Hybrid Regularized GPS Tropospheric Sensing Using 3-D Ray Tracing Technique

Abstract

Tropospheric water vapor plays an important role in the atmospheric processes. It is a major uncertainty parameter in numerical weather models due to the large and fast variation in spatiotemporal scales. Because of its variations, modeling the spatiotemporal distribution of water vapor is a complex problem. The 4-D Global Navigation Satellite Systems tropospheric tomography is a powerful technique to compute the accurate spatiotemporal distribution of water vapor. Tropospheric tomography usually is an ill-conditioned inverse problem. Therefore, using the regularization methods is inevitable. The precision of a tomography technique depends on the accuracy of the GPS-derived tropospheric slant wet delay, the ray tracing technique, and the regularization method. In this letter, the 3-D ray tracing technique based on Eikonal equations and hybrid regularization frameworks is used to increase the precision of the 4-D tropospheric tomography technique. After solving the tropospheric tomography problem, the modeled water vapor is validated using radiosonde observations and meteorological data obtained from synoptic stations. Using hybrid regularization techniques reduces the root mean square of the results of common regularization methods by an average of 0.39 g/m3.