Tropospheric Correction Models for Local Events

Abstract

The Atmospheric Performance Assessment Facility (APAF) is a Standalone Test Case lead by DLR and performed together with the University of Armed Forces, Munich, in the frame of ESA’s Galileo System Testbed (GSTB V1). One task within APAF was to analyse the performance of tropospheric correction models during local events, i. e. in the case of extreme weather situations or rapid changes of local weather conditions. Different user correction models for the zenith delay have been compared with 15 years of radiosonde measurements for 11 globally distributed sites in order to find the distributions of the residual zenith delay for different correction models, in particular the edges of the distributions with large residual delays. Three different methods of radiosonde data processing have been tested and compared, in order to investigate the error due to the data processing itself. Nine correction user models have been analyzed and compared, which include models using surface measurements of meteorological parameters like Hopfield, modified Hopfield models and the Saastamoinen model as well as “blind” models based on longtime statistics and not using actual surface measurements. The latter include the RTCA-MOPS model and a new blind model developed by ESA. Radiosonde measurements are taken only 2 times a day, so that short time events cannot always be discovered by them. Therefore, from 1 year meteorological measurements with 10 sec temporal resolution recorede at Oberpfaffenhofen, Germany, severe events have been selected and compared with IGS zenith delays, which were derived from post-processing of GPS measurements of the IGS network and are available for the site Oberpfaffenhofen with 2 hour temporal resolution. The paper will summarize the results and main findings of the data analysis