CLIGEN天气发生器在黄河流域的适应性研究

Adaptability of Climate Generator of CLIGEN in Yellow River Basin

根据年降水及其年内分布特征在全美范围内选择参照站的基础上,用黄河流域3个气象站30年降水和气温的月平均资料,对CLIGEN天气发生器在黄河流域的适应性进行了检验。结果表明利用参证站、通过月资料的输入,CLIGEN可以较好地模拟年降水、降水的月分布、最高温度和最低温度,模拟值的标准差普遍低于实际值的标准差,参照站的气候特征对模拟结果有显著影响,因此,在参证站选择时应综合考虑多个气象参数。

CLIGEN is one stochastic weather generator to produce the daily future weather scenarios for hydrological model and soil erosion model. It had been well tested in many locations across the United States. The long period daily input of history climate files limited it to be widely used in other countries. This study was conducted to calibrate CLIGEN with 30-year monthly precipitation and temperature data of three climate sites in Yellow River basin of China. The reference climate sites were selected based on the annual and monthly distribution of precipitation. The generated data by CLIGEN was compared to observed values. The results indicated that the CLIGEN was successful in modeling the annual and monthly precipitation. The average relative error of annual precipitation for three sites was 2.57%, and the generated and observed monthly precipitation was closely correlated (R^2≥(0.98)). The generated maximum temperature was also well fitting to observed data with a slight average greater of 0.06℃. The generated minimum temperature of Xining and Yinchuan was 0.06℃ lower than observed values. Due to the influence of reference site, the generated minimum temperature of Xi'an was significant difference from observed data. There were some differences in most of standard deviations between generated and observed values. The results also indicated that CLIGEN was capable of generating stochastic weather data based on monthly input files. However, more parameters should be used for reference climate site selection.