V3.7 were taken from the NCEP–NCAR reanalysis dataset with a 2.5° × 2.5° resolution. The data were bilinearly interpolated onto the model domain as initial and boundary conditions, and the boundary conditions were updated every 6 h.
b. Experimental design
The experiment was continuously integrated from 1 January 2002 to 31 December 2012 (11 yr) based on the different satellite-based urban surface data images for the 1980s, 1990s, 2000s, and 2010s in China (abbreviated as U80, U90, U00, and U10, respectively). The same driving data for 2002–12, including sea surface temperatures as well as NCEP–NCAR atmospheric data, were used for each simulation, with the only differences being that the urban land use data over China represented the different decades. The first year was the model spinup, and the results for the remaining 10 yr were analyzed. The central latitude and longitude of the simulated domain were 35°N and 108.5°E. The horizontal coarse mesh had 259 longitudinal grid points and 199 latitudinal grid points, including a 15-grid point buffer zone that was not used in the analysis (Fig. 1a). The horizontal resolution was 30 km for the coarse domain and 10 and 3.3 km for the two nested domains. The first nested domain covered most of eastern China, with 222 longitudinal grid points and 312 latitudinal grid points. The second nested domain concentrated on the three city clusters [Beijing–Tianjin–Hebei (BTH) and the Yangtze and Pearl River deltas (YRD and PRD, respectively)], with 150 longitudinal grid points and 120 latitudinal grid points. The air pressure at the top of the model was 10 hPa, and there were 23 levels in the vertical direction. The main physical parameterization schemes for the current experiment, based on previous studies (Zhao 2012, 2013; Zhao et al. 2007, 2009), included the Kain–Fritsch cumulus scheme (Kain and Fritsch 1993); the Medium Range Forecast Model (MRF) planetary boundary layer (PBL; Hong and Pan 1996), which uses enhanced vertical flux coefficients in the PBL and the PBL height is determined from a critical bulk Richardson number; and the RRTM radiation scheme (Mlawer et al. 1997). The Noah land surface scheme (Chen and Dudhia 2001a,b) was adopted, which considers urban effects and emissivity in computing surface temperature through changing urban-surface-related physical parameters, including roughness length, minimum stomatal resistance, areal fractional coverage of green vegetation, and soil thermal properties (Dudhia 2005), and has been proven to display nice performance in SAT simulation over the city of Beijing (Z. X. Liu et al. 2011).
(a) Model domain and terrain height distribution (m) with nested domains (BTH, YRD, and PRD) and (b) diagram of subregions in China (NE, NC, EC, SC, NWE, SW, NWW, and TP).
Citation: Journal of Climate 30, 3; 10.1175/JCLI-D-15-0604.1
To more concisely express the subregional trends, eight subregions in China were considered (Fig. 1b), including northeast China (NE), north China (NC), east China (EC), south China (SC), the eastern part of northwest China (NWE), southwest China (SW), the western part of northwest China (NWW), and the Tibetan Plateau (TP).
c. Satellite-based urban surface data
The urban surface data were established by integrating census information and multiple-source satellite images, and national land cover datasets were obtained from the Chinese Data Sharing Infrastructure of Earth System Science (DSIESS). Analysis of these data was undertaken to determine the general trends of the urbanization dynamics and their spatial patterns. Further, nighttime light datasets from the Defense Meteorological Satellite Program (DMSP) Operational Linescan System (OLS) were used to delineate socioeconomic activities, such as urbanization and population growth. The datasets, which performed best in mapping urban land cover in the evaluation process, were then selected and combined to create 1-km-resolution data for urban land cover fusion (Hu et al. 2015; Jia et al. 2014). To provide more accurate information for climate model simulations at different spatial scales, the fractional urban cover was calculated at spatial resolutions of 30, 10, and 3.3 km, which is consistent with the model domains. With the fractional urban cover and the corresponding values from other land use categories, the dominant land category at a certain grid cell was calculated according to Guo and Chen (1994).
3. Results
a. Comparisons of different urban surface data
The default LULC characteristics (U92 and U01) and new derived LULC data (U80, U90, U00, and U10) by Hu et al. (2015) and Jia et al. (2014) over the three city clusters in the model are shown in Figs. 2a–r. The numbers of urban grid cells in BTH, YRD, and PRD are 58, 42, and 16 from U92 and 733, 996, and 1013 from U01, respectively, which reveals that the numbers of urban grid cells from U92 are far lower than that from U01 (Table 1). The fixed-in-time LULC data have been adopted to perform simulations on numerical weather forecast and climate studies, which are adequate on annual or seasonal scales. However, it is inadequate for long-term integration in the areas with rapid economic development and human activities, in which variations for urban surface are necessary.
Distributions of land use category (colors referring to different land use category, where red is for urban surface grids) at 3.3-km resolution in the (a)–(f) BTH, (g)–(l) YRD, and (m)–(r) PRD city clusters for the (a),(g),(m) 1980s; (b),(h),(n) 1990s; (c),(i),(o) 2000s; and (d),(j),(p) 2010s, as well as the default values from (e),(k),(q) USGS (U92) and (f),(l),(r) MODIS (U01). Land use categories: 1) URBAN, urban and built in; 2) DRCRP, dryland cropland and pasture; 3) IRCRP, irrigated cropland and pasture; 4) MIXCP, mixed dryland/irrigated cropland and pasture; 5) CRGRM, cropland/grassland mosaic; 6) CRWDM, cropland/woodland mosaic; 7) GRSLD, grassland; 8) SHRLD, shrubland; 9) SHRGR, mixed shrubland/grassland; 10) SAVAN, savanna, 11) DBFST, deciduous broadleaf forest; 12) DNFST, deciduous needleleaf forest; 13) EBFST, evergreen broadleaf; 14) ENFST, evergreen needleleaf; 15) MXFST, mixed forest; 16) WATER, water bodies; 17) HWTLD, herbaceous wetland; 18) WWTLD, wooded wetland; 19) BARSP, barren or sparsely vegetated; 20) HRTUN, herbaceous tundra; 21) WDTUN, wooded tundra; 22) MXTUN, mixed tundra; 23) BGTUN, bare ground tundra; and 24) SNWIC, snow or ice.
Citation: Journal of Climate 30, 3; 10.1175/JCLI-D-15-0604.1
Numbers of urban grid cells at a 3.3-km spatial resolution (the area of each grid cell is 10.89 km2) in the three city clusters for the new variation of urban surface data in the 1980s (U80), 1990s (U90), 2000s (U00), and 2010s (U10) and the default values from the USGS (U92) and MODIS (U01). The total number of grid points is 18 000 over the three city clusters, within which 86.1%, 69.6%, and 74.6% are land grid cells over the BTH, YRD, and PRD, respectively.
Variations of urban surface expansion in the last 30 yr in the three city clusters clearly indicate a slow expansion for the first 20 yr, with a dramatic increase from the 2000s to the 2010s (U10 − U00). The numbers of urban grid cells are severely underestimated in the default U92. However, the values from U01 are closer to those from U00, which expresses the reliability of the new urban surface data to a certain degree. The numbers of urban grid cells increases by 5.2–13.9 times for U90 − U80 because of fewer cells from U80 (Table 1). Meanwhile, the number increases by 1.0–1.5 times for U00 − U90 and 2.7–3.4 times for U10 − U00. Therefore, a very similar increasing tendency for the numbers of urban grid cells between different urban surface data can be detected among the three city clusters, except for the much greater expansion in PRD for U90 − U80, which is due to the fewer urban grid cells for U80 and rapid development of township enterprises with the national reform and opening policy in the early 1980s (Fang 2009).
b. Influence on surface air temperature from U10 − U80
1) Annual differences
Spatial distributions for annual SAT changes were consistent with the urban surface distributions from U10 − U80, in which there were greater values in the urban surface areas (Fig. 3). The impact over East Asia showed that the urban-related warming was not limited to a single city or city clusters but extended with a SAT-increased belt covering the eastern coast of China (Figs. 3p–t). Spatial distributions for annual SAT changes passed a significance test (the 90% confidence level paired t test) mainly in the urban surface areas.
Spatial distribution of (left) annual and (columns 2–5) seasonal SAT changes in the three city clusters (a)–(e) BTH, (f)–(j) YRD, and (k)–(o) PRD and (p)–(t) the eastern part of China due to the difference in urban surface from the 1980s to the 2010s (U10 − U80) for (b),(g),(l),(q) MAM; (c),(h),(m),(r) JJA; (d),(i),(n),(s) SON; and (e),(j),(o),(t) DJF (°C). The vertical lines denote areas passing the 90% confidence level t test.
Citation: Journal of Climate 30, 3; 10.1175/JCLI-D-15-0604.1
(i) In the three city clusters
The annual urban-related warming values were 0.38° (BTH), 0.30° (YRD), and 0.28°C (PRD) (Fig. 4). Further analysis under different surface characteristics in the three city clusters is also discussed. In BTH, the contributions to subregional warming were 0.43° and 0.073°C on land and in the ocean, respectively. When only urban surface areas (URBAN), including areas with LULC transition from nonurban to urban land use (N2U) and areas that were classified as urban in both time periods (U2U), were considered, the annual urban-related warming increased to 1.06°C. The value in URBAN was closer to that from N2U (1.07°C) than it was to that from U2U (0.62°C). In YRD, the contributions were 0.40° and 0.051°C over the land and ocean areas. In URBAN, the value reached 0.84°C, which was also closer to the value from N2U (0.84°C) than to that from U2U (0.46°C). In PRD, the contributions were 0.35° and 0.067°C over land and ocean areas. In URBAN, the value increased to 0.92°C, which was also closer to the value from N2U (0.92°C) than to that from U2U (0.66°C).
Annual urban-related warming in the subregions of the three city clusters BTH, YRD, and PRD for U10 − U80: Average (black), WATER (red), LAND (blue), U2U (green), N2U (pink), and URBAN (brown). One, two, and three asterisks mean data passing the 80%, 90%, and 95% confidence level t tests, respectively.
Citation: Journal of Climate 30, 3; 10.1175/JCLI-D-15-0604.1
In the three city cluster subregions, the annual urban-related warmings that occurred in subregional and land-only areas were low (0.28°–0.43°C). For U2U, contributions were larger (0.46°–0.66°C). The contributions in N2U were much greater (0.84°–1.07°C), which were closer to the values in URBAN (0.84°–1.06°C). Differences among the three subregions might result from the numbers of urban grid cells in the 1980s and the degree of urbanization from U10 − U80.
(ii) In subregions of China
For China as a whole, the annual contribution was less at 0.075°C. Because the urbanization had primarily concentrated in eastern China, urban-related warming in eastern (including NE, NC, EC, and SC), central (including NWE and SW), and western (including NWW and TP) parts of China were 0.14°, 0.056°, and 0.025°C, respectively, which displayed a decreased intensity moving from east to west.
For the eight subregions of China, greater annual urban-related warming could be detected in eastern China, especially in areas where the three city clusters were located (Table 2). The contributions in other subregions of China were smaller.
Annual urban-related warmings (SAT, °C), the precipitation changes (mm day−1), and RFB and RFT (W m−2) in the subregions of China for U10 − U80 (An asterisk means data passing the 80% confidence level t test).
(iii) Across the simulated domain
The annual urban-related warming was much lower across all simulated regions [East Asia (EA), 0.031°C], with values of 0.031° and 0.0085°C over land (EAL) and ocean (EAO), respectively.
Significance tests (the 80%, 90%, and 95% confidence level t tests) on annual SAT changes at different spatial scales were passed mainly in the urban surface areas, especially in N2U and URBAN (Fig. 4, Table 2). However, SAT changes in subregions of China and across the simulated domain, which mainly resulted from the contributions of the urban-surface-expansion-induced SAT changes, were less and were not significant.
2) Seasonal trends
SAT changes revealed marked seasonal variations, which were greatest in summer [June–August (JJA)] and passed the 90% confidence level t test in the urban surface areas of all three city clusters (Fig. 3). There were weakest SAT changes in winter [December–February (DJF)], which passed this significance test only in PRD. In spring [March–May (MAM)] and autumn [September–November (SON)], areas passing the significance test were similar to those from annual results.
Because the contributions to regional warming resulted primarily from urban surface within the three urban cluster subregions, seasonal variations of urban-related warming are further discussed (not shown).
In URBAN, urban-related warming expressed marked seasonal variations: they were higher in JJA and lower in DJF. The maximum (minimum) contribution was 1.2°C (0.86°C) in JJA (DJF) in BTH. The values in PRD were 1.06° and 0.77°C, which were close to those in BTH. For YRD, there was also a maximum (1.23°C) in JJA and a minimum (0.37°C) in DJF.
The differences of urban-related warming between JJA and DJF in URBAN were 0.34° (BTH), 0.86° (YRD), and 0.29°C (PRD). The value in YRD was higher than those in BTH and PRD. Though the urban-related warming in URBAN for JJA in YRD was greater than those in BTH and PRD, the annual contribution was lower (0.84°C) in YRD because of the lower DJF contribution.
3) Interannual variations
Interannual variations for urban-related warming reveal that there were consistent positive contributions due to urban surface expansion in the three city cluster subregions (not shown). However, small negative contributions could be detected in subregions of China and EA, which occurred in 2005 for EA, EAL, China, and NC, as well as during 2004–07 for NE. As a typical monsoon subregion in eastern China (Fu et al. 2000), the negative values and the substantially varied values from year to year may have a connection with the intensity of the monsoon and the corresponding rain belt movement, as well as wind speeds, cloud cover, etc., which need further analysis.
c. Influences on precipitation from U10 − U80
1) Spatial distributions
The spatial distributions of annual precipitation changes (Fig. 5) revealed two slightly intensified rainfall belts. One covered the areas from the northern Indian subcontinent, southwestern China, southeastern China, and the East China Sea to Japan. The other was located in the center of Russia. However, the precipitation was weakened in other areas, especially in the northern part of China.
Spatial distribution of (a),(e),(i),(m) annual and seasonal [(b),(f),(j),(n) MAM; (c),(g),(k),(o) JJA; and (d),(h),(l),(p) SON] precipitation changes in the three city clusters [(a)–(d) BTH, (e)–(h) YRD, and (i)–(l) PRD] and (m)–(p) eastern China for U10 − U80 (mm day−1). The vertical lines denote areas passing the 90% confidence level t test.
Citation: Journal of Climate 30, 3; 10.1175/JCLI-D-15-0604.1
Spatial distributions for annual precipitation changes passed the 90% confidence level t test mainly in BTH and PRD, as well as central Russia. Seasonal changes showed that the precipitation changes were less in winter and passed this significance test only in PRD (not shown). However, areas passing the significance test were wider in JJA in BTH and in SON in PRD.
2) Annual variations
Latitude–month cross sections of monthly mean precipitation from (top) U80 (mm day−1) and (bottom) the corresponding percent difference for U10 − U80 (the gray shading indicates values near 0%) in (a),(d) SR1; (b),(e) SR2; and (c),(f) SR3.
Citation: Journal of Climate 30, 3; 10.1175/JCLI-D-15-0604.1
The latitude–month cross sections of monthly precipitation in the three subregions from U80 (Figs. 6a–c) showed the typical monsoon precipitation. The corresponding differences from U10 − U80 (Figs. 6d–f) revealed marked changes in the precipitation during rainy seasons but smaller changes during dry seasons.
In SR1, changes in the precipitation were clearly different between the south and north; they were intensified to the south of 38°N but weakened to the north, especially between 38° and 48°N. For the intensified precipitation in the south, changes during the shift of monsoon system northward were less than that during the southward retreat of the monsoon system. In SR2, similar distributions to SR1 were shown, with the exception especially near 42°N. In SR3, the precipitation changes were less than those in SR1 and SR2, indicating that the influence of urbanization on the precipitation became weaker in the western part, in which the likely mechanisms for the changes in precipitation are to be discussed later.
3) Subregional characteristics
(i) In the three city clusters
The influences on the annual precipitation were −0.13 (BTH), 0.12 (YRD), and 0.17 mm day−1 (PRD; Fig. 7). In BTH, the contributions were −0.10 and −0.31 mm day−1 on land and in the ocean, respectively. The values in U2U, N2U, and URBAN were −0.026, −0.047, and −0.047 mm day−1. In YRD, the changes were 0.13 and 0.097 mm day−1 over the land and ocean areas. The values in U2U, N2U, and URBAN were 0.18, 0.17, and 0.17 mm day−1. For PRD, the changes were 0.20 and 0.12 mm day−1 over land and ocean areas. The values in U2U, N2U, and URBAN were 0.36, 0.29, and 0.29 mm day−1.
Changes in the annual averaged precipitation in the subregions of the three city clusters BTH, YRD, and PRD for U10 − U80: Average (black), WATER (red), LAND (blue), U2U (green), N2U (pink), and URBAN (brown). An asterisk means data passing the 80% confidence level t test.
Citation: Journal of Climate 30, 3; 10.1175/JCLI-D-15-0604.1
(ii) In subregions of China
In China as a whole, the changes were smaller. The contributions in eastern, central, and western China were 0.015, 0.042, and −0.0030 mm day−1, which were greater in the eastern and central parts because the concentrated urbanization was primarily in eastern China.
For subregions of China (Table 2), the precipitation decreased in the north (NE and NC) and increased in the south (EC and SC). Similar to the distributions in the eastern part of China, the precipitation was intensified in SW and weakened in NWE. In western China, changes in the precipitation were low. In summary, changes in annual precipitation revealed marked subregional characteristics in China, particularly in the East Asian summer monsoon (EASM) subregions containing the three city clusters. As a result, overall precipitation across China increased by only 0.016 mm day−1 (0.65%), even though some subregions showed much greater changes.
Changes in the precipitation over subregions of China from U10 − U80 also showed clear seasonal characteristics, which were greater in JJA, in all subregions except EC (not shown). In EC, changes in the precipitation in SON were stronger, which corresponded to the rain belt movement due to the EASM retreat. Furthermore, the changes in SON over NE, NC, and SC showed secondary seasonal intensity, which was comparable with JJA values. The interannual variations over subregions of China revealed that changes in the precipitation were mainly consistent, in which the prevalences of different signs were 50%, 70%, 70%, 70%, 70%, 80%, 50%, 60%, and 60% in NE, NC, EC, SC, NWE, SW, NWW, TP, and China overall, respectively, based on the 10-yr integration.
(iii) Across the simulated domain
Changes in the annual precipitation across the simulated domain were much less; the values were 0.010 (0.27%), 0.13 (2.37%), and 0.011 mm day−1 (0.28%) over EAL, EAO, and EA, respectively. These showed intensified precipitation in all three subregions, with the largest values over EAO.
Changes in seasonal precipitation were intensified across the four seasons over different subregions, with the exception of the slightly weakened results in DJF over EAL and EA (not shown). Changes in the precipitation were larger in JJA over EAL and EA. The interannual variations over EAL, EAO, and EA revealed that changes in the precipitation were mainly positive, where the prevalences of different signs were 60%, 80%, and 60% for the 10 yr.
Significance tests (the 80%, 90%, and 95% confidence level t tests) on annual precipitation changes at different spatial scales were passed over smaller areas and more weakly than for SAT changes (Figs. 3–5 and 7). Changes in precipitation due to urban surface expansion were significant in BTH and PRD, it being weakened for the former and intensified for the latter, but the changes in YRD were not significant (Fig. 7).
d) The impact on radiation budget from U10 − U80
1) At the surface
(i) In the three city clusters
In BTH (Fig. 8a), the annual radiation forcing at surface [RFB; net radiation from net shortwave (SWB) and net longwave (LWB) flux] was 1.88 W m−2, with values of 1.81 and 2.30 W m−2 on land and in the ocean. The RFB in URBAN increased to 12.0 W m−2. In YRD (Fig. 8c), the annual RFB was 1.41 W m−2, with values of −0.553 and 2.26 W m−2 on land and in the ocean, respectively. The RFB in URBAN increased to 7.31 W m−2. In PRD (Fig. 8e), the annual RFB was 1.20 W m−2, with values of 0.120 and 1.57 W m−2 on land and in the ocean, respectively. The RFB in URBAN increased to 8.42 W m−2.
Changes in radiation budget (W m−2) for U10 − U80 at the (left) surface and (right) top of the atmosphere under different surface characteristics (see the different colors) in the three urban clusters [(a),(b) BTH; (c),(d) YRD; (e),(f) PRD]. One, two, and three asterisks mean data passing the 80%, 90%, and 95% confidence level t tests, respectively, for RFBs and RFTs.
Citation: Journal of Climate 30, 3; 10.1175/JCLI-D-15-0604.1
The RFBs in U2U were less, while much stronger in N2U and closer to that in URBAN, which revealed the main contributions of N2U to subregional RFBs. Further analysis on the corresponding radiation budget components in N2U revealed that the strongly enhanced SWB and weakly decreased LWB resulted in positive RFBs. Changes in the upward component of SWB (SWOB) revealed more negative values for N2U and less negative values in U2U, except for a positive value in BTH. The positive RFBs in N2U and URBAN mainly came from the strongly negative SWOB because of changes in albedo from other land use categories to the urban surface. However, the RFBs in U2U were much weaker because of less change in albedo, which resulted in less change in SWOB.
(ii) In subregions of China
For the whole of China, the RFB was less (0.0323 W m−2). The corresponding values in the eastern, central, and western parts were 0.23, −0.17, and −0.024 W m−2, respectively, which also indicated greater intensity in eastern parts.
For the eight subregions of China (Table 2), the RFBs in the eastern part of China were positive in the north (NE and NC) and negative in the south (EC and SC). Though the RFBs in the central and western parts of China were occasionally stronger than those in the subregions of eastern China, the changes in the radiation budget components were greater for the latter.
(iii) Across the simulated domain
With the rapid urban surface expansion in China, the RFB was also much less across the simulated domain (0.014 W m−2), with values of 0.014 and −0.45 W m−2 over EAL and EAO, respectively.
2) At the top of the atmosphere
(i) In the three city clusters
Changes for the radiation budget at the surface further had an effect on the corresponding values at the top of the atmosphere. The annual radiation forcing at the top [RFT; net radiation from net shortwave (SWT) and net longwave (LWT) flux] due to urban surface expansion at local and regional scales was then quantitatively evaluated, which could be used to make comparisons with the results at global scale from IPCC (IPCC 2013). In BTH (Fig. 8b), the annual RFT was 2.64 W m−2, with values of 2.73 and 2.05 W m−2 on land and in the ocean, respectively. The RFT in URBAN increased to 14.5 W m−2. In YRD (Fig. 8d), the annual RFT was 3.14 W m−2, with values of 4.25 and 0.60 W m−2 on land and in the ocean, respectively. The RFT in URBAN increased to 11.2 W m−2. In PRD (Fig. 8f), the annual RFT was 2.61 W m−2, with values of 3.07 and 1.25 W m−2 on land and in the ocean, respectively. The RFT in URBAN increased to 11.7 W m−2.
(ii) In subregions of China
For China as a whole, the RFT was less at 0.42 W m−2. The corresponding RFTs in eastern, central, and western parts of China were 0.98, 0.16, and −0.0064 W m−2, which revealed weakened RFTs moving from east to west. Therefore, the main contribution to RFT from the eastern part could be further validated.
For the eight subregions of China (Table 2), the RFTs were all positive, except in NWW, which were stronger in NC, EC, and SC, with much higher changes in the radiation budget components. However, the RFTs in other subregions of China were less.
(iii) Across the simulated domain
The RFT was less across the simulated domain (EA, 0.23 W m−2), with values of 0.23 and −0.21 W m−2 for EAL and EAO, respectively. The positive RFTs over EA and EAL resulted from the weakened shortwave (SWOT) and longwave (LWOT) upward flux for EA and EAL.
The 80%, 90%, and 95% confidence level t tests on annual RFB and RFT (Fig. 8, Table 2) at different spatial scales revealed similar results as those for SAT changes, in which areas passing these tests were mainly urban surface areas, especially N2U and URBAN. Changes in SAT, precipitation, RFB, and RFT could be interpreted in terms of urban-surface-expansion-induced changes in albedo and latent and sensible heat flux (Fig. 9) at different spatial scales, which revealed similar results in significance tests as for SAT changes. However, area-averaged changes in total cloud amount were less (not significant) because of the spatial inhomogeneity for the impacts on the cloud amount.
Changes in (a) albedo (percent divided by 100), (b) total cloud amount (tenths divided by 10), and (c) sensible (W m−2) and (d) latent heat fluxes (W m−2) for U10 − U80 in the three urban cluster subregions: Average (black), WATER (red), LAND (blue), U2U (green), N2U (pink), and URBAN (brown).
Citation: Journal of Climate 30, 3; 10.1175/JCLI-D-15-0604.1
e. Changes in albedo, total fractional cloud amount, sensible and latent heat fluxes from U10 − U80
Changes due to urban surface expansion were found to be greater in eastern China, especially in the three city clusters; nonetheless, differences could be detected among different city clusters. Changes for the albedo, total cloud amount, and sensible and latent heat fluxes were further explored under different surface characteristics in the three city clusters, especially in N2U and U2U; the former had marked albedo changes due to the urban surface expansion, whereas the latter had fewer albedo changes due to a nearly unchanging urban surface.
1) In nonurban to urban areas
In N2U (Fig. 9), SWOB strongly weakened with the decreased albedo due to the conversion from other land use categories to the urban surface, and the RFB was positive. With the strong urban-related warming, the strengthened lesser downward component of LWB (LWDB) and greater upward component of LWB (LWOB) resulted in overall decreased LWB. Meanwhile, the sensible (latent) heat flux was strengthened (weakened). However, differences could be detected in the impact on SWB components and total cloud amounts among three subregions.
Given the typical monsoon climate system over East Asia, with the urban-related warming, especially the SAT-increased belt along the eastern coast of China, atmospheric circulation and moisture flux in the East Asian monsoon (EAM) subregions were impacted. The precipitable water (PWAT) in the eastern part of China, as shown in Fig. 11a, demonstrated decreased values in the north and increased values in the south, with less change in between. Therefore, the total cloud amounts were decreased in the north and increased in the south (Fig. 9b), which further had a significant influence on the radiation budget.
In BTH, with the decreased total cloud amount, the downward component of SWB (SWDB) was strengthened. Meanwhile, the decreased albedo resulted in a decreased SWOB. The stronger decreased SWOB and lesser enhanced SWDB resulted in a positive SWB. As for the LWB, with the urban-related warming, the increased surface temperature benefited the enhancement of LWOB, which further contributed to the increasing LWDB through absorption and reemission by the air in the boundary layer. Therefore, the greater strengthened SWB and lesser weakened LWB resulted in a positive RFB.
In PRD, with the increased total cloud amount, SWDB was weakened. Meanwhile, the decreased albedo resulted in the weakened SWOB. The stronger decreased SWOB and less weakened SWDB resulted in a positive SWB. As for the LWB, with the urban-related warming, the increased surface temperature was beneficial to the enhancement of LWOB. Meanwhile, the LWDB was also enhanced, which resulted from both the strengthened LWOB and the reflected LWOB through absorption and reemission by the air (the increased total cloud amount) in the boundary layer. Therefore, the greater strengthened SWB and lesser weakened LWB resulted in an overall positive RFB.
In YRD, changes for the total cloud amount increased little and were between the values of BTH and PRD. With the little increased total cloud amount, SWDB was weakened. The stronger decreased SWOB and lesser weakened SWDB resulted in a positive SWB. Meanwhile, the LWOB was intensified because of the urban-related warming, which further contributed to the LWDB increase in the form of the reflected LWOB through absorption and reemission by the air (the increased total cloud amount) in the boundary layer. Therefore, the greater strengthened SWB and lesser weakened LWB resulted in a positive RFB.
As a result, although the RFBs in the N2U were all positive and the LWB and SWOB components expressed similar distributions, the SWDB increased in BTH but decreased in YRD and PRD because of the changes in the total cloud amount. These features further influenced other radiation budget components and thus affected the RFB.
2) In urban to urban areas
In U2U (Fig. 9), the albedo change was less and negligible. The influence on the radiation budget mainly resulted from changes in the total cloud amount. Urban-related warming resulted in the enhancement of LWOB, which in turn increased LWDB. Furthermore, the LWOB changes were greater than that of the LWDB for all of the U2U areas. For SWB components, the SWDB and SWOB were both strengthened in BTH but weakened in YRD and PRD. The SWDB changes in U2U were similar to those in N2U, while the SWOB changes in U2U were less because of less albedo changes.
In BTH, the increased SWDB and SWOB resulted in a positive SWB. With the greater negative LWB component, the RFB was negative. In YRD and PRD, the weakened SWDB and SWOB resulted in a negative SWB; together with the negative LWB, the RFB was negative.
f. The impact on atmospheric circulations
Given that the precipitation concentrates on JJA and that the urban surface expansion is mainly located in the eastern part of China, the impacts of urban surface expansion on EASM circulation are further discussed here.
1) The circulation fields at 850 hPa
In terms of the influence on zonal wind at 850 hPa (Figs. 10a,b), there was a marked negative-value belt ranging from southwestern China and the East China Sea to Japan, while mainly positive values were presented in other subregions, especially in the low latitudes. The negative-value belt represented the weakened westerly wind, while the positive-value areas revealed the enhanced westerly wind or the weakened easterly wind from the northwestern Pacific to southern China. For changes of meridional wind, similar distributions were detected as those of zonal wind, except for the difference in the South China Sea (Figs. 10c,d). Therefore, the EASM circulation was weakened from southwestern China and the East China Sea to Japan, while slightly strengthened between the South China Sea and southern China, as well as to the south of Japan. These results were consistent with the strengthened southerly winds to the east of 135°E, with the exception of the weakened circulation over northeastern and southwestern Taiwan. The circulation in northern and northeastern China was less enhanced because of the intensified zonal and meridional winds.
(left) JJA mean results for U80 and (right) the corresponding changes for U10 − U80 winds (color scale) for (a),(b) zonal (m s−1); (c),(d) meridional (m s−1); (e),(f) total (m s−1) and vectors and (g),(h) moisture flux (color scale) and vectors (10−3 kg hPa−1 m−1 s−1) at 850 hPa. (i) JJA mean moisture flux (color scale) and vectors for U80 and (j) its changes for U10 − U80 across the whole troposphere (10−3 kg m−1 s−1).
Citation: Journal of Climate 30, 3; 10.1175/JCLI-D-15-0604.1
2) Sea level pressure
The influence of urban surface expansion on sea level pressure (SLP) in JJA shown in Fig. 11c revealed that SLP generally declined over the simulated domain, especially from southwestern China and the East China Sea to Japan. As it was controlled by the continental low over land and the northwestern Pacific subtropical high over the ocean, the negative-value belt was located at the boundary between the two systems, and the decreased SLP limited the southeasterly wind in the southern boundary of the northwestern Pacific subtropical high. These features induced considerable influence on the atmospheric circulation.
Changes in the (a) annual and (b) JJA precipitable water (PWAT, mm) and (c) JJA sea level pressure (hPa) for U10 − U80.
Citation: Journal of Climate 30, 3; 10.1175/JCLI-D-15-0604.1
First, the weakened southeasterly wind to southern China resulted in a decreased contribution to the southwesterly monsoon circulation from southeastern China, the East China Sea, and Japan. Second, the weakened southeasterly wind was accompanied by enhanced southwesterly wind on the northwestern side of the northwestern Pacific subtropical high, which was located to the southeast of the weakened circulation. Third, with the weakened easterly wind located in the low latitudes, the westerly wind from the Bay of Bengal was intensified, but with the decreased SLP, the pressure gradient between the South China Sea and southeastern China was increased, which resulted in the enhanced southerly wind shown by Fig. 10d. The strengthened westerly wind from the Bay of Bengal and the intensified southerly wind from the South China Sea contributed to the enhanced circulation in the south China coastal subregions. In summary, the EASM-related circulation was weakened from southwestern China and the East China Sea to Japan, with the exception of the intensification in the southern China coastal areas, as well as to the east of 135°E.
3) Moisture flux
(i) Precipitable water
In terms of the JJA PWAT changes due to the urban surface expansion from U10 − U80 (Fig. 11b), an increase was apparent across the simulated domain, with the exception of the decrease from northern China to northeastern Asia. The greater increased subregions covered the area from southwestern China and the East China Sea to Japan. In the EASM subregions, the intensified PWAT areas were consistent with the weakened EASM, which limited the northward moisture transport to northern parts of China.
(ii) Moisture flux at 850 hPa
The moisture flux in the EASM system originates from four air currents (Huang et al. 1998). Figures 10g and 10h show that the moisture flux was weakened for the southwesterly and southeasterly branches, but intensified for the southerly branch from the South China Sea and for the westerly current in the mid-to-high latitudes. The moisture flux was intensified along the coast of southern China, which arose from two aspects: the enhanced westerly moisture flux due to the weakened easterly current from the southeastern branch and the intensified southerly branch. These all contributed to the moisture flux changes in southern China, in which the strengthened values were greater than the weakened values. Furthermore, the weakened southeastern moisture flux was accompanied by stronger southwesterly transportation on the northwestern side of the northwestern Pacific subtropical high. A similar distribution to the results at 850 hPa could be found for the results across the whole of the troposphere (Figs. 10i,j).
g. Comparisons on the influences under different urban surface expansion backgrounds
To further evaluate the impact of urban surface expansion, comparisons of the annual urban-related warming and the precipitation changes in the three city clusters under different urban surface backgrounds were performed.
1) SAT changes
In BTH, the increased annual SATs were 0.11°, 0.093°, and 0.17°C for U90 − U80, U00 − U90, and U10 − U00 (Fig. 12), respectively, which resulted in 0.38°C for U10 − U80. For YRD (PRD), the corresponding values were 0.043° (0.0061°), 0.086° (0.051°), and 0.17°C (0.23°C), respectively, which resulted in 0.30°C (0.28°C) for U10 − U80. The maximum increases were from U10 − U00, which was consistent with the rapid development of urban surface shown in Figs. 2a–r and Table 1. However, the minimum increases were from U90 − U80 in YRD and PRD and from U00 − U90 in BTH.
Subregional trends for SAT (°C) changes in the subregions of the three city clusters (a) BTH, (b) YRD, and (c) PRD under different urban surface expansion backgrounds: U90 − U80 (black), U00 − U90 (red), U10 − U00 (blue), and U10 − U80 (green).
Citation: Journal of Climate 30, 3; 10.1175/JCLI-D-15-0604.1
For SAT changes in the U2U, N2U, and URBAN areas, similar distributions could be detected. What should be addressed is that the SAT changes for U10 − U80 were not equal to the added values of U90 − U80, U00 − U90, and U10 − U00 in the urban surface subregions because of the differences in the urban surface grid cells under different urban surface backgrounds.
Subregions passed significance tests (the 80%, 90%, and 95% confidence level t test) on annual SAT changes under different urban surface backgrounds at varied spatial scales, mainly in the urban surface areas, especially in N2U and URBAN (Fig. 12).
2) Changes in precipitation
(i) Changes in precipitation
For annual precipitation under different urban surface expansion backgrounds over EAL, EAO, and EA, the precipitation was increased for U90 − U80 and U10 − U00 and decreased for U00 − U90. Because the increased values were greater than the decreased values, this resulted in slightly strengthened precipitation for U10 − U80.
(ii) Changes in rain belt movement in EAM subregions
The latitude–month cross sections for the changes of monthly precipitation in the EAM subregions under different urban surface expansion backgrounds (Figs. 13a–c) revealed quite different spatial characteristics. Changes in the precipitation were greater from U10 − U00 with quite similar spatial distributions to those from U10 − U80. The precipitation in the north was enhanced from U00 − U90 but decreased from U90 − U80 and U10 − U00. The precipitation changes in the south were also varied; precipitation was enhanced during the northward movement of the rain belt and weakened during the southward retreat of the rain belt from U90 − U80. However, this was contrary to the results from U00 − U90 and U10 − U00.
(top) Latitude–month cross sections for the differences of monthly precipitation (%) in EASM subregions and (bottom) changes of JJA moisture flux (color scale) and vectors at 850 hPa (10−3 kg hPa−1 m−1 s−1) under different urban surface expansion backgrounds: (a),(d) U90 − U80; (b),(e) U00 − U90; and (c),(f) U10 − U00.
Citation: Journal of Climate 30, 3; 10.1175/JCLI-D-15-0604.1
As the precipitation concentrates in JJA, changes in JJA precipitation are further discussed. The precipitation was generally decreased in NE and NC and increased in EC and SC from U90 − U80 and U10 − U00. However, changes from U00 − U90 were opposite to this.
The impact on the EASM circulation (not shown) and moisture flux (Figs. 13d–f) further expressed varied characteristics under different urban surface expansion backgrounds. A negative moisture flux belt occurred over EASM subregions; however, there were differences in the location and intensity for the negative belt. Meanwhile, the westerly moisture fluxes in the mid-to-high latitudes were enhanced from U90 − U80 but were less weakened from U00 − U90 and U10 − U00, with diverse effects on the precipitation changes in eastern China (Figs. 13a–c).
Comparisons among different urban surface expansion backgrounds revealed that the EASM-related moisture flux change from U10 − U80 in the mid-to-low latitudes was quite similar to that from U10 − U00; however, changes in the westerly moisture flux in the mid-to-high latitudes for U10 − U80 (Fig. 10h) were mainly from U90 − U80. Furthermore, the intensity for the moisture flux change was larger from U10 − U00 and lower from U90 − U80 and U00 − U90.
4. Discussion and conclusions
Based on satellite-based urban surface data from the 1980s, 1990s, 2000s, and 2010s in China, the contributions to regional warming and precipitation changes by urban surface expansion were explored using the nested MM5V3.7 with urban effects considered. These are quite different from previous integrations, which were either based on virtual urban surface data (Miao et al. 2010) or focused on a certain city (Ye et al. 2014) or city clusters (N. Zhang et al. 2010) or concentrated on a particular weather event (Miao et al. 2011). The urbanization was concentrated in the eastern part of China, especially in the three city clusters (BTH, PRD, and YRD), which demonstrated a clear and gradual expansion for the first 20 yr followed by a dramatic increase for U10 − U00.
The spatial impact on SAT over East Asia for U10 − U80 revealed that the increased SAT areas were not limited to a single city or city clusters but extended over a SAT-increased belt that covered the eastern coast of China. SAT changes displayed a decreasing intensity moving from east to west.
The annual contribution of the rapid urban surface expansion in China from U10 − U80 to regional warming was small when averaged over the simulated East Asian region (0.031°C) but was slightly larger at 0.075°C over China, in which the impact weakened when moving from east to west. With respect to the subregions of China, greater contributions to regional warming could be detected in eastern China (0.14°C), increasing by 0.042°, 0.20°, 0.18°, and 0.19°C in NE, NC, EC, and SC, respectively. Overall urban-related warming of the three city cluster subregions was 0.38° (BTH), 0.30° (YRD), and 0.28°C (PRD). The annual urban-related warming in areas which were or became urbanized (URBAN) was much greater, at 1.06°, 0.84°, and 0.92°C.
Changes in the SAT for U10 − U80 could be explained by the radiative forcings (RFs) and their components. Similar trends for the surface (RFB) and top-of-atmosphere (RFT) could be detected due to urban surface expansions, though the RF trends were weak when averaged across the simulated East Asian domain. The impact over China was greater, with the greatest intensity observed in eastern China. Even larger values could be detected in the subregions of eastern China, especially in the three city cluster subregions. For annual RFTs as an example, the values were 0.23 W m−2 in East Asia and 0.43 W m−2 in China. The RFT in eastern China (0.98 W m−2) was greater than that in the central and western parts. The RFTs in the three city cluster subregions were much greater, with values of 2.64, 3.14, and 2.61 W m−2 in BTH, YRD, and PRD, respectively, with even higher values of 14.5, 11.2, and 11.7 W m−2 in the urban areas. Therefore, the subregional RFTs, especially in the urban areas, were much greater than global forcings (IPCC 2013). The positive RFBs in N2U originated primarily from the strong SWOB changes arising from large changes in albedo when rural surfaces were urbanized, and differed from the small RFBs in U2U where albedo changes were small. Therefore, the RFs in N2U in the three city clusters were closer to those in URBAN and much greater than those in other areas, which revealed the importance of the contributions of N2U to subregional RFs.
The analysis of the changes in albedo, total cloud amount, and sensible and latent heat fluxes from U10 − U80 helped to explain the impacts of different surface characteristics. In the urban areas, with decreased albedo, the RFBs were enhanced, which resulted in strengthened sensible heat fluxes. Meanwhile, the increased rapid runoff of rainwater from impervious surfaces into drains, as well as decreased vegetation, contributed to weakened latent heat fluxes. Meanwhile, urban-related warming induced increased LWOB, which in turn contributed to the strengthened LWDB. However, differences between city clusters could be detected in the impacts of the SWDB and total cloud amounts, thus affecting other radiation components and the RFB. Both the urban-surface-expansion-related changes in albedo and the total cloud amount contributed to changes in the radiation budget, with significant differences between the U2U and N2U areas, thus affecting the radiation flux.
Because of the warming effect from urban surface expansion, especially on the coast of eastern China, the precipitation was slightly intensified from U10 − U80 over the simulated East Asian land domain (0.010 mm day−1; 0.27%) and over China (0.016 mm day−1; 0.65%). For EASM subregions, the precipitation was weakened in the north but intensified in the south, mainly as a result of EASM southward retreat. The precipitation change also showed reducing influence from urbanization from east to west. Changes in precipitation could be explained by the influence on the EASM circulation.
Although the overall intensity of the seasonal and annual changes was small, there were much greater changes in individual years or seasons. Meanwhile, the analysis of the radiation budget showed that although the overall RFs were generally small, the corresponding radiation budget component changes could be much greater, especially at local scales. Furthermore, the urban influences showed clear subregional, seasonal, and interannual trends, as well as decadal variations due to the different urban surface expansion phases from the 1980s to 2010s.
Based on the 10-yr period nested numerical simulations with fine-resolution urban surface data, the impacts on regional climate at different spatial and temporal scales due to the rapid spread of urbanization in the eastern China have been discussed. Such methods were essential for assessing urban-related warming and changes in the precipitation, in which the use of fine model grid cells allowed for more detailed spatial information to be obtained. Although the satellite-derived urban surface data were able to concisely represent the urban surface expansion, there might be errors between the retrieved results and the actual urban surface distribution because of the limitations of the remote sensing technology and retrieval methods (Hu et al. 2015), which will need much improvement and further validation.
Acknowledgments
This work was supported by the Chinese Academy of Sciences Strategic Priority Program under Grant XDA05090206, the National Key Research and Development Program of China under Grant 2016YFA0600403, the Chinese Natural Science Foundation (41275162), the National Key Basic Research Program on Global Change under Grant 2011CB952003, and the Jiangsu Collaborative Innovation Center for Climatic Change. The authors also thank Y. H. Hu from the Institute of Remote Sensing and Digital Earth (funded by the project XDA05090203) and G. S. Jia from the Institute of Atmospheric Physics (funded by the project XDA05090201), Chinese Academy of Sciences for the urban surface data. The authors thank the reviewers for their numerous valuable comments to improve the manuscript.
REFERENCES
Barth, A., A. Alvera-Azcarate, M. Rixen, and J. M. Beckers, 2005: Two-way nested model of mesoscale circulation features in the Ligurian Sea. Prog. Oceanogr., 66, 171–189, doi:10.1016/j.pocean.2004.07.017.
Bukovsky, M. S., and D. J. Karoly, 2009: Precipitation simulations using WRF as a nested regional climate model. J. Appl. Meteor. Climatol., 48, 2152–2159, doi:10.1175/2009JAMC2186.1.
Changnon, S. A., R. T. Shealy, and R. W. Scott, 1991: Precipitation changes in fall, winter, and spring caused by St. Louis. J. Appl. Meteor., 30, 126–134, doi:10.1175/1520-0450(1991)030<0126:PCIFWA>2.0.CO;2.
Chen, F., and J. Dudhia, 2001a: Coupling an advanced land-surface/hydrology model with the Penn State/NCAR MM5 modeling system. Part I: Model implementation and sensitivity. Mon. Wea. Rev., 129, 569–585, doi:10.1175/1520-0493(2001)129<0569:CAALSH>2.0.CO;2.
Chen, F., and J. Dudhia, 2001b: Coupling an advanced land-surface/hydrology model with the Penn State/NCAR MM5 modeling system. Part II: Preliminary model validation. Mon. Wea. Rev., 129, 587–604, doi:10.1175/1520-0493(2001)129<0587:CAALSH>2.0.CO;2.
Chen, H., and Y. Zhang, 2013: Sensitivity experiments of impacts of large-scale urbanization in East China on East Asian winter monsoon. Chin. Sci. Bull., 58, 809–815, doi:10.1007/s11434-012-5579-z.
Dong, X. C., J. P. Tang, Y. Wang, and J. W. Yu, 2008: Numerical simulation of the climate effect of the urbanzation course along the middle and lower reaches of Yangtze River (in Chinese). Sci. Meteor. Sin., 28, 147–154.
Dudhia, J., 2005: MM5 version 3.7 (the final version). Sixth WRF/15th MM5 Users’ Workshop, Boulder, CO, NCAR, 1.1. [Available online at http://www2.mmm.ucar.edu/wrf/users/workshops/WS2005/abstracts/Session1/1-Dudhia.pdf.]
Fang, C. L., 2009: The urbanization and urban development in China after the reform and opening-up (in Chinese). Econ. Geogr., 29, 19–25.
Findell, K. L., E. Shevliakova, and P. C. D. Milly, 2007: Modeled impact of anthropogenic land cover change on climate. J. Climate, 20, 3621–3634, doi:10.1175/JCLI4185.1.
Fu, C. B., H. L. Wei, Y. Qian, and M. Chen, 2000: Documentation on regional integrated environmental model system (RIEMS version 1). TEACOM Science Rep. 1, START Regional Committee for Temperate East Asia, 26 pp.
Gallo, K., T. Owen, D. Easterling, and P. F. Jamason, 1999: Temperature trends of the U.S. Historical Climatology Network based on satellite-designated land use/land cover. J. Climate, 12, 1344–1348, doi:10.1175/1520-0442(1999)012<1344:TTOTUS>2.0.CO;2.
Gibbard, S., K. Caldeira, G. Bala, T. J. Philips, and M. Wickett, 2005: Climate effects of global land cover change. Geophys. Res. Lett., 32, L23705, doi:10.1029/2005GL024550.
Grell, G. A., J. Dudhia, and D. R. Stauffer, 1994: A description of the fifth generation Penn State/NCAR Mesoscale Model (MM5). NCAR Tech. Note NCAR/TN-398+STR, 121 pp., doi:10.5065/D60Z716B.
Guo, Y.-R., and S. Chen, 1994: Terrain and land use for the fifth-generation Penn State/NCAR Mesoscale Modeling System (MM5): Program TERRAIN. NCAR Tech. Note NCAR/TN-397+IA, 114 pp., doi:10.5065/D68C9T67.
Han, Z. Q., Z. W. Yan, Z. Li, W. D. Liu, and Y. C. Wang, 2014: Impact of urbanization on low-temperature precipitation in Beijing during 1960–2008. Adv. Atmos. Sci., 31, 48–56, doi:10.1007/s00376-013-2211-3.
Hansen, J., R. Ruedy, M. Sato, and K. Lo, 2010: Global surface temperature change. Rev. Geophys., 48, RG4004, doi:10.1029/2010RG000345.
Hong, S.-Y., and H.-L. Pan, 1996: Nonlocal boundary layer vertical diffusion in a medium-range forecast model. Mon. Wea. Rev., 124, 2322–2339, doi:10.1175/1520-0493(1996)124<2322:NBLVDI>2.0.CO;2.
Hu, Y. C., W. J. Dong, and Y. He, 2010: Impact of land surface forcings on mean and extreme temperature in eastern China. J. Geophys. Res., 115, D19117, doi:10.1029/2009JD013368.
Hu, Y. H., and Coauthors, 2015: Improved monitoring of urbanization processes in China for regional climate impact assessment. Environ. Earth Sci., 73, 8387, doi:10.1007/s12665-014-4000-4.
Huang, R. H., Z. Z. Zhang, G. Huang, and B. H. Ren, 1998: Characteristics of the water vapor transport in East Asian monsoon region and its differences from that in South Asian monsoon regions in summer (in Chinese). Sci. Atmos. Sin., 22, 460–469.
Ichinose, T., K. Shimodozonob, and K. Hanakib, 1999: Impact of anthropogenic heat on urban climate in Tokyo. Atmos. Environ., 33, 3897–3909, doi:10.1016/S1352-2310(99)00132-6.
IPCC, 2013: Climate Change 2013: The Physical Science Basis. Cambridge University Press, 1535 pp., doi:10.1017/CBO9781107415324.
Jia, G. S., R. H. Xu, Y. H. Hu, and Y. T. He, 2014: Multi-scale remote sensing estimates of urban fractions and road widths for regional models. Climatic Change, 129, 543–554, doi:10.1007/s10584-014-1114-3.
Jiang, X. Y., C. L. Zhang, H. Gao, and S. G. Miao, 2007: Impacts of urban albedo change on urban heat island in Beiing—A case study (in Chinese). Acta Meteor. Sin., 65, 301–307.
Jones, P. D., D. H. Lister, and Q. Li, 2008: Urbanization effects in large-scale temperature records, with an emphasis on China. J. Geophys. Res., 113, D16122, doi:10.1029/2008JD009916.
Kain, J. S., and J. M. Fritsch, 1993: Convective parameterization for mesoscale models: The Kain-Fritsch scheme. The Representation of Cumulus Convection in Numerical Models, Meteor. Monogr., No. 46, Amer. Meteor. Soc., 165–170.
Kalnay, E., and M. Cai, 2003: Impact of urbanization and land-use change on climate. Nature, 423, 528–531, doi:10.1038/nature01675.
Li, Q. X., W. Li, P. Si, X. R. Gao, W. J. Dong, P. D. Jones, J. Y. Huang, and L. J. Cao, 2010: Assessment of surface air warming in northeast China, with emphasis on the impacts of urbanization. Theor. Appl. Climatol., 99, 469–478, doi:10.1007/s00704-009-0155-4.
Li, Y., Y. H. Song, A. Mochida, and T. Okaze, 2014: WRF environment assessment in Guangzhou city with an extracted land-use map from the remote sensing data in 2000 as an example. J. Harbin Inst. Technol., 21, 26–32.
Lin, C.-Y., W.-C. Chen, S. C. Liu, Y. A. Liou, G. R. Liu, and T. H. Lin, 2008: Numerical study of the impact of urbanization on the precipitation over Taiwan. Atmos. Environ., 42, 2934–2947, doi:10.1016/j.atmosenv.2007.12.054.
Liu, J. Y., M. L. Liu, H. Q. Tian, D. F. Zhuang, Z. X. Zhang, W. Zhang, X. M. Tang, and X. Z. Deng, 2005: Spatial and temporal patterns of China’s cropland during 1990–2000: An analysis based on Landsat TM data. Remote Sens. Environ., 98, 442–456, doi:10.1016/j.rse.2005.08.012.
Liu, J. Y., and Coauthors, 2014: Spatiotemporal characteristics, patterns, and causes of land-use changes in China since the late 1980s. J. Geogr. Sci., 24, 195–210, doi:10.1007/s11442-014-1082-6.
Liu, W. D., B. L. Zhang, H. L. You, and P. Yang, 2014: Preliminary analysis of urbanization effects on temperature change in Beijing during 1978–2008 (in Chinese). Meteor. Mon., 40, 94–100.
Liu, Z. H., Y. Li, and J. Peng, 2011: Progress and perspective of the research on hydrological effects of urban impervious surface on water environment (in Chinese). Progress Geogr., 30, 275–281.
Liu, Z. X., S. H. Liu, F. Hu, J. Li, Y. J. Ma, and H. P. Liu, 2011: A comparison study of the simulation accuracy between WRF and MM5 in simulating local atmospheric circulations over greater Beijing. Sci. China Earth Sci., 55, 418–427, doi:10.1007/s11430-011-4310-2.
Loveland, T. R., B. C. Reed, J. F. Brown, D. O. Ohlen, Z. Zhu, L. Yang, and J. W. Merchant, 2000: Development of a global land cover characteristics database and IGBP-DISCover from 1 km AVHRR data. Int. J. Remote Sens., 21, 1303–1330, doi:10.1080/014311600210191.
Miao, S. G., F. Chen, Q. C. Li, and S. Y. Fan, 2010: Month-averaged impacts of urbanization on atmospheric boundary layer structure and precipitation in summer in Beijing area (in Chinese). Chin. J. Geophys., 53, 1580–1593.
Miao, S. G., F. Chen, Q. C. Li, and S. Y. Fan, 2011: Impacts of urban processes and urbanization on summer precipitation: A case study of heavy rainfall in Beijing on 1 August 2006. J. Appl. Meteor. Climatol., 50, 806–825, doi:10.1175/2010JAMC2513.1.
Mlawer, E. J., S. J. Taubman, P. D. 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, doi:10.1029/97JD00237.
Mote, T. L., M. C. Lacke, and J. M. Shepherd, 2007: Radar signatures of the urban effect on precipitation distribution: A case study for Atlanta, Georgia. Geophys. Res. Lett., 34, L20710, doi:10.1029/2007GL031903.
Pielke, R. A., Sr., R. Avissar, M. Raupach, A. J. Dolman, X. Zeng, and A. S. Denning, 1998: Interactions between the atmosphere and terrestrial ecosystems: Influence on weather and climate. Global Change Biol., 4, 461–475, doi:10.1046/j.1365-2486.1998.t01-1-00176.x.
Portman, D., 1993: Identifying and correcting urban bias in regional time series: Surface temperature in China’s northern plains. J. Climate, 6, 2298–2308, doi:10.1175/1520-0442(1993)006<2298:IACUBI>2.0.CO;2.
Ren, G. Y., Y. Zhou, Z. Chu, J. Zhou, A. Zhang, J. Guo, and X. Liu, 2008: Urbanization effects on observed surface air temperature trends in north China. J. Climate, 21, 1333–1348, doi:10.1175/2007JCLI1348.1.
Ren, K., J. He, and L. Qi, 2010: The establishment characteristics of the East Asian subtropical monsoon rain-belt and an analysis of its nature of precipitation (in Chinese). Acta Meteor. Sin., 68, 550–558.
Ren, Y. Y., and G. Y. Ren, 2011: A remote-sensing method of selecting reference stations for evaluating urbanization effect on surface air temperature trends. J. Climate, 24, 3179–3189, doi:10.1175/2010JCLI3658.1.
Shao, H., J. Song, and H. Ma, 2013: Sensitivity of the East Asian summer monsoon circulation and precipitation to an idealized large-scale urban expansion. J. Meteor. Soc. Japan, 91, 163–177, doi:10.2151/jmsj.2013-205.
Shem, W., and M. Shepherd, 2009: On the impact of urbanization on summertime thunderstorms in Atlanta: Two numerical model case studies. Atmos. Res., 92, 172–189, doi:10.1016/j.atmosres.2008.09.013.
Shepherd, J. M., H. Pierce, and A. J. Negri, 2002: Rainfall modification by major urban areas: Observation from space borne rain radar on the TRMM satellite. J. Appl. Meteor., 41, 689–701, doi:10.1175/1520-0450(2002)041<0689:RMBMUA>2.0.CO;2.
Skamarock, W. C., and Coauthors, 2008: A description of the Advanced Research WRF version 3. NCAR Tech. Note NCAR/TN-475+STR, 113 pp., doi:10.5065/D68S4MVH.
Thielen, J., W. Wobrock, A. Gadian, P. G. Mestayer, and J.-D. Creutin, 2000: The possible influence of urban surfaces on rainfall development: A sensitivity study in 2D in the meso-gamma scale. Atmos. Res., 54, 15–39, doi:10.1016/S0169-8095(00)00041-7.
Wang, B., and H. Lin, 2002: Rainy season of the Asian–Pacific summer monsoon. J. Climate, 15, 386–398, doi:10.1175/1520-0442(2002)015<0386:RSOTAP>2.0.CO;2.
Wang, F., and Q. S. Ge, 2012: Estimation of urbanization bias in observed surface temperature change in China from 1980 to 2009 using satellite land-use data. Chin. Sci. Bull., 57, 1708–1715, doi:10.1007/s11434-012-4999-0.
Weaver, C. P., and R. Avissar, 2001: Atmospheric disturbances caused by human modification of the landscape. Bull. Amer. Meteor. Soc., 82, 269–2281, doi:10.1175/1520-0477(2001)082<0269:ADCBHM>2.3.CO;2.
Yan, Z. W., Z. Li, Q. X. Li, and P. D. Jones, 2010: Effects of site change and urbanisation in the Beijing temperature series 1977–2006. Int. J. Climatol., 30, 1226–1234, doi:10.1002/joc.1971.
Yan, Z. W., J. Wang, Z. Li, and W. D. Liu, 2014: Effects of urbanization based on homogenized daily observations (in Chinese). Adv. Meteor. Sci. Tech., 4, 41–48.
Yang, X. C., Y. L. Hou, and B. D. Chen, 2011: Observed surface warming induced by urbanization in east China. J. Geophys. Res., 116, D14113, doi:10.1029/2010JD015452.
Yang, Y. J., and Coauthors, 2013: Impacts of urbanization and station-relocation on surface air temperature series in Anhui Province, China. Pure Appl. Geophys., 170, 1969–1983, doi:10.1007/s00024-012-0619-9.
Ye, L. M., Z. H. Jiang, and F. Huo, 2014: Numerical simulation of the impact of land cover change on the urban heat island effect in Nanjing (in Chinese). Trans. Atmos. Sci., 37, 642–652.
Zhang, A. Y., G. Y. Ren, J. X. Zhou, Z. Y. Chu, Y. Y. Ren, and G. L. Tang, 2010: Impacts of urbanization on trends of surface air temperature change in China (in Chinese). Acta Meteor. Sin., 68, 957–966.
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, doi:10.1007/s00704-010-0263-1.
Zhao, D. M., 2012: Performance of Regional Integrated Environment Modeling System (RIEMS) in the simulation of surface air temperature over East Asia. Atmos. Ocean. Sci. Lett., 5, 145–150, doi:10.1080/16742834.2012.11446981.
Zhao, D. M., 2013: Performance of Regional Integrated Environment Modeling System (RIEMS) in precipitation simulations over East Asia. Climate Dyn., 40, 1767–1787, doi:10.1007/s00382-012-1660-1.
Zhao, D. M., B. K. Su, and J. P. Tang, 2007: Testing the ability of numerical model to simulate climate and its change with 4D-EOF. Acta Meteor. Sin., 21, 179–191.
Zhao, D. M., C. B. Fu, and X. D. Yan, 2009: Testing the ability of RIEMS2.0 (Regional Integrated Environment Modeling System) to simulate multi-year precipitation and air temperature in China. Chin. Sci. Bull., 54, 3101–3111, doi:10.1007/s11434-009-0178-3.
Zhao, J., A. D. Gong, and J. Xu, 2008: Progress in study on urban heat island at home and abroad (in Chinese). J. Nat. Disasters, 17, 64–67.
Zhou, J. P., Z. B. Sun, D. H. Ni, and Z. X. Li, 2013: Impact of meteorological station relocation on homogeneity of annual mean temperature in China (in Chinese). Trans. Atmos. Sci., 36, 139–146.
Zhou, L., R. E. Dickinson, Y. Tian, J. Fang, Q. Li, R. K. Kaufmann, C. J. Tucker, and R. B. Myneli, 2004: Evidence for a significant urbanization effect on climate in China. Proc. Natl. Acad. Sci. USA, 101, 9540–9544, doi:10.1073/pnas.0400357101.
Zhuang, B., and Coauthors, 2014: Optical properties and radiative forcing of urban aerosols in Nanjing, China. Atmos. Environ., 83, 43–52, doi:10.1016/j.atmosenv.2013.10.052.
