GTDM:一种获取全球对流层延迟的新模型

GTDM:a new model for generating global tropospheric delay

对流层延迟是GNSS定位的主要误差源。现有的各对流层延迟模型大多存在过度拟合的弊端,不能反映延迟在短时间内的细节变化。本文利用2011—2017年ECMWF气象资料分析了对流层延迟的变化特征,发现同一格网相邻年份之间全球对流层延迟偏差绝大多数在5 cm内。在此基础上,本文提出了一种非参数拟合的对流层延迟模型GTDM。经验证,GTDM模型具有较好的拟合效果。本文将2016—2017年IGS分析中心提供的对流层延迟产品数据、探空气象数据解算的对流层延迟作为外检核数据,验证结果表明,GTDM模型在全球范围的精度均优于GZTD、GPT2w、UNB3m模型。GTDM模型建模方法简单,可避免过度拟合对流层延迟值的问题,能够有效地反映对流层延迟变化特征。

Tropospheric delay is the main error source in GNSS positioning.Various existing tropospheric delay models have the disadvantage of overfitting,the details of delay changes in a short time.In this paper,the variation characteristics of tropospheric delay are analyzed by using the meteorological data of ECMWF from 2011 to 2017.It is found that the global tropospheric delay deviation between adjacent years of the same grid is mostly within 5 cm.Accordingly,a new tropospheric delay model GTDM is proposed.It has been verified that GTDM has a good fitting effect.In this paper,the data from 2016—2017 are used as the external compliance inspection data including the tropospheric delay product data of IGS stations provided by the IGS analysis center and the tropospheric delay calculated by the reanalysis of meteorological data by the sounding data.The results show that the accuracy of GTDM is higher than that of GZTD,GPT2w,UNB3m in global.A simple modeling method,GTDM can avoid overfitting the value of tropospheric delay and reflect the characteristics of tropospheric delay changes effectively.