顾及多因子的中国区域加权平均温度模型  

China ’s Regional Weighted Average TemperatureModel Considering Multiple Factors

作　　者：占果 王盼 黄嘉艇 唐旭 ZHAN Guo;WANG Pan;HUANG Jiating;TANG Xu(School of Remote Sensing and Geomatics Engineering,Nanjing University of Information Science and Technology,Nanjing 210044,China;China Design Group Co.Ltd.,Nanjing 210014,China)

机构地区：[1]南京信息工程大学遥感与测绘工程学院,南京210044 [2]华设设计集团股份有限公司,南京210014

出　　处：《遥感信息》2024年第1期112-120,共9页Remote Sensing Information

基　　金：华设设计集团股份有限公司科技开放基金(KY2021074);华设设计集团股份有限公司D类科研基金(KY2021043)。

摘　　要：针对中国地区地形起伏较大等情况影响T m值特别是短期T m值的高精度获取的问题,提出一种顾及多因子的大气加权平均温度模型。首先,通过探空数据利用积分法计算出加权平均温度,拟合分析选出特征因子加入建模;其次,将利用积分法计算的探空数据与NCEP数据分别作为参考值,引入Bevis模型、GPT3-Z模型和时空模型,将模型计算的预测值和参考值作对比,分析各模型精度;之后,将各模型用于水汽反演,分析反演精度。结果表明:多因子T m模型反演的T m值精度最高,且多因子T m模型中加入了日周期变化,相比全年在短期时间内提升更加明显;将各模型用于GNSS水汽计算,多因子T m模型精度最好,理论RMS误差和相对误差分别为0.23 mm和1.69%。Aiming at the problem that the terrain fluctuation in China affects the high-precision acquisition of T m value,especially the short-term T m value,this paper proposes an atmospheric weighted average temperature model considering multiple factors.Firstly,the weighted average temperature is calculated by the sounding data using the integral method,and the characteristic factors are selected by fitting analysis to join the modeling.Secondly,the sounding data calculated by the integral method and the NCEP data are used as reference values respectively,and the Bevis model,GPT3-Z model and space-time model are introduced.The predicted values and reference values calculated by the model are compared to analyze the accuracy of each model.After that,each model is used for water vapor inversion to analyze the inversion accuracy.The results show that the T m value retrieved by the multi-factor T m model has the highest accuracy,and the daily cycle change is added to the multi-factor T m model,which is more obvious than the short-term improvement in the whole year.When each model is used for GNSS water vapor calculation,the multi-factor T m model has the best accuracy,and the theoretical RMS error and relative error are 0.23 mm and 1.69%,respectively.

关 键 词：加权平均温度 精化模型 探空数据 NCEP再分析资料 GNSS水汽探测 

分 类 号：P228[天文地球—大地测量学与测量工程]