中文摘要：

就江淮梅雨锋低涡预报基于奇异矢量目标观测作了观测系统模拟试验,目的在于对基于奇异矢量目标观测实际实施作预先研究,寻找目标观测中所要遵循原则和实施细节,以及用奇异矢量确定目标观测区的恰当方法。经分析实际奇异矢量相关误差如何影响预报特征,得出在实施目标观测时应遵循的原则：只对奇异矢量相关误差进行订正,不对非奇异矢量相关误差订正 对奇异矢量强相关误差区域优先订正能更为高效率地改进预报 对整个垂直气柱进行订正,而不只对满足阈值区域进行订正 应优先采用效率较高的第一类斜压订正方案。文中两种方法确定的目标观测区与实际奇异矢量相关误差区域在位置、大小、形状上比较相似,两种方法的目标观测区误差影响预报方式与真实奇异矢量相关误差影响预报方式很相近。

英文摘要：

With the aim to pre-analyze the implementation of singular-vectors-based adaptive observation, disclose its principles, fix some problems worthy of special attention, and find proper methods to locate the adaptive region, an observation system simulation experiment（OSSE） is presented in this paper for singular-vectors-based adaptive observation of low pressure on the Mei-yu front. Four principles are gotten in this paper. Firstly, only the errors related with singular vectors are corrected while all others are ignored that are not about the singular vectors. Secondly, when errors related with relatively strong singular vectors are given the priority in correction, forecasts can be improved more efficiently. Thirdly, the errors in the entire vertical atmospheric column should be taken into account, instead of correcting those fitting the criterion only. Finally, the fn＇st kind of baroclinic correction scheme is better than all the others in efficiency. Two methods are used to decide on the adaptive region. The determined adaptive region is similar to the singular vectors related error regions in terms of position, area, and shape. Meanwhile, the error in these adaptive regions influences the forecast in the same manner as the true singular vectors related error does.