Measuring and modeling the slant wet delay with GPS and the ECMWF NWP model

Abstract

We describe the data fusion of ground-based Global Positioning System (GPS) slant total delay values and ray-traced slant dry delay values from the European Centre for Medium-Range Weather Forecasts (ECMWF) model to determine wet delay values for elevations down to −0.51°. They have a magnitude up to several meters and provide valuable information on the distribution of water vapor. The slant wet delay values for one GPS station and one GPS satellite are determined with the Bernese Software V4.2 using a modified elevation dependent weighting scheme to handle GPS observations below the horizon. We compare the eigenvectors of the wet delay correlation matrix to predictions from the Treuhaft–Lanyi (TL) model and find a good agreement for the three largest eigenvalues representing the major wet delay fluctuations. This result shows that the TL-model describes certain aspects of refractivity fluctuations within the troposphere accurately. We use it to propose a new model for ground-based slant wet delay values at very low elevations. In addition to the zenith delay value the model contains so-called fluctuation parameters. The model is optimal in the sense that, depending on the number of fluctuation parameters, the slant wet delay can be modeled with an elevation independent accuracy.