A variational data assimilation system for ground‐based GPS slant delays

Abstract

Tropospheric delay affects the propagation of the microwave signals broadcast by the Global Positioning System (GPS) satellites. Geodetic processing software enable estimation of this effect on the slanted signal paths connecting the satellites with the ground-based receivers. These estimates are called slant delay (SD) observations and they are potentially of benefit in numerical weather prediction.The three-dimensional variational data assimilation system of the High-Resolution Limited-Area Model (HIRLAM 3D-Var) has been modified for data assimilation of the SD observations. This article describes the ground-based GPS observing system, the SD observation operator, the estimation of the observation-and background-error standard deviations, the methodology of accounting for the observation-error correlation, and the tuning of the background quality control for this observation type.The SD data assimilation scheme is evaluated with experiments utilizing hypothetical observations from a single receiver station, as well as a single-case experiment utilizing real observations from a regional GPS receiver network. The ability of the data assimilation system to extract the asymmetric information from the SD observations is confirmed. In terms of analysis increment structure and magnitude, SD observations are found to be comparable with other observation types currently in use, provided that the observation-error correlation is taken into account.

A Numerical Study on a Mesoscale Convective System over a Subtropical Island with 4D-Var Assimilation of GPS Slant Total Delays

“…They calculated this error using data from a German observation network and European mesoscale model predictions. Järvinen et al (2007) inferred a ZTD error of 11 mm from a comparison of modeled and observed STDs.…”

“…Furthermore, the observational error depends on the elevation angle. Järvinen et al (2007) and Bauer et al (2011) used the same type of error model. By comparing GPS-STD assimilation with GPS-PWV assimilation and considering the GPS-ZTD error, we estimated that the GPS-STD error in the zenith direction was 50 mm and adopted the error model below (Eq.…”

“…The error of 50 mm and that calculated with Eq. (4) are large compared with the values suggested by Järvinen et al (2007) and Bauer et al (2011), but we chose to use a conservative value and function for this first GPS-STD assimilation trial because our assumption of linear propagation (see Section 3.2a) might result in large observational errors at low elevation angles. In addition, when the elevation angle exceeds 85°, the standard deviation of the departure value (observation minus model; Fig.…”

“…Using a fifth-generation mesoscale model (MM5)-4D-Var system (Zou and Kuo 1996) with 27-km horizontal grid spacing to assimilate GPS-SW observations, Ha et al (2003) obtained good squall-line simulation results. Järvinen et al (2007) used a high resolution local-area-modeling three-dimensional variational system (HIRLAM; Gustafsson et al 2001) with 9-km horizontal grid spacing for GPS-STD assimilation, and demonstrated that the magnitude of the GPS-STD assimilation analysis increments exceeded that of the GPS-ZTD assimilation analysis increments, and that the horizontal distribution of analysis increments differed between Vol. 91, No.…”

Development and optimization of the IPM MM5 GPS slant path 4DVAR system

“…Recently, JÄRVINEN et al (2007) analyzed the data assimilation of hypothetical and real STD (Slant Total Delay) observations in the framework of the HIRLAM (High-Resolution Limited Area Modeling) 3DVAR (3-dimensional variational assimilation) system. The experiments with hypothetical data showed that their assimilation results in reasonable specific humidity analysis increments.…”