《Table 1.Comparison of the simulated and observed station-averaged near-surface air temperature in J

  提示：宽带有限、当前游客访问压缩模式

本系列图表出处文件名：随高清版一同展现

《"Modeling the Warming Impact of Urban Land Expansion on Hot Weather Using the Weather Research and Forecasting Model: A Case Study of Beijing, China"》

1. 获取 高清版本忘记账户？点击这里登录

1. 下载图表忘记账户？点击这里登录

To evaluate the model，we compared the simulated results with direct observations from the corresponding period.According to the locations of the 17 meteorological stations，we calculated the station-averaged simulated results in JJA（control run）.The statistics showed that the model tended to overestimate the mean temperature，with a systematic bias of1.14?C（Table 1），which was at least as good as what had been shown in prior studies[1.55?C in Wang et al.（2013b）；1.55?C in Wang et al.（2015a）；1.0?C–1.5?C in Cao et al.（2016）].Previous studies attributed the positive systematic discrepancies to the inaccuracies of the boundary conditions and sensitivities of the physics parameterization schemes（Cao et al. （2016）；Wang et al.（2015a）） .In this study，the warm bias of the temperature would have been primarily caused by the simulation without consideration of atmospheric aerosol.This is because atmospheric aerosol can reduce the fraction of solar shortwave radiation arriving at the surface，reduce the ground-stored heat，and accordingly decrease the air temperature.The station-based spatial correlation of mean temperature between the observations and simulated data was 0.94（P<0.01）.The model also overestimated the observed maximum air temperature，with an absolute error of～1?C（Table1）.Its station-based spatial correlation（0.71）was smaller than that of the mean and minimum temperature.The simulated minimum air temperature agreed well with the observed data，with an absolute error of～0.2?C（Table 1）.Figure 3indicates that the temporal evolution of the simulated daily temperature agreed with observations.