Comparisons of Different Methods to Determine Starting Altitudes for Dry Air Atmosphere by GNSS-RO Data

Abstract

Boundary profile evaluation (BPV) is an approach proposed in order to estimate water vapor content in the atmosphere. It exploits radio occultation (RO) observations of the signals emitted by the satellites of global navigation systems (GNSS) which are eclipsing (rising) as viewed by a low earth orbit satellite (LEO). BPV requires, as a preliminary step, the estimation of the dry background atmosphere model of refractivity (i.e., obtained from bending angle profiles) to be subtracted from the real observations in order to extract water vapor profiles. The determination of the lowest layer of the atmosphere over which the concentration of water vapor can be deemed negligible is particularly crucial for a correct application of the BPV method. In this study, we have applied three methods to set the starting altitudes for the dry air layers of the atmosphere: (1) by air temperature below some threshold values (for example, 250 K); (2) by “smooth” bending angle profiles in ROs; (3) by saturated water vapor pressure. These methods were tested with thermodynamic and bending angle profiles from 912 radiosonde excursions colocated with RO observations. For every dry air starting altitude we determined the best estimator from each of the three methods. In particular, by comparing those estimators with the quantiles and momenta of the dry air starting altitude distributions, we achieved improvements of up to 50% of the humidity profiles.