Recent improvements in the retrieval of precipitable water vapor

Atmospheric parameters derived from Global Navigation Satellite System (GNSS) are increasingly being used in numerical weather prediction and long-term climate monitoring. We characterize the impact of three analysis strategies on the retrieval of precipitable water vapor (PW). Two of these strategies integrate GNSS analysis techniques with numerical weather prediction (NWP) analysis fields. The first uses direct mapping functions computed from NWP fields to model and estimate the atmospheric state within the GNSS software. The second uses vertical profile information from NWP fields to compute the mean temperature of the atmosphere used to scale zenith wet delay into PW. The third modification assesses the impact of horizontal gradient estimates in the retrieval of PW. The impact of these changes is quantified using extended time series of observations for ground stations within North America, and compared against observations collected with microwave radiometers. The impact on seasonal and diurnal signals is investigated.