Near realtime tomographic reconstruction of atmospheric water vapour using multi-GNSS observations in Central Europe

Continuously operating reference stations (CORS) are widely used to provide near realtime estimates of zenith tropospheric delays and tropospheric gradients to be assimilated in numerical weather models. Due to the latest evolution of global satellite navigation systems and ground based augmentation services, it became feasible to estimate the 3D distribution of wet refractivities in near realtime.

This paper presents a near realtime tomographic reconstuction algorithm utilizing the zenith tropospheric delays and tropospheric gradients obtained from the processing of several GNSS networks in Hungary (HU), Slovakia (SK), Romania (RO) and Ukraine (UK). The estimated zenith tropospheric delays (ZTDs) and tropospheric gradients are used to restore the slant wet delays (SWD) affecting the observed satellite-receiver range. The SWDs are used as input for a tomographic reconstruction algorithm based on the multiplicative algebraic reconstruction technique. The developed software tool includes a stepwise outlier detection module to select the most reliable slant wet delays for the tomographic reconstruction. It provides the wet refractivities in a pre-defined voxel model on an hourly basis over the HU-SK-RO-UA cross-border region.

The derived refractivity profiles have been validated with radiosonde observations. The results show that our GNSS tomography approach could reconstruct the refractivities with the standard deviation of 5 ppm below 3 km of altitude, while the standard deviation decreased to the level of 0.3 ppm at the altitude of 10 km.

The estimated tropospheric delays as well as the refractivity profiles are made available online to the meteorologists community in Little-R format and can be directly assimilated in the Weather Research & Forecast numerical model.