The precision of humidity in GNSS tomography

The GNSS signal passing through the layers of the atmosphere, including troposphere is subjected to many effects. One of the effect is a change of signal phase propagation speed due to the change of density. The large fluctuation of density is due to water vapor content - the most important greenhouse gas. The ground GNSS tomography is a tool to obtain the wet refractivity of the discrete model above the network of receivers. The wet refractivity could be converted into water vapor partial pressure, using the additional temperature information. This paper presents rigorous approach to obtain uncertainties of the final product - water vapor partial pressure in the voxel. All factors influencing the final uncertainty has been taken into account, mainly uncertainty of: Zenith Troposphere Delay, pressure, antenna phase center height, satellite elevation angle, mapping function, tomography solution and temperature. The analyzed error budget shows, great influence of cofactor matrix topology on tomography solution uncertainty. It is also shown that ground GNSS tomography is feasible to obtain water vapor partial pressure in the troposphere up to 0.5 hPa so in average 6 km above the ground.