湖南地区加权平均温度的影响因素分析及建模

摘要

In this study,3 consecutive years of radiosonde observation data since 2012 from three sites of Changsha,Huaihua and Chenzhou in Hunan province has been used to establish multi-factor Tm local models based on the numerical integration and least squares principle.The results showed that Tm has a positive correlation with ground temperature (Ts) and water vapor pressure (es),while it has a negative correlation with atmospheric pressure (Ps).Compared with the bias of single-factor local Tm model (-2～3 K),the bias of Bevis model is mainly in 0～4 K,which means that Bevis model exists a system deviation in Hunan region.The research also reveals that the two-factor model and three-factor model have an approximate accuracy,but they are all better than single-factor model.Summarily,the two-factor local Tm model is better than single-factor model which is better the Bevis model at the same time.Therefore,it can be applied to researches and applications of GPS meteorology in Hunan region.%利用湖南地区长沙、怀化、郴州3个探空站连续3 a(2012～2014年)的观测数据,以数值积分法计算的大气加权平均温度(Tm)作为参考值,分析Tm与地面气温Ts、水汽压es及大气压Ps之间的相关性,并基于最小二乘原理建立湖南地区多因子Tm模型.研究结果表明,Tm与地面温度Ts和水汽压es呈正相关性,与大气压Ps呈负相关;Bevis模型的Tm偏差主要分布在0～4 K之间,存在一定的系统偏差,而基于Ts单因子的Tm模型偏差主要分布在-2～3 K之间;双因子与三因子精度相当,但均优于单因子模型.总体而言,基于地面气温Ts和水汽压es的双因子Tm模型的精度优于基于Ts的单因子模型,更优于Bevis模型,适用于湖南地区的GPS气象研究及业务应用.