1. Introduction
In the past three decades, the Tibetan Plateau (TP) has been warming at a faster pace than the global average temperature (Wu et al. 2007; Kang et al. 2010). The warming has already exerted great impacts on the cryosphere and water cycle in the TP (Solomon et al. 2007; Krause et al. 2010; Moore 2012; Gao et al. 2014) that feed major rivers supporting the large population and economic development in Asia. Complex topography has a dominating influence on the TP climatology (Xu et al. 2008; Wu et al. 2013), but observations are sparse in the expansive plateau due to the harsh conditions. With observation sites predominantly located in the eastern mountain valley with accessibility, the vast northwestern TP is mostly devoid of observations. Therefore, climate change analysis is severely limited by poorly gauged or ungauged conditions. Although global reanalyses are appropriate tools for assessing past climate change, they have limited value in the TP because global models are not well constrained by observations and they lack sufficient spatial resolution to represent the large spatial heterogeneity associated with topography (Fig. 1a), land cover (Fig. 1b), and soil that influence atmospheric and land processes. Land cover in the TP is dominated by grassland, shrubland, and sparse vegetation or barren ground, denoted as land cover types 7, 9, and 19, respectively, in Fig. 1b. These land copy types are distributed from the southeast to the northwest, with a corresponding decrease in leaf area index (LAI; Fig. 1c).
(a) Topography (unit: m) and (b) land cover types (LU_index) and leaf area index (LAI) over the TP. In (a), 1 stands for the Tarim River basin (TRB), 2 for Qilian Mountain (QLB), 3 for the Qaidam basin (QDB), 4 for the Chang Tang Plateau (CTB), 5 for the Yangtze River (YTR), 6 for the Yellow River (YLR), 7 for the Mekong River (MKR), 8 for the Salween River (SWR), and 9 for the Brahmaputra River (BPR). In (b), land cover types 7, 9, and 19 respectively denote grassland, shrubland, and sparse vegetation or barren ground.
Citation: Journal of Climate 28, 10; 10.1175/JCLI-D-14-00581.1
Observations and global reanalyses of the net precipitation [i.e., precipitation − evapotranspiration (P − E)] revealed a trend toward drier and wetter conditions over the humid southeastern TP and the vast arid northwestern TP, respectively (Yang et al. 2011; Yin et al. 2012; Gao et al. 2014). Among four global reanalysis products analyzed, Gao et al. (2014) found that ERA-Interim (ERA-Int) best captured the general moistening in the northwestern TP and drying in the southeastern TP in the past three decades, but a large discrepancy is still notable. For example, ERA-Int does not capture the decrease of P − E in the Yellow River headwater and the Rouergai plateau located in the east. Also the contrast in P − E changes between the northwestern and southeastern TP is not well replicated. None of the global reanalyses yields a statistically significant pattern correlation with the observed changes at the 70% confidence level. Since the headwater of major rivers originating from the TP is widely distributed across the eastern and central plateau, uncertainty in estimating the spatial distribution of P − E changes limits the usefulness of the global reanalysis products in understanding the mechanisms responsible for historical changes across the broad Asian region.
Regional climate models (RCMs) are useful for a host of climate change applications that require finer-scale information. They have been widely used to study historical and projected regional climate changes (Mearns et al. 2009; Nikulin et al. 2012; Déqué et al. 2005; Duffy et al. 2006; Giorgi et al. 1992; Kim et al. 2002; Leung et al. 2003a,b; Plummer et al. 2006; Zhang et al. 2009; Laprise et al. 1998) and are generally shown to provide useful skill in simulating surface air temperature, precipitation, and water resources in East Asia (Gao et al. 2008, 2011; X. Gao et al. 2012; Xu and Gao 2014). A recent 33-yr-long regional climate simulation has been conducted over East Asia and comparisons with observations show that the high-resolution simulation not only improves the pattern correlations but also reproduces finer-scale changes in the surface air temperature in the wet season over the TP compared to the global reanalyses (Gao et al. 2015). This study analyzes the precipitation (P), evapotranspiration (E), and net precipitation (P − E) changes from the regional simulation between 1979 and 2011. Differences between the high-resolution simulation and the coarse-resolution reanalysis that provided meteorological forcing are investigated to explore the mechanisms behind the regional-scale P − E changes. Section 2 introduces the methods and data used. Results are presented in section 3, with conclusions and discussion provided in section 4.
2. Data and methods
The Weather Research and Forecasting (WRF) Model (http://www.wrf-model.org/index.php; Skamarock et al. 2005) was used in the high-resolution regional climate simulation. WRF is a nonhydrostatic model with various choices of physics parameterizations suitable for applications across a wide range of scales. The simulation was performed at 30-km horizontal grid resolution with 159 × 196 grids cells covering East Asia for 1 January 1979 to 31 December 2011 using lateral boundary conditions and SST interpolated from ERA-Int (Dee and Uppala 2009; Dee et al. 2011) since ERA-Int ranks the best among the examined reanalyses in describing temperature and elements of the water cycle over the TP (Wang and Zeng 2012; Li et al. 2012; Gao et al. 2014). ERA-Int is an improved version of the ERA-40 reanalysis (Simmons et al. 2006). The ERA-Int archive maintained by the Computational and Information Systems Laboratory (CISL) at the National Center for Atmospheric Research (NCAR) has a horizontal resolution of 0.7° and available at 6-hourly intervals. The NCAR Community Atmospheric Model (CAM) shortwave and longwave scheme (Collins et al. 2004), the single-moment three-class cloud microphysics (WSM3), Grell–Devenyi ensemble cumulus convective scheme (Grell 1993), the Yonsei University (YSU) boundary layer scheme (Hong and Pan 1996), and the Noah land surface model (Chen and Dudhia 2001) were used in this simulation. The regional simulation produced model outputs at 3-hourly intervals for analysis of moisture budgets. Results of the regional simulation are interpolated onto the grid cells of ERA-Int for plotting.
The Global Land Data Assimilation Systems dataset (GLDAS; Rodell et al. 2004) with a spatial resolution of 1 degree is used as a reference for P − E changes. GLDAS assimilated a variety of conventional data (radiosonde, buoy, ship, and airborne) and satellite-derived observations using a four-dimensional multivariate approach (Rodell et al. 2004). Near-real-time satellite-derived precipitation data were obtained from the U.S. Naval Research Laboratory (NRL) and Goddard Space Flight Center (GSFC). NRL’s precipitation fields were based on geostationary satellite infrared (IR) cloud-top temperature measurements and microwave observation techniques (Turk et al. 2000). The microwave product merged data from the Special Sensor Microwave/Imager (SSM/I), NASA–JAXA (Japan Aerospace Exploration Agency)’s Tropical Rainfall Measuring Mission (TRMM), and the Advanced Microwave Sounding Unit (AMSU) instruments. GSFC’s precipitation fields were based on optimal merging of microwave measurements from the Advanced Microwave Scanning Radiometer (AMSR) and the Advanced Earth Observation Satellite II (AdEOS-II) besides SSM/I and TRMM with the more frequent IR measurements from the international constellation of geosynchronous-Earth-orbit (GEO) satellites (Huffman et al. 2003). Driven by observed temperature and precipitation, GLDAS captures both daily and monthly precipitation variability and exhibits satisfactory performance for many surface variables (Wang and Zeng 2012). Most importantly, the GLDAS product replicates the observed contrast in P − E changes between the northwestern and southeastern TP (Gao et al. 2014). The ensemble mean of the multiland model products is used as the reference E in this study.
3. Results
a. P − E changes
For comparison, P − E changes from GLDAS and ERA-Int are presented alongside those from WRF and the results are shown in Fig. 2. The difference P − E exhibits general increases in the vast northwestern TP and decreases in the southeastern TP in all three datasets. However, the pattern of P − E changes in the WRF simulation (Fig. 2h) resembles GLDAS (Fig. 2g) more than ERA-Int (Fig. 2i) does. The pattern correlation of P − E changes between GLDAS and WRF is 0.17 (Table 1), which passes the statistical significance t test at a 98% confidence level. In testing the statistical significance of the pattern correlation, we estimated the degree of freedom based on the autocorrelation length of the dataset following Leith (1973). In contrast, the correlation (0.04) between GLDAS and ERA-Int does not pass statistical significance even at 70% confidence level, although ERA-Int is already reproducing the observed P − E climatology and trend in the TP better than other popular reanalyses (Gao et al. 2014). Most notably, the WRF simulation captures the larger increase in the Qiangtang plateau over the northwest comparable to GLDAS, but ERA-Int shows increases that are more evenly distributed over the TP and a slight positive gradient from the northwestern to southeastern TP (Fig. 2i). Furthermore, the decrease in the southeastern TP has smaller values and spreads across the lower reaches of the Upper Brahmaputra River basin in WRF (Fig. 2h) compared to the larger, more concentrated decreases in ERA-Int (Fig. 2i). Interestingly, the pattern correlation coefficients between P − E in GLDAS with HGT and MODIS annual mean LAI averaged from 2001 to 2012 are 0.20 and −0.27, respectively. The positive correlation of the GLDAS P − E change with topography indicates larger P − E increase at higher elevation, while the negative correlation of the GLDAS P − E change with LAI suggests that P − E increases more over the dry land cover type (type 19) with smaller LAI in GLDAS. These relationships of P − E changes in GLDAS with elevation and LAI are well captured by the WRF simulations showing high pattern correlations of P − E with topography (Fig. 1a) and LAI (Fig. 1c). However none of these relationships is captured in ERA-Int (Table 1). This suggests a potential shift toward topography and vegetation distribution as a stronger driver of P − E changes at regional scale, and further supports the importance of resolving topography and land cover using models with higher resolution to simulate regional changes in P − E.
Changes in (a)–(c) P, (d)–(f) E, and (g)–(i) P − E (unit: mm day−1) of 1998–2011 compared to 1979–97 for OBS/GLDAS, WRF simulation, and ERA-Int, respectively, over the TP.
Citation: Journal of Climate 28, 10; 10.1175/JCLI-D-14-00581.1
Pattern correlations between annual P − E change in GLDAS and ERA-Int/WRF simulation in 1998–2011 compared to 1979–97 and pattern correlations between ERA-Int/WRF simulation with topography (HGT) and leaf area index (LAI) climatology in 2000–11. The degree of freedom is 189. Correlation coefficients with an asterisk are statistically significant at 98% confidence level based on a two-tailed t test, and the same applies in other tables.
Comparing the P and E changes with P − E changes separately for GLDAS, WRF, and ERA-Int (Figs. 2a–f), P changes significantly contribute to the P − E change pattern in all three datasets (Table 2). Furthermore, Table 2 shows that the P − E changes are explained much more by the P changes than the E changes in GLDAS and WRF, but in ERA-Int the P − E changes explained by the P and E changes are more comparable. Like most GCMs, ERA-Int greatly overestimates P in the TP (Fig. 2c) but the wet biases at high elevation are reduced in WRF at high resolution (Fig. 2b). Similar bias reduction is also found in P − E in many subregions of the TP except QDB and QLB (Figs. 2h, 3, and 1a; regions are defined in Fig. 1). Larger improvements are most notable in the southern (Figs. 3a–d) rather than the northern TP (Figs. 3e–i), hinting that the WRF improvements may be related to better representation of the sharp topographic gradient associated with the Himalayas in the southern slope and its impacts on moisture transport and precipitation.
Pattern correlations between annual P − E change with P and E changes in GLDAS and ERA-Int/WRF simulation in 1998–2011 compared to 1979–97. Correlation coefficients with an asterisk are statistically significant at 98% confidence level based on a two-tailed t test, and the same applies in other tables.
Time series of annual mean P, E, and P − E (unit: mm day−1) for nine subregions labeled in Fig. 1a corresponding mostly to the headwater of major rivers in Asia that originate from the TP.
Citation: Journal of Climate 28, 10; 10.1175/JCLI-D-14-00581.1
b. P − E changes contributions
Figures 4a–h present the P − E changes estimated from the column integrated moisture flux convergence and its three contributors (δMCD, δTH, and δTE) in the WRF simulation and ERA-Int. As demonstrated in Gao et al. (2014), changes in the mean circulation dynamics (δMCD) contribute the most to P − E changes (Figs. 4b,d). The WRF simulation presents the same predominant contribution of δMCD to P − E changes (Figs. 4a,c), which explains 41% of the P − E change pattern (Table 3). Unlike ERA-Int, however, changes in the thermodynamics (δTH) and the transient eddy (δTE) also contribute importantly to P − E changes in the high-resolution simulation (Figs. 4e,g and Table 3 vs Figs. 4f,h). Pattern correlations between δTH and δTE with the P − E changes are 0.40 and −0.26, respectively, both statistically significant at 99.9% confidence level (Table 3).
Distributions of annual (a),(b) P − E changes and contributions from (c),(d) mean circulation dynamics (δMCD), (e),(f) thermodynamics (δTH), and (g),(h) transient eddy (δTE) from (left) the WRF simulation and (right) ERA-Int in 1998–2011 compared to 1979–1997 (Unit: mm day−1).
Citation: Journal of Climate 28, 10; 10.1175/JCLI-D-14-00581.1
Pattern correlations between annual P − E changes and annual changes in mean circulation contribution (δMCD), thermodynamic contribution (δTH), and transient eddy contribution (δTE) in WRF simulation in 1998–2011 compared to 1979–97. Correlation coefficients with an asterisk are statistically significant at 98% confidence level based on a two-tailed t test, and the same applies in other tables.
1) δTE
Having demonstrated important differences between ERA-Int and the regional simulation in their depiction of P − E changes and the contributors to those changes, it would be interesting to contrast how moisture budget terms differ between regional models and their global counterparts in the TP and the Rocky Mountains, two important mountain ranges in the midlatitudes. Y. Gao et al. (2012) found a similar opposite contribution of transient eddy changes to P − E changes in the high-resolution simulation over the Rocky Mountains region, which offsets the drying trend in the southwestern United States under climate change, but in the global models the transient eddy changes contribute to the drying trend, exacerbating the vulnerability of the southwestern United States to droughts in the global model projections. For the TP, in contrast, δTE in the regional simulation offsets the wetter trend in the northwestern TP and drier trend in the southeastern TP in the historical period, but it has almost no contribution to the P − E changes in ERA-Int. Previous studies of extreme precipitation changes have reported an upward trend in the southern TP and a downward trend in the central TP (Wu et al. 2013; You et al. 2008). Since transient eddy changes are associated with changes in storms, similarities between the distributions of δTE changes in the WRF simulation with observed pattern of extreme precipitation changes suggest that the WRF simulated δTE changes may be more realistic than those of ERA-Int.
In the Rocky Mountains, the difference in the transient eddy changes dominates the difference in P − E changes between the high- and coarse-resolution moisture budgets (Gao 2012a). However, in the TP, differences in the thermodynamic changes also contribute to differences in P − E changes between the high-resolution simulation and the coarse-resolution forcing. In the narrower, north–south-oriented Rocky Mountains with prevailing westerly winds, higher resolution has important effects on a model’s ability to simulate orographic uplift and partitioning of precipitation between upwind and leeside along the winter storm tracks. Hence, it is not surprising that the primary difference between the high- and coarse-resolution moisture budgets is related to the transient eddy changes. However, the TP is a massive highland with altitude above 5000 m across a vast mountaintop, with its annual water cycle changes dominated more by the warm season changes (Gao et al. 2014). Unlike the Rocky Mountains where the main difference between the regional and global simulations lies in the transient eddies changes, in the TP the regional simulation and ERA-Int differ in both changes in transient eddies and thermodynamic effects.
2) δMCD and δTH
Figure 5 shows the annual changes of the four terms and Table 4 summarizes the pattern correlation coefficients between the changes in mean circulation/thermodynamics (δMCD/δTH) and the contributions due to the advection terms (δMCDA/δTHA) and the convergence or divergence terms (δMCDD/δTHD), respectively. The correlation between δMCD and δMCDA is negative (Table 4) so the changes in the advection of the mean circulation tend to offset the MCD changes. This is also apparent from Figs. 4c and 5a. However, such offsetting is absent in ERA-Int (Figs. 4d and 5b), which leaves a concentrated region of large MCD changes over the southern TP. Advection of the mean circulation change moisturizes almost the whole TP except for the upper reaches of the Tarim basin and lower reaches of the Brahmaputra River basin in the WRF simulation (Fig. 5a). Given the high pattern correlation of δ( P − E) with δMCD (Table 3) and the high correlation of δMCD with δMCDD (Table 4), it is clear that changes in P − E are largely attributed to the convergence or divergence changes of the mean circulation (Figs. 4a,c and 5c), consistent with analysis based on ERA-Int (Fig. 5d). Comparing the high- and coarse-resolution changes, the WRF simulation is closer to GLDAS than ERA-Int (Figs. 2a–i). Hence while WRF inherits the correct broad-scale mean circulation changes from ERA-Int through lateral boundary conditions, the added finer-scale circulation change pattern such as the wetter trend in the Qiangtang plateau gives WRF an overall more skillful representation of the P − E changes seen in GLDAS.
Contributions to P − E changes from (a)–(d) mean circulation dynamics and (e)–(h) thermodynamic due to advection [in (a),(b),(e),(f)] and convergence or divergence [in (c),(d),(g),(h) ] from (left) the WRF simulation and (right) ERA-Int in 1998–2011 compared to 1979–97 in WRF simulation (Unit: mm day−1).
Citation: Journal of Climate 28, 10; 10.1175/JCLI-D-14-00581.1
Pattern correlations between annual changes in mean circulation/thermodynamic contribution (δMCD/δTH) and annual mean circulation dynamics/thermodynamic contributions due to the advection of moisture (δMCDA/δTHA), convergence or divergence of moisture (δMCDD/δTHD), and seasonal changes in mean circulation/thermodynamic contributions (δMCDwet/δTHwet and δMCDDRY/δTHDRY) in 1998–2011 compared to 1979–97 in the WRF simulation. Correlation coefficients with an asterisk are statistically significant at 98% confidence level based on a two-tailed t test, and the same applies in other tables.
Divergence changes in the mean circulation could be separated into changes in the zonal and meridional wind components. Figure 6 shows the circulation changes at 200 mb in the high-resolution simulation and the coarse-resolution forcing. Similar to ERA-Int, the high-resolution simulation presents a weaker westerly, which controls the general wetter trend in the TP, as Gao et al. (2014) demonstrated. Differences exist in the meridional component changes. Stronger northwesterly changes over the northwestern TP are presented in the high-resolution simulation (Fig. 6a). The stronger northwesterly changes push the moisture air from the south to the northwestern TP in the high-resolution simulation, which causes the wetter trend in the Qiangtang plateau (Figs. 2b,h) than in the forcing from ERA-Int. In ERA-Int, circulation changes in the northwestern TP are almost subdued (Fig. 6b). In the southern TP, there are even slight increases in northeasterly winds. The almost subdued changes in the northwestern TP and slight increase in northeasterly winds in the southern TP hold the moisture-rich air in the south, which forms the pattern in P and P − E changes in ERA-Int (Figs. 2c,i). As Table 4 demonstrated in Gao et al. (2014), the circulation changes in the upper level are highly correlated to the thermal wind changes, which are triggered by the thermal gradient changes above. The circulation changes in the high-resolution simulation are constrained by the same mechanism related to the thermal changes as in ERA-Int. However, major differences are found and interpreted in the following discussion.
Distributions of changes in circulation (unit: m s−1) at 200 mb in 1998–2011 compared to 1979–97 in (a) the WRF simulation and (b) ERA-Int.
Citation: Journal of Climate 28, 10; 10.1175/JCLI-D-14-00581.1
As mentioned above, a major difference between the WRF simulation and ERA-Int is the thermodynamic contribution. There is almost no contribution of thermodynamic change to P − E changes in ERA-Int (Figs. 4b, 4f), but thermodynamic change contributes to P − E changes in the WRF simulation, especially for the wetter trend in the Qiangtang plateau and Yangtze River headwater and the drying in the southern TP (Figs. 4a,e). When the thermodynamic change is further decomposed into δTHA and δTHD, the pattern correlation between δTHD and δTH is higher than the value between δTHA and δTH (Table 4), suggesting that moisture changes driven by the divergence are more responsible for the thermodynamic changes. Note that δTHD exhibits broad positive changes in the TP except for the upper reaches of the Tarim basin, Qaidam, and lower reaches of the Brahmaputra River basin (Fig. 5g). In contrast, δTHA exhibits mostly negative changes except for the Qiangtang plateau and Qaidam (Fig. 5e).
In recent decades, the largest and second largest inner river basin in the arid northern marginal TP, the Tarim basin and the Heihe River basin, experienced discharge increases (Figs. 2g, 3h,i) revealed by historical records (Wang and Meng 2008; Jiang and Xia 2007). In Figs. 4a,b and 5c,g, the P − E increase over the upper reaches of Tarim basin comes from the divergence change of the mean circulation (Fig. 6a) while the P − E increase over the Qilian Mountain stems from the moisture change indirectly driven by the convergence at low levels. The P − E decrease in the Qaidam basin also comes from the moisture change.
Annual changes in MCD and TH are both more highly correlated with changes in the wet season than the dry season (Table 4; Gao et al. 2014). Therefore, changes in the wet season are further investigated. Changes in divergence at 500 and 100 mb, average vertical motion, moisture, and microphysics diabatic latent heat averaged between 500 and 200 mb are presented in Fig. 7. The patterns in Fig. 7 are all broadly similar to the δMCD (Fig. 4c) and δMCDD (Fig. 5c) patterns. More convergence and divergence are noted in the northwestern and southeastern TP, respectively (Fig. 7a). So moisture is advected to the northwestern TP from its surroundings and enhances moisture from the Qiangtang plateau to Qilian Mountain (Fig. 7d) as shown in Fig. 6a. Larger convergence triggers upward motion (Fig. 7c), resulting in condensation and release of diabatic latent heat by cloud microphysics processes (Fig. 7e) that further heats the air masses over the northwestern TP and enhances divergence at upper levels (Fig. 7b). The patterns of column integrated moisture, divergence, vertical motion, and microphysics diabatic latent heat all match the δMCDD and P − E changes patterns very well.
Distributions of changes in (a),(b) divergence (unit: 10−6 s−1) at 500 and 100 mb, (c) vertical motion (unit: mm s−1), (d) column integrated moisture (unit: g kg−1), and (e) microphysics diabatic latent heat (unit: K s−1) in 1998–2011 compared to 1979–97 in the WRF simulation.
Citation: Journal of Climate 28, 10; 10.1175/JCLI-D-14-00581.1
Comparing Figs. 7a and 7b herein to Figs. 11a and 11b in Gao et al. (2014), we see different divergence changes in the WRF simulation and ERA-Int, especially at low levels. Unlike the evenly increased convergence at low levels in ERA-Int, contrasting changes in divergence between the northwestern and southeastern TP are clear in the WRF simulation. This heterogeneity in divergence is likely related to heterogeneity in topography and overlaid vegetation captured by the finer-scale simulation but not the coarse forcing (Figs. 1, 2h, 2i). This uneven pattern is strengthened through feedback by the latent heat release associated with the convergence and condensation. Herein we identified the precipitation changes and subsequent changes in convergence (divergence) as an important mechanism leading to MCD increases (decreases) in the vast northwestern (southeastern) TP in the regional simulation (Fig. 2h). Furthermore, the larger convergence in the northwestern TP at low levels attracts more moisture to the Qiangtang plateau and Qilian Mountain from the surroundings, which enhances moisture in a swath extending from the Qiangtang plateau to Qilian Mountain, leading to TH changes in the TP (Fig. 4c). This mesoscale contrast in divergence at low levels is captured by the regional simulation but not coarse-resolution products like ERA-Int.
4. Conclusions
The P − E changes from a high-resolution regional climate simulation and the global reanalysis that provided boundary forcing are evaluated by comparing them to GLDAS in the TP. The high-resolution WRF climate simulation not only improves the pattern of P − E changes compared to GLDAS over the best available reanalysis, but it also provides new and substantial findings regarding the contributions of the thermodynamic and the transient eddy components of moisture flux convergence. Most notably, the WRF simulation better represents the observed positive (negative) changes in the vast northwestern (southeastern) TP than the coarse-resolution reanalysis forcing. The improved P − E change pattern is attributed to improved P changes and reduced P biases at high elevations in the high-resolution simulation. Furthermore, the WRF simulation captures the correlation between P − E changes with topography and LAI in the TP similar to that found in GLDAS. This demonstrates that over the TP with complex terrain, high-resolution climate simulations can provide important insights into regional water cycle changes.
Besides the predominant contribution of the mean circulation dynamic changes inherited from the large-scale forcing, thermodynamic and transient eddy changes also contribute importantly to P − E changes in the WRF simulation but not the reanalysis. Contrasting divergence changes at low levels between the northwestern and southeastern TP induced by land surface heterogeneity trigger stronger (weaker) upward motion in the vast northwestern (southeastern) TP, leading to P − E increases (decreases) by MCD changes. The larger convergence in the northwestern TP attracts more moisture to Qiangtang plateau and Qilian Mountain from the surroundings and strengthens P − E changes through diabatic latent heat released and TH change contribution.
Although the WRF simulation provided improved depictions of P − E changes in the TP, the P − E increase in the Yellow River basin and Rouergai plateau is still uncertain because the horizontal resolution of 30 km used in the WRF simulation is still not fine enough to resolve the complexity in land surface properties such as the wetland in the Rouergai plateau and land and atmospheric processes over the TP. Furthermore, many studies reported that the Yellow River basin has experienced large land cover/land use changes in recent decades (Li and Liu 2004; Dong et al. 2005; Cuo et al. 2013), which are not included in our simulation. Increasing model resolution (Jiménez and Dudhia 2012; Ikeda et al. 2010; Rasmussen et al. 2011) and including variable land characteristics/parameters (Gao et al. 2008; Meng et al. 2009) might provide additional realisms to better understand P − E changes in the eastern TP.
Acknowledgments
The reanalysis data used in this study are from the Research Data Archive (RDA), which is maintained by the Computational and Information Systems Laboratory (CISL) at the National Center for Atmospheric Research (NCAR). We acknowledge the Super Computing Center of Chinese Academy of Sciences for providing resources for conducting the simulations. This work is jointly supported by National Basic Research Program of China (2013CB956004), National Natural Science Foundation of China (41322033), and “100-Talent” program granted by the Chinese Academy of Sciences to Yanhong Gao and Lan Cuo. Ruby Leung is supported by the Office of Science of the U.S. Department of Energy through the Regional and Global Climate Modeling program. Pacific Northwest National Laboratory is operated for Department of Energy by Battelle Memorial Institute under Contract DE-AC05-76RL01830.
REFERENCES
Chen, F., and J. Dudhia, 2001: 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.
Collins, W. D., and Coauthors, 2004: Description of the NCAR Community Atmosphere Model (CAM 3.0). NCAR Tech. Note NCAR/TN-464+STR, 226 pp.
Cuo, L., Y. Zhang, Y. Gao, Z. Hao, and L. Cairang, 2013: The impacts of climate change and land cover/use transition on the hydrology in the upper Yellow River basin, China. J. Hydrol., 502, 37–52, doi:10.1016/j.jhydrol.2013.08.003.
Dee, D. P., and S. Uppala, 2009: Variational bias correction of satellite radiance data in the ERA-Interim reanalysis. Quart. J. Roy. Meteor. Soc., 135, 1830–1841, doi:10.1002/qj.493.
Dee, D. P., and Coauthors, 2011: The ERA-Interim reanalysis: Configuration and performance of the data assimilation system. Quart. J. Roy. Meteor. Soc., 137, 553–597, doi:10.1002/qj.828.
Déqué, M., and Coauthors, 2005: Global high resolution versus limited area model climate change projections over Europe: Quantifying confidence level from PRUDENCE results. Climate Dyn., 25, 653–670, doi:10.1007/s00382-005-0052-1.
Dong, L., W. Wang, J. Kong, H. Shi, Z. Guo, and L. Mai, 2005: Supervision of ecoenvironmental change based on RS methods and causal analysis in Maduo County in upriver regions of the Yellow River (in Chinese). Bull. Soil Water Conserv., 25 (4), 68–72.
Duffy, P. B., and Coauthors, 2006: Simulations of present and future climates in the western United States with four nested regional climate models. J. Climate, 19, 873–895, doi:10.1175/JCLI3669.1.
Gao, X., Y. Shi, D. Zhang, J. Wu, F. Giorgi, Z. Ji, and Y. Wang, 2012: Uncertainties in monsoon precipitation projections over China: Results from two high-resolution RCM simulations. Climate Res., 52, 213–226, doi:10.3354/cr01084.
Gao, Y., and Coauthors, 2008: Enhancement of land surface information and its impact on atmospheric modeling in the Heihe River Basin, northwest China. J. Geophys. Res., 113, D20S90, doi:10.1029/2008JD010359.
Gao, Y., Y. Xue, W. Peng, H. Kang, and D. Waliser, 2011: Assessment of dynamic downscaling of China regional summer climate using regional climate model. Adv. Atmos. Sci., 28, 1077–1098, doi:10.1007/s00376-010-0039-7.
Gao, Y., L. R. Leung, E. P. Salathé, Jr., F. Dominguez, B. Nijssen, and D. P. Lettenmaier, 2012: Moisture flux convergence in regional and global climate models: Implications for droughts in the southwestern United States under climate change. Geophys. Res. Lett., 39, L09711, doi:10.1029/2012GL051560.
Gao, Y., L. Cuo, and Y. Zhang, 2014: Changes in moisture flux over the Tibetan Plateau during 1979–2011 and possible mechanisms. J. Climate,27, 1876–1893, doi:10.1175/JCLI-D-13-00321.1.
Gao, Y., J. Xu, and D. Chen, 2015: Evaluation of WRF mesoscale climate simulations over the Tibetan Plateau. J. Climate, 28, 2823–2841, doi:10.1175/JCLI-D-14-00300.1.
Giorgi, F., G. T. Bates, and S. J. Nieman, 1992: Simulation of the arid climate of the southern Great Basin using a regional climate model. Bull. Amer. Meteor. Soc., 73, 1807–1822, doi:10.1175/1520-0477(1992)073<1807:SOTACO>2.0.CO;2.
Grell, G. A., 1993: Prognostic evaluation of assumptions used by cumulus parameterizations. Mon. Wea. Rev.,121, 764–787, doi:10.1175/1520-0493(1993)121<0764:PEOAUB>2.0.CO;2.
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.
Huffman, G. J., R. F. Adler, E. F. Stocker, D. T. Bolvin, and E. J. Nelkin, 2003: Analysis of TRMM 3-hourly multi-satellite precipitation estimates computed in both real and post-real time. Preprints, 12th Conf. on Satellite Meteorology and Oceanography, Long Beach, CA, Amer. Meteor. Soc., P4.11. [Available online at https://ams.confex.com/ams/pdfpapers/54906.pdf.]
Ikeda, K., and Coauthors, 2010: Simulation of seasonal snowfall over Colorado. Atmos. Res., 97, 462–477, doi:10.1016/j.atmosres.2010.04.010.
Jiang, Y., and J. Xia, 2007: The hydrological characteristics of runoff and its response to climate changes in Tarim. Resources Science, 29 (3), 45–52.
Jiménez, P. A., and J. Dudhia, 2012: Improving the representation of resolved and unresolved topographic effects on surface wind in the WRF Model. J. Appl. Meteor. Climatol., 51, 300–316, doi:10.1175/JAMC-D-11-084.1.
Kang, S., Y. Xu, Q. You, W.-A. Flügel, N. Pepin, and T. Yao, 2010: Review of climate and cryospheric change in the Tibetan Plateau. Environ. Res. Lett., 5, 015101, doi:10.1088/1748-9326/5/1/015101.
Kim, J., T. K. Kim, R. W. Arritt, and N. L. Miller, 2002: Impacts of increased atmospheric CO2 on the hydroclimate of the western United States. J. Climate, 15, 1926–1942, doi:10.1175/1520-0442(2002)015<1926:IOIACO>2.0.CO;2.
Krause, P., S. Biskop, J. Helmschrot, W.-A. Flügel, S. Kang, and T. Gao, 2010: Hydrological system analysis and modelling of the Nam Co basin in Tibet. Adv. Geosci., 27, 29–36, doi:10.5194/adgeo-27-29-2010.
Laprise, R., D. Caya, M. Giguere, G. Bergeron, H. Côté, J. P. Blanchet, G. J. Boer, and N. A. McFarlane, 1998: Climate and climate change in western Canada as simulated by the Canadian Regional Climate Model. Atmos.–Ocean, 36, 119–167, doi:10.1080/07055900.1998.9649609.
Leith, C. E., 1973: The standard error of time-average estimates of climatic means. J. Appl. Meteor., 12, 1066–1069, doi:10.1175/1520-0450(1973)012<1066:TSEOTA>2.0.CO;2.
Leung, L. R., Y. Qian, and X. Bian, 2003a: Hydroclimate of the western United States based on observations and regional climate simulation of 1981–2000. Part I: Seasonal statistics. J. Climate, 16, 1892–1911, doi:10.1175/1520-0442(2003)016<1892:HOTWUS>2.0.CO;2.
Leung, L. R., Y. Qian, J. Han, and J. O. Roads, 2003b: Intercomparison of global reanalyses and regional simulations of cold season water budgets in the western United States. J. Hydrometeor., 4, 1067–1087, doi:10.1175/1525-7541(2003)004<1067:IOGRAR>2.0.CO;2.
Li, D., and C. Liu, 2004: Study on land-cover change of the source regions of the Yellow River Basin in the past 10 years. J. Beijing Norm. Univ., 40, 269–275.
Li, R., S. Lu, B. Han, and Y. Gao, 2012: Preliminary comparison and analyses of air temperature at 2m height between three reanalysis data-sets and observation in the east of Qinghai-Xizang Plateau (in Chinese). Plateau Meteor., 31, 1488–1502.
Mearns, L. O., W. Gutowski, R. Jones, R. Leung, S. McGinnis, A. Nunes, and Y. Qian, 2009: A regional climate change program for North America. Eos, Trans. Amer. Geophys. Union, 90, 311, doi:10.1029/2009EO360002.
Meng, X., and Coauthors, 2009: Numerical simulations of the atmospheric and land conditions over the Jinta oasis in northwestern China with satellite-derived land surface parameters. J. Geophys. Res., 114, D06114, doi:10.1029/2008JD010360.
Moore, G. W. K., 2012: Surface pressure record of Tibetan Plateau warming since the 1870s. Quart. J. Roy. Meteor. Soc., 138, 1999–2008, doi:10.1002/qj.1948.
Nikulin, G., and Coauthors, 2012: Precipitation climatology in an ensemble of CORDEX-Africa regional climate simulations. J. Climate, 25, 6057–6078, doi:10.1175/JCLI-D-11-00375.1.
Plummer, D., D. Caya, A. Frigon, H. Côté, M. Giguère, D. Paquin, S. Biner, R. Harvey, and R. de Elia, 2006: Climate and climate change over North America as simulated by the Canadian RCM. J. Climate, 19, 3112–3132, doi:10.1175/JCLI3769.1.
Rasmussen, R., and Coauthors, 2011: High-resolution coupled climate runoff simulations of seasonal snowfall over Colorado: A process study of current and warmer climate. J. Climate, 24, 3015–3048, doi:10.1175/2010JCLI3985.1.
Rodell, M., and Coauthors, 2004: The Global Land Data Assimilation System. Bull. Amer. Meteor. Soc., 85, 381–394, doi:10.1175/BAMS-85-3-381.
Seager, R., N. Naik, and G. A. Vecchi, 2010: Thermodynamic and dynamic mechanisms for large-scale changes in the hydrological cycle in response to global warming. J. Climate, 23, 4651–4668, doi:10.1175/2010JCLI3655.1.
Simmons, A., S. Uppala, D. Dee, and S. Kobayashi, 2006: ERA-Interim: New ECMWF reanalysis products from 1989 onwards. ECMWF Newsletter, No. 110, ECMWF, Reading, United Kingdom, 25–35.
Skamarock, W. C., J. B. Klemp, J. Dudhia, D. O. Gill, D. M. Barker, W. Wang, and J. G. Powers, 2005: A description of the Advanced Research WRF, version 2. NCAR Tech. Note 468+STR, 88 pp. [Available online at http://wrf-model.org/wrfadmin/docs/arw_v2.pdf.]
Solomon, S., D. Qin, M. Manning, Z. Chen, M. Marquis, K. Averyt, M. M. B. Tignor, and H. L. Miller Jr., 2007: Climate Change 2007: The Physical Science Basis. Cambridge University Press, 996 pp.
Turk, F. J., G. Rohaly, J. D. Hawkins, E. A. Smith, A. Grose, F. S. Marzano, A. Mugnai, and V. Levizzani, 2000: Analysis and assimilation of rainfall from blended SSM/I, TRMM, and geostationary satellite data. Preprints, 10th Conf. on Satellite Meteorology and Oceanography, Long Beach, CA, Amer. Meteor. Soc., 66–69.
Wang, A., and X. Zeng, 2012: Evaluation of multi-reanalysis products with in situ observations over the Tibetan Plateau. J. Geophys. Res., 117, D05102, doi:10.1029/2011JD016553.
Wang, J., and J. Meng, 2008: Characteristics and tendencies of annual runoff variations in the Heihe River Basin during the past 60 years. Sci. Geogr. Sin., 28, 83–88.
Wu, G., A. Duan, X. Zhang, Y. Liu, Y. Ma, and K. Yang, 2013: Extreme weather and climate changes and its environmental effects over the Tibetan Plateau. Chin. J. Nat., 35 (3), 167–171.
Wu, S., Y. Yin, D. Zheng, and Q. Yang, 2007: Climatic trends over the Tibetan Plateau during 1971–2000. J. Geogr. Sci.,17, 141–151, doi:10.1007/s11442-007-0141-7.
Xu, J., and Y. Gao, 2014: Validation of summer surface air temperature and precipitation simulation over Heihe River Basin. Plateau Meteor., 33, 937–946, doi:10.7522/j.issn.1000-0534.2013.00149.
Xu, Z. X., T. L. Gong, and J. Y. Li, 2008: Decadal trend of climate in the Tibetan Plateau—Regional temperature and precipitation. Hydrol. Processes, 22, 3056–3065, doi:10.1002/hyp.6892.
Yang, K., B. Ye, D. Zhou, B. Wu, T. Foken, J. Qin, and Z. Zhou, 2011: Response of hydrological cycle to recent climate changes in the Tibetan Plateau. Climatic Change, 109, 517–534, doi:10.1007/s10584-011-0099-4.
Yin, Y., S. Wu, D. Zhao, D. Zheng, and T. Pan, 2012: Impact of climate change on actual evapotranspiration on the Tibetan Plateau during 1981–2010 (in Chinese). Acta Geogr. Sin., 67, 1471–1481.
You, Q., S. Kang, E. Aguilar, and Y. Yan, 2008: Changes in daily climate extremes in the eastern and central Tibetan Plateau during 1961–2005. J. Geophys. Res., 113, D07101, doi:10.1029/2007JD009389.
Zhang, Y., V. Dulière, P. Mote, and E. P. Salathé, 2009: Evaluation of WRF and HadRM mesoscale climate simulations over the United States Pacific Northwest. J. Climate, 22, 5511–5526, doi:10.1175/2009JCLI2875.1.
