【摘要】 基于1960—2012年的地面常规气象指标、大尺度气候指数和NOAA气候分析产品,利用功率谱周期分析、时间尺度分离分析和交叉检验的逐步回归分析等,把吉林省春旱期(4—5月)降水量分离成不同时间尺度的值,并在年代际尺度和年际尺度下分别找到显著相关的影响因子。结果表明:吉林春旱期降水存在着2~4 a的年际变化和10 a左右的年代际变化。在年代际尺度上,4月降水与前期3月南半球环球状态指数以及俄罗斯东部高纬地区的低空经向风有关,5月降水与同期北半球环球状态指数和前期4月太平洋中高纬环流有关;在年际尺度上,4月降水与前期3月混合ENSO指数和同期当地相对湿度、华东华北沿海地区的低层经向风有关,5月降水与同期北大西洋涛动指数以及局地相对湿度、地面气压有关。利用选出的影响因子对降水进行预报,估计值和真实值的相关系数分别为0.67(4月)和0.81(5月),且选择合适的影响因子比模型结构更加重要。更多还原

【Abstract】 Based on the ground meteorological indicators,the climate indexes and the NOAA reanalysis data during 1960—2012,spring-drought-period( April—May) monthly precipitation amounts in Jilin Province are analyzed by using power spectrum analysis,time-scale decomposition,cross-examination stepwise regression analysis,etc. This paper aims to find proper potential annual influence factors and decadal ones. The results show that April—May precipitation in Jilin Province has two significant periods,namely,2—4 years and about 10 years. In terms of the decadal scale,precipitation in April is closely related to the Southern Hemisphere Annular Mode Index in March and the 850 h Pa meridional wind in March over the high-latitude area in East Russia,while that in May is related to the Northern Hemisphere Annular Mode Index in May and the Pacific mid-high-latitude circulations in April. As for the interannual scale,precipitation in April has close relationship with the Multivariate ENSO Index in March,the local relative humidity in April and the 850 h Pa meridional wind in April over coastal areas in East and North China,whilethat in May does with the contemporary North Atlantic Oscillation Index,the local relative humidity and the surface pressure. The coefficients of correlation between the real values and the fitted ones are 0. 67(April) and 0. 81(May) separately,and proper potential influence factors play more important roles than the model structure.更多还原