An Evaluation of the practicability of current mapping functions using ray-traced delays from JMA Mesoscale Numerical Weather Data

The the Japan Meteorological Agency (JMA) meso-scale analysis data (MANAL data) which we used in our study provides temperature, humidity, and pressure values at the surface and at 21 height levels (which vary between several tens of meters and about 31 km), for each node in a 10km by 10 km grid that covers Japan islands, the surrounding ocean and eastern Eurasia. The 3-hourly operational products are available by JMA since March, 2006. We have simultaneously evaluated atmospheric parameters (equivalent zenith total delay and linear horizontal delay gradients) and position errors derived from slant path delays obtained by the KAshima RAytracing Tools (KARAT) through the MANAL data. Most of the early mapping functions developed for VLBI and GPS were based on the assumption of azimuthal isotropy. On the other hand, the recent geodetic analyses are carried out by applying the modern mapping functions based on the numerical weather analysis fields. The Global Mapping Function (GMF) by Boehm et al. (2006), and Vienna Mapping Function (VMF) by Boehm and Schuh (2004) have been successfully applied to remove the zenith hydrostatic delay in the recent years. In addition, the lateral spatial variation of wet delay is reduced by linear gradient estimation. Comparisons between KARAT-based slant delay and empirical mapping functions indicate large biases ranging from 18 to 90 mm, which is considered to be caused by significant variability of water vapor. Position error simulation reveal that the highly variability of the errors is clearly associated with severe atmospheric phenomena. Such simulation are very useful to investigate the characteristics of positioning errors generated by local atmospheric disturbances. Finally, we compared PPP processed position solutions using KARAT with those using the latest mapping functions covering a period of two week GEONET data. The KARAT solution is almost identical to the solution using GMF with linear gradient model, but some cases tends to be slightly worse under the extreme atmospheric condition. Though we need further investigations to evaluate the capability of KARAT to reduce atmospheric path delay under the various topographic and meteorological regimes, KARAT will promise an efficient reduction of atmospheric path delays considering that the numerical weather model will be improved concerning spatial and temporal resolution