On the Study of Influences of Different Factors on the Rapid Tropospheric Tomography

Abstract

A rapid tropospheric tomography system was developed by using algebraicreconstruction technique. Influences of different factors on the tomographic results, including theground meteorological data, the multi-Global Navigation Satellite System (GNSS) observations, theground station distribution and the tomographic horizontal resolution, were systematicallyinvestigated. In order to exclude the impacts from discrepancies of water vapor informationbetween input observations and references on the tomographic results, the latest reanalysisproducts, ERA5, which were taken as references for result evaluations, were used to simulate slantwet delay (SWD) observations at GNSS stations. Besides, the slant delays derived from GNSSprocessing were also used to evaluate the reliability of simulated observations. Tomography resultsshow that the input both SWD and ground meteorological data could improve the tomographicresults where SWD mainly improve the results at middle layers (500 to 5000m, namely 2 to 16 layer)and ground meteorological data could improve the humidity fields at bottom layers further (0 to500m, namely 0 to 2 layer). Compared to the usage of Global Positioning System (GPS) only SWD,the inclusion of multi-GNSS SWD does not significantly improve the tomographic results at alllayers due to the almost unchanged dispersion of puncture points of GNSS signals. However,increases in the ground GNSS stations can benefit the tomography, with improvements of morethan 10% at bottom and middle layers. Higher tomographic horizontal resolution can furtherslightly improve the tomographic results (about 3-6% from 0.5 to 0.25 degrees), which, however,will also increase the computational burden at the same time.