Accuracy Variability of GNSS PWV in the Range of Small and Medium Scale Areas
【摘要】：Meteorological data is a very important data source in GNSS data processing and result verification, but due to the high cost, only a few GNSS stations are equipped with parallel-site meteorological sensors. This results in GNSS users using meteorological data from tens or even hundreds of kilometers away, which brings errors to GNSS high-precision data processing. To obtain the variation of GNSS Precipitable Water Vapor(PWV) accuracy in small and medium scale regions, 28 radiosonde stations were selected around the world, each of which has more than 50 GNSS stations within 100 km. The PWV of these radiosonde stations and a total of over 2,000 GNSS stations for the period 1994-2020 were extracted, and the mean bias(BIAS) and root mean square error(RMS) of GNSS PWV was calculated and counted using the data from the central radiosonde station as a reference, then these accuracy metrics were analyzed from multiple spatial scales, and the results show that the RMS error that distance brings to PWV varies greatly between regions, with a 1 mm for a distance of less than 40 km on a tropical island, while it takes 200 km in a desert region in the middle of a continent.