摘要
应用作物产量历史丰歉气象影响指数、作物气候适宜指数、作物生长模拟模型分别建立早稻产量动态预报方法,利用这3种方法分别对1996―2005年湖南早稻产量进行动态模拟预报,在分析预报误差的基础上,确定每种方法的预报权重,建立动态集成预报方法。拟合检验结果表明,集成预报方法的丰歉趋势预报正确率、产量预报准确率都较任意单一预报方法稳定,且有一定程度提高。利用上述方法分别对2006―2008年湖南早稻产量进行预报检验,结果表明,除2006年的丰歉趋势预报因早稻收获阶段遭遇台风出现错误外,集成预报法的丰歉趋势预报正确率、产量预报准确率都较高,且好于任意单一预报方法,能够满足业务服务的要求。
The theories of historical meteorological influence index for bumper or poor harvest of crop yield,crop climatic suitability index and crop growth simulation model were respectively applied to establish the methods for dynamic prediction of early rice yield.These three methods were used to predict early rice yield in Hunan,China from 1996 to 2005.Based on the analysis of prediction errors,a dynamic integrated prediction method was developed by using the weights of each method.The results of fit test showed that the prediction accuracy of integrated method for the bumper or poor harvest trend and the actual yield was more stable and higher than any single prediction method.Tests for yield of early rice in Hunan,China from 2006 to 2008 indicated that the prediction accuracy of the dynamic integrated prediction method was higher than any single prediction method except for the bumper or poor harvest trend in 2006 because of a typhoon hazard encountered at the harvest stage,and it can satisfy the needs of operational services.
出处
《中国水稻科学》
CAS
CSCD
北大核心
2011年第3期307-313,共7页
Chinese Journal of Rice Science
基金
中国气象局基础建设项目
关键词
气象影响指数
气候适宜指数
ORYZA2000模型
早稻
产量
动态预报
丰歉趋势
meteorological influence index
climatic suitability index
ORYZA2000 model
early rice
yield
dynamic prediction
bumper or poor harvest trend
