中国天气发生器非降水变量模拟参数分布特征

Spatial variability of the parameters of the Chinese stochastic weather generator for daily non-precipita-tion variables simulation in China

对基于马尔可夫链的理查森型中国天气发生器降水模拟已经有过比较系统的研究,但对非降水变量的模拟及其参数的分布特征等的研究还有待进一步深入。文中根据中国669个站点1971—2000年的逐日降水、最高气温、最低气温和日照时数资料,分干、湿两种状态计算了中国天气发生器各非降水变量的模拟参数——干、湿日条件下平均值和标准差的傅立叶系数以及各变量残差序列之间当天和后延一天的自相关、互相关系数,并分析了这些模拟参数在中国的空间分布规律,为中国天气发生器的进一步推广应用以及模拟参数的空间插值提供了技术支撑。

has been Simulation of daily precipitation in China with the ＂Richardson-type＂ stochastic weather generator BCC/RCG-WG introduced in previous works. But more works need to be done for non-precipitation variables simulation. In this pa- per, a set of parameters for daily non-precipitation variables including maximum temperature, minimum temperature and sun- shine hours is determined from observations during 1971 - 2000 over China within the framework of the BCC/RCG-WG. The geographic distribution of the Furrier coefficients of mean and standard deviation of the non-precipitation variables for rainy and dry states respectively and autocorrelation coefficients and cross-correlation coefficients of the non-precipitation variables on the same day and with one day lag for the 669 stations all over the China are described and analyzed.