CLIGEN气候生成器模型在紫色土地区的适应性研究

Study on Adaptability of CLIGEN Model in Purple Soil Area

随机气候生成器CLIGEN模型可以为当前一些综合模型所需要的气象资料提供必要的数据支持,为检验CLIGEN在四川紫色土地区的适用性,以CLIGEN模型为研究平台,根据四川省遂宁水土保持实验站1987～2002年降水和气温的月平均资料及在全美范围内选择参照站点得到的其他气象因子,利用CLIGEN模型分别模拟出遂宁市的月平均降水量、月平均最高温度和月平均最低温度。结果证明用所选的3个参证站点,利用CLIEGN模拟出来的降水、气温在数值上和年内的分布上均与实测值较为符合,选择模型有效系数E和平均绝对百分比误差（MAPE）为评价指标,作为3个站点最优选择的指标,最终选择CENTERVILLE站点为最优参证站,CEN-TERVILLE站点的模拟月平均降雨量、月平均最高温度和月平均最低温度的相对平均绝对值（MAPE）分别为7.15%,7.03%和6.17%,而有效系数分别为0.996 2,0.926 6和0.995 7,从而确定了CLIGEN模型在遂宁紫色土地区的适用性。

Daily weather data are required as climate input to many eompositive, CLIGEN model is a stochastic weather generator to meet this need. The paper focused on the adaptability of climate generator in Suining purple soil based on the CLIGEN model as a research tool,the paper utilized CLIGEN model to simulate respectively the average precipitation monthly, the average highest temperature monthly and the average lowest temperature monthly by using Suining average meteorological data of 15 years and referencing three meteorological stations in America（CENTERVILLE, MADISONVILLE and CROCKETT）. Results show, using the CLIGEN model to simulate, the value and distribution of precipitation and temperature in a year are fitted with actual value well. We select the available coefficient E of model and meaning absoluted percentage error（MAPE） as the index sign of most excellent selection station,finally select CENTERVILLE station as the most excellent station,the meaning absoluted percentage error（MAPE）,which use the CLIGEN model to simulate the average precipitation monthly, the average highest temperature monthly and the average lowest temperature monthly are only 7.15%, 7.03% and 6. 17% respectively;and the available coefficient are 0. 9962,0. 9266 and 0. 9957 respectively. Thus definite the applicability of CLIGEN model in Suining purple soil finally .