利用ECMWF再分析数据验证PRECIS对中国区域气候的模拟能力

Validating PRECIS with ECMWF Reanalysis Data over China

本文利用1979-1983年的ECMWF再分析数据作为准观测侧边界条件驱动:Hadley气候预测和研究中心的区域气候模式系统PRECIS,验证PRECIS对中国区域的气候模拟能力。选择典型的观测站点北京模拟的日最高/最低气温、太阳短波总辐射和月均降水量与观测结果进行直接比较,显示:PRECIS具有很强的模拟地面气候季节变化的能力。全国740个台站的观测与模拟值的统计分析表明:尽管模拟的最高,最低气温在0℃附近有一‘锯齿’状的偏差,PRECIS能够很好地模拟全国范围最高/最低气温的型态分布特征;从全国范围看,模拟的降水值偏高,但显示出很强的模拟极端降水事件的能力;模拟的地面太阳短波总辐射与全国122个台站的观测结果的比较显示:PRECIS模拟的辐射值偏高,全国范围内约高22％。因此,当应用PRECIS输出结果进行气候变化的影响评价时,需要对模式的输出结果进行订正。

In this paper, 5-year (1979 -1983) ECMWF reanalysis data is employed as quasi-observational boundaries to drive PRECIS to validate PRECIS capacity of simulating present climate over China. It is indicated from the direct day-today comparison of the simulated and observed daily maximum/minimum surface air temperature (Tmax/Tmin) and total downward shortwave surface flux in Beijing station that PRECIS possesses strong capacity to simulate the seasonal cycles of these variables; direct comparison of monthly mean precipitation shows that the simulated seasonal cycle of precipitation fits the observation very well. The statistical analyses for whole country show that PRECIS can simulate Tmax/Tmin very well, though there is an odd zig-zag in the occurrence curve around 0 ℃, the simulated precipitation is normally larger, however, the statistical comparison of simulated and observed daily precipitation indicates that the capacity of PRECIS to simulate rainfall is strong; statistical analysis shows PRECIS overestimates the SW radiation about 22% over the whole country. It is necessary to re-calibrate the outputs of PRECIS when they are used as input data for potential impacts assessment work.