A Global Assessment of Precipitable Water Vapor Derived From GNSS Zenith Tropospheric Delays With ERA5, NCEP FNL, and NCEP GFS Products

Abstract

In precipitable water vapor (PWV) retrievals from Global Navigation Satellite System (GNSS) data, the two essential parameters, namely, surface pressure ((Formula presented.)) and weighted mean temperature ((Formula presented.)), are often not available due to the lack of collocated meteorological sensors or improper data retention. Hence, this study presents a comprehensive assessment of the GNSS PWV retrieval using alternative (Formula presented.) and (Formula presented.) data from the European Centre for Medium-Range Weather Forecasts (ECMWF) ReAnalysis 5, National Centers for Environmental Prediction Final (NCEP FNL) Analysis, and NCEP Global Forecast System (GFS) products. The assessment was based on 691 globally distributed GNSS stations over the entire year of 2019. The zenith hydrostatic delay (ZHD) and (Formula presented.) integrated from the three types of numerical weather prediction (NWP) atmospheric profiles achieve varying accuracies in ranges of 2.4–3.0 mm and 1.1–1.5 K, respectively. PWVs estimated using ZHD and (Formula presented.) integrated from the NWP profiles obtain accuracies of about 1.6–2.0 mm. These PWVs are slightly better than those using ZHD and (Formula presented.) calculated by empirical models with surface pressure and temperature from the NWP datasets. The assessment of PWVs with the global pressure and temperature 2 wet model yields a root mean square (RMS) error of 3.73 mm. The relative RMS decreases from 30%-40% at high latitudes (70–80°S/N) to ∼5% around the equator. The monthly variations of relative RMS show that (a) low-latitude regions outperform the high-latitude regions, and (b) winter months have significantly worse performance than other months in both hemispheres.