Reducing ZHD–ZWD mutual absorption errors for blind ZTD model users

Abstract

In precise point positioning (PPP), the zenith hydrostatic delay (ZHD) is traditionally treated as a priori value from the ZHD model but the zenith wet delay (ZWD) is treated as an unknown parameter to estimated. For some near real-time PPP tasks, priori ZHD from the blind ZHD models are often used because external meteorological measurement or information about atmospheric pressure or ZHD can not be obtained in such conditions. On the other hand, a priori ZHD errors can project into GNSS height estimates errors in the PPP analysis, because unmodeled part of the ZHD is usually absorbed into the estimated ZWD while there is the difference between the hydrostatic mapping function and the wet mapping function. In this study, we found the errors of the ZHD seasonal models are not always less than those of the ZWD seasonal models, which implies that the traditional strategy for treating ZTD (priori ZHD and estimated ZWD) is not always the best choice when using the blind zenith tropospheric delay (ZTD) models. A decision model for the ZTD blind model (DMZBM), which is a new strategy of dealing with ZTD in PPP when using blind ZTD models, was proposed. For most cases, the traditional strategy works while for some exception cases, it is recommended that priori ZWD from the blind ZWD model was set but ZHD is treated as an unknown parameter to estimated. The DMZBM can directly reduce ZHD–ZWD mutual absorption errors which potentially reduce GNSS height estimates errors due to the difference between the hydrostatic mapping function and the wet mapping function. The results show that there are significant reductions of ZHD–ZWD mutual absorption errors in polar regions when using the new ZTD-treating strategy.