摘要
基于 5 33年旱涝指数场序列 ,采用了滑动累积频率法用于降水序列重建工作 ,并选用了 3种不同均值用于序列重建试验 ,得到了 1 4 99～ 2 0 0 2年共 5 0 4年连续的滑动平均降水场序列。分析表明 ,重建的 5 0 4年降水场序列与 4 9年实测降水场序列相关性较好 ,其平均相对均方误差在 1 89%以下 ;重建的序列与原旱涝指数序列的非线性关系 ,表明重建序列是合理的 。
Based on the 533 year dryness and wetness grades series, the Moving Cumulative Frequency Method (MCFM) was developed, and a moving average precipitation field series from 1499 to 2002 was reconstructed. The results show that there is good relationship between the reconstructed series and the primitive 49 year observation precipitation series, their relative root mean square errors are below 1 89%; and the reconstructed series has a nonlinear relation to the dryness and wetness grades series. It means that the reconstructed series is reliable. The reconstruction of the moving average precipitation field series laid a foundation for researching the regularities of drought and flood evolution.
出处
《气象科技》
北大核心
2004年第3期163-167,共5页
Meteorological Science and Technology
基金
国家科技部社会公益类项目 2 0 0 2DIB2 0 0 67课题资助
