Inversion of Water Vapor Variation During Typhoon by Quad-Constellation GNSS Tomography

摘要

The ground-based GNSS tomography technology can obtain three-dimensional distribution of water vapor. However, many voxels may not be passed through by satellite rays because of the limitation of satellite orbits and the topological distribution of ground GNSS stations, which results in the rank defect of tomography equations. In this paper, the geometric relationship between the elevation angle of the satellites and the satellite ray height intersecting with tomography boundary is deduced. The distribution of the GNSS satellite signals that can be used for the tomography at different stations is shown for the selected tomography area. In order to increase more usable GNSS rays and get stable and reliable results, the quad-constellation GNSS including GPS, GLONASS, Galileo and BDS (BeiDou navigation satellite system) systems are used simultaneously. Further, the Gaussian function is used as horizontal restriction while the mean value of three-day radiosonde profiles prior to the experiment is used as vertical restriction. To validate the four-constellation integrated tomography method, the raw GNSS observations were collected from 26 August, 2017 to 28 August, 2017 in Hong Kong to carry out the tomography experiment. During this period, there was a typhoon named Pakhar to pass the tomographic area, causing water vapor to vary dramatically. This process was descripted and captured by the presented tomographic technique. The result indicates that the current quad-constellation GNSS improves the tomographic accuracy by 5% compared with the GPS-only approach.