| 本文已被：浏览 1809次 下载 1108次 |
码上扫一扫！ |
| 福建旱涝短期气候预测统计方法试验研究 |
|
张容焱1, 李永尧2, 徐宗焕3, 吴滨1, 高建芸1
|
|
1.福建省气候中心, 福建, 福州, 350001;2.福建师范大学数学与计算机学院, 福建, 福州, 350007;3.福建省气象科学研究所, 福建, 福州, 350001
|
|
| 摘要: |
| 利用最优子集、神经网络、EOF-CCA和均生函数4种常用统计方法,对福建春季、前汛期、后汛期和秋冬季进行年度、季度旱涝的短期气候预测试验,采用统一的评分标准评定。结果表明:最优子集、神经网络模型具有较好的预测能力,EOF-CCA模型有一定的预测能力,均生函数模型相对较差。 |
| 关键词: 旱涝 短期气候预测 统计模型 预测能力 |
| DOI： |
| 分类号: |
| 基金项目:中国气象局新技术推广项目(CMATG2007M36)“东南沿海农业气象灾害预警及减灾模式研究”资助 |
|
| EXPERIMENTAL RESEARCH WITH STATISTICS METHOD INTO SHORT-TERM CLIMATE PREDICTION ABOUT FLOOD/DROUGHT IN FUJIAN PROVINCE |
|
ZHANG Rong-yan1, LI Yong-yao2, XU Zong-huan3, WU Bing1, GAO Jian-yun1
|
|
1.The Climate Center of Fujian, Fuzhou 350001, China;2.Math and Computer College, Fujian Normal University, Fuzhou 350007, China;3.Meteorological Institute of Fujian, Fuzhou 350001, China
|
| Abstract: |
| Experimental research has been made into short-term climate prediction of flood/drought in spring,monsoon and typhoon seasons and autumn and winter seasons.It lasted a year and a quarter in Fujian with four usual statistics methods,namely,the optimal subset,neural network,EOF-CCA and mean generating function.The result is graded by a unified scoring criterion,which indicates that the optimal subset and neural network have better forecast ability than the model of EOF-CCA and the mean generating function has the worst forecast ability. |
| Key words: flood/drought short-term climate prediction statistical model forecast ability |