CAJ | 학술논문

利用奇异值分解(SVD)方法、500 hPa高度场、太平洋海温场和降水资料,建立起汛期降水的预测方程;经过适应本地化的Z指数修正,将预测结果转化为旱涝等级;将SVD技术与修正的Z指数结合起来,实现旱涝的气候预测;将研究成果推广应用到气象、防汛抗旱部门.结果表明:1)影响江淮分水岭地区汛期降水的因子有5个,分别是太平洋地区2个,印度半岛附近2个,欧洲地区1个;2)理论上的Z指数等级不符合江淮分水岭地区的实际状况,因而必须对Z指数进行修正.经过修正后的各个旱涝等级的划分概率较为合理,说明Z指数的5级指标是可靠的;3)利用5个影响因子可以建立汛期降水量与影响因子之间的预报方程,在共计8年的旱涝滚动预测和实况检验中,等级相符的有7年,只有2003年的预测试验相差一个等级,5级的预测准确率达到87.5%;4)经过气象、防汛抗旱部门2008年的应用,旱涝等级的预测意见和实际基本吻合,说明预测技术的应用情况良好.
이용기이치분해(SVD)방법、500 hPa고도장、태평양해온장화강수자료,건립기신기강수적예측방정;경과괄응본지화적Z지수수정,장예측결과전화위한로등급;장SVD기술여수정적Z지수결합기래,실현한로적기후예측;장연구성과추엄응용도기상、방신항한부문.결과표명:1)영향강회분수령지구신기강수적인자유5개,분별시태평양지구2개,인도반도부근2개,구주지구1개;2)이론상적Z지수등급불부합강회분수령지구적실제상황,인이필수대Z지수진행수정.경과수정후적각개한로등급적화분개솔교위합리,설명Z지수적5급지표시가고적;3)이용5개영향인자가이건립신기강수량여영향인자지간적예보방정,재공계8년적한로곤동예측화실황검험중,등급상부적유7년,지유2003년적예측시험상차일개등급,5급적예측준학솔체도87.5%;4)경과기상、방신항한부문2008년적응용,한로등급적예측의견화실제기본문합,설명예측기술적응용정황량호.
A forecast equation for drought/flood during the flood season is established, by using the Singular Value Decomposition (SVD) and the data of 500-hPa height field, Sea Surface Temperature(SST) in the Pacific Ocean, and precipitation. Prediction results are transformed into drought/flood grade after the modified Z-index is used. SVD method combined with modified Z-index is used to predict the drought and flood trend. The results are used in the meteorological department, the flood prevention and drought resisting department. The results are as follows:1) There are five factors affecting the precipitation during the flood season in the Changjiang-Huaihe watershed. Two factors are in the Pacific region, two near the Indian Peninsula, and one around Europe. 2) The theoretical Z-index grade is not suitable for the Changjiang-Huaihe watershed, so Z-index has to be modified. Modified Z-index is more reasonable, which shows that the five grades of Z-index are reliable. 3) By using the above five factors, the prediction equation between the precipitation during the flood season and the effective factors is established. Among the drought/flood rolling prediction and practical test, the match grades can be seen in eight years,except 2003. The accuracy of the five grade reaches 87.5%. 4) According to the application in the meteorological department, the flood prevention and drought resisting department, the prediction opinions are in agreement with practical situation, which proves the application of the technology is in good condition.