Impact analysis of processing strategies for long-term GPS zenith tropospheric delay (ZTD)

Abstract

Homogenized atmospheric water vapour data are an important prerequisite for climate analysis. Compared with other techniques, GPS has an inherent homogeneity advantage, but it still requires reprocessing and homogenization to eliminate impacts of applied strategy and observation environmental changes where a selection of proper processing strategies is critical. This paper comprehensively investigates the influence of the mapping function, the elevation cut-off angle and homogenization on long-term reprocessing results, in particular for zenith tropospheric delay (ZTD) products, using GPS observations at 44 IGS (International GNSS Service) stations during 1995 to 2014. In the analysis, for the first time, we included the mapping function (the latest Vienna mapping function, VMF3) and exploited homogenized radiosonde data as a reference for ZTD trend evaluations. Our analysis shows that both site position and ZTD solutions achieved the best accuracy when using VMF3 and the 3∘ elevation cut-off angle. Regarding the long-term ZTD trends, homogenization reduced the trend inconsistency among different elevation cut-off angles. ZTD trend results show that the impact of mapping functions is very small, with a maximum difference of 0.19 mm yr−1. On the other hand, the discrepancy can reach 0.60 mm yr−1 using different elevation cut-off angles. Low-elevation cut-off angles (3∘ or 7∘) are suggested for the best estimates of ZTD reprocessing time series when compared to homogenized radiosonde data or ERA5 reference time series.