The Use of NWP Products in NearReal-Time GPS Data Processing

摘要

We investigate the eu000bect of the choice of a priori"rninformation for the troposphere in the evaluation ofrnGlobal Positioning System (GPS) data, since the commonlyrnused standard atmosphere values may diu000ber considerablyrnfrom the actual atmospheric conditions. Arnmeans to obtain more realistic initial values may bernthe use of numerical weather prediction (NWP) models.rnThus, the impact of the ingestion of zenith hydrostaticrn(ZHD) and zenith wet delay (ZWD) valuesrnderived from a operational NWP model into the GPSrndata processing is analysed.rnAs meteorological information we use 1 hour time seriesrnof ZHD and ZWD that have been forecasted byrnthe NWP model HIRLAM. The investigation is donernfor the three International GPS Service for Geodynamicsrn(IGS) stations Villafranca, Wettzell, and Zimmerwald.rnIn the case of not estimating the troposphere,rn1-h solutions furnish signifcantly better results whenrnusing the HIRLAM derived tropospheric information.rnThe repeatabilities of the vertical coordinate componentrndrop from 0:1 : : : 0:2 m with standard a-prioris torn0:02 : : : 0:05 m in the HIRLAM case. In case the troposphericrnparameters are being estimated as well, thernresults from hourly batches become unclear. While twornstations have smaller repeatabilities with the standardrna-priories, one station shows the opposite behaviour.rnThus, for small estimation intervals (1 hour and less)rnthe choice of initial values for the troposphere may haverna signifcant eu000bect on the estimation results.