GNSS_ATom

The lowermost layers of the troposphere are transport media of rapidly variable moisture fields which are essential for regional weather forecast. GNSS signals (Global Navigation Satellite Systems) experience a delay when passing the troposphere and conversely allow for estimation of humidity.

Current approaches are based on the determination of vertical signal delays above each GPS/GLONASS observing site. These are integral numbers and therefore do not reflect the change of refractivity with height within the tropospheric layers.

Moreover the horizontal resolution is limited by the mean distance of the observing sites (usually 50km-80km). Numerical weather forecast- (or nowcast-) models require 3D- information of pressure, temperature and humidity with spatial resolution of the model in close to real-time.

Project GNSS-ATom aims at the 3D-modelling of the lower troposphere in the alpine region with a horizontal resolution of 10km x 10km (about 15 vertical layers) with an update rate of 15min. The troposphere tomography is based on processing of GPS/GLONASS signals collected by an subnet of reference stations of an Austrian GNSS service provider (ÖBB).

Remaining range residual along the signal path reflect the tropospheric signal delay not captured by an apriori hydrostatic model. We make use of these residuals to reconstruct the wet refractivity of each voxel in a 3D-grid above the area of interest. This problem is ill-conditioned due to the small number of observations compared to the number of voxels. To stabilize the inversion an initial wet refractivity field is introduced which is improved by each iteration step. The computation of range residuals in each step is performed by raytracing (in-house software). The finally determined grid of wet refractivity is transformed to pressure, temperature and humidity fields in the various height levels. These fields are used for assimilation by ZAMG.

This project also investigates the potential of significant densification of the observing network by means of cheap single-frequency receivers. Furthermore the added value of processing Galileo signals (E1 and E5) will be investigated. A high density network increases the number of signals passing the voxels and therefore strengthens the ill-conditioned parameter estimation problem. The Galileo signals can be used to improve the observation geometry further. Both the single frequency network observations as well as the additional Galileo data will established by means of a signal generator available at the institute.

The resulting parameters are assimilated and validated within the high-resolution weather model AROME operated by ZAMG. This allows for a general assessment of GNSS observations as innovative sensor to contribute to weather forecast.

Coordinator

University of Technology Vienna (TU Wien)

Project partner

Central Institute for Meteorology and Geodynamics (ZAMG)

University of Technology Vienna (TU Wien)
Karlsplatz 13
A-1040 Wien
Web: www.tuwien.at