Comparison of software and techniques for water vapor estimation using German near real-time GPS data

To deliver tropospheric water vapor contents derived from GPS ground networks for the assimilation into numerical weather prediction models the GPS data have to be analyzed within near real-time with a delay of about one hour. Two near real-time processors for the trophospheric water vapor monitoring developed at IEEC Barcelona and GFZ Potsdam are compared using one week of data acquired in the German near real-time network. The resulting small differences in the integrated water vapor being on the level of 1–2 mm precipitable water vapor reflect that both processors fulfill the requirements for numerical weather predictions. Therefore the standardization for processors should not be too strict not to obstruct a fruitful competition in the software developments and improvements.