Multi-observation meteorological and GNSS data comparison with Numerical Weather Prediction model

In this study we have investigated a number of meteorological and GNSS data sources; our long-term plan is to construct an integrated model of neutral atmosphere with high temporal and spatial resolution. The integrated model will be based on a Numerical Weather Prediction (NWP) model with high spatial resolution and vertical information about various meteorological parameters. In the first step (covered by this study) we have compared the NWP model outputs with the reference data consisting of: high-quality ground-based meteorological observations, radiosonde profiles and GNSS products and we have established the accuracy of the provided parameters. The radiosonde and the GNSS observations have been found to be consistent (Zenith Total Delay (ZTD) discrepancy below 10 mm), whereas the ground-based observations and soundings detect in NWP data: 1) a negative bias in air pressure, 2) a negative bias in water vapor partial pressure, and 3) a positive bias in temperature. The standard deviation of ZTD residuals of reference data and the NWP model is on order of 20 mm. The NWP model COAMPS (Coupled Ocean/Atmosphere Mesoscale System) used in our study does not provide meteorological parameters with acceptable accuracy for positioning.