A rapid ray tracing method to evaluate the performances of ERA5 and MERRA2 in retrieving global tropospheric delay

Abstract

Atmospheric reanalysis plays an important role in retrieving the atmospheric tropospheric delays with ray tracing for space geodetic techniques. In order to represent the complex weather and climate conditions better, the spatiotemporal resolutions of the newly developed atmospheric reanalysis products are improved significantly. The increased spatiotemporal resolution provides a great opportunity to improve the accuracy of the tropospheric delays derived from ray tracing, but it remains a challenge due to the highly increased computation costs. In this paper, we develop a rapid ray tracing method with refined height interval determination to accommodate the increased spatiotemporal resolution of the atmospheric reanalysis products. The accuracy of this method was validated by the 2010 International Association of Geodesy (IAG) Working Group 4.3.3 ray tracing Comparison Campaign reference results. Zenith and slant delays were obtained by tracing 342 global International GNSS Service (IGS) stations. Compared to the traditional method, this reduced memory footprint by 16.1%, global refractivity field construction time by 13.6%, and per ray trace time by 22.5% while maintaining accuracy. Based on this methodology, ray tracing using state-of-the-art fifth-generation European Centre for Medium-Range Weather Forecasts (ECMWF) Re-Analysis (ERA5) and second Modern-Era Retrospective Analysis for Research and Applications (MERRA2) at 342 IGS stations assessed tropospheric delay performance in 2021. Results showed significant ERA5 and MERRA2 slant delay and mapping factor differences up to the decimeter level, especially for the wet component. Additionally, using IGS Zenith Total Delay (ZTD) as a reference, ERA5 ZTD bias and Root Mean Square Error (RMSE) were 2.3 and 11.9 mm, versus that of 1.8 and 16.2 mm for MERRA2 ZTD. At extreme weather-affected AIRA stations over August 5-9, 2021, ERA5 ZTD mean and RMSE differences were -3.0 and 19.8 mm, and -5.3 and 21.7 mm for MERRA2 ZTD. Tropospheric delays and models derived from ERA5 can support space geodetic applications given improved performance and temporal resolution.