An investigation of a voxel-based atmospheric pressure and temperature model

Atmospheric pressure, temperature and zenith wet delay (ZWD) are important meteorological parameters for GNSS (Global Navigation Satellite Systems) data processing and GNSS meteorology. Due to the fact that not all GNSS stations are physically equipped with meteorological sensors, empirical models are widely used for the prediction of the atmospheric pressure (which can be used for the calculation of zenith hydrostatic delay (ZHD)), temperature and ZWD. However, if the reference height significantly differs from the height of the user site, the quality of the predicted meteorological parameters may be poor due to the height difference between the user site and the reference level. To address these issues, a voxel-based atmospheric pressure and temperature model, named PVoxel, was developed to improve the accuracy in the determination of both atmospheric pressures and temperatures. Ten-year ERA5 monthly mean reanalysis data were used for the development of PVoxel, and each monthly mean atmospheric pressure and temperature at each of the four selected reference heights (0, 4, 10 and 15 km) and the ZWD with its vertical decay parameter at 0 and 4 km over all globally distributed grid points (horizontal resolution 1° × 1°), i.e., at the nodes of the 3D voxels, were determined. Then, the characteristics of the annual and semi-annual variations of these parameters in the temporal domain for each node were modeled. The PVoxel model can be used to predict atmospheric pressure, temperature and ZWD at any geographic location and time. The model was evaluated by comparing the model-predicted results for the sites of all globally distributed radiosonde stations against the corresponding radiosonde data in 2019. The results showed that the atmospheric pressure and temperature predicted by PVoxel were improved significantly compared to four advanced publicly available empirical models, i.e., UNB3m, GPT3, IGPT and GTrop, especially at high altitudes. The significant improvement in