Modeling weather fronts to improve GPS heights: A new tool for GPS meteorology?

The precision of vertical position and atmospheric water vapor content dete rmined by the Global Positioning System (GPS) is limited by errors due to t ropospheric delay. One factor is the spatial and temporal variability in tr opospheric refractivity caused by passing weather fronts. We can explain so me of the temporal characteristics of estimated tropospheric delay in terms of a simple path delay model as a function of frontal parameters. These re sults suggest that GPS could be used to estimate the geometry and passage t ime of a frontal zone. We have developed indices which detect tropospheric variability from GPS data alone; the detection rate of fronts with this app roach is up to 70%. Once detected, we eliminated days affected by fronts or other tropospheric variability from the time series of station height esti mates, resulting in improved long-term repeatability. The additional varian ce attributable to fronts is estimated to be up to 10 mm(2) at Herstmonceux ;, England, where fronts occur every 2-3 days. The effect of fronts on the horizontal station component is up to 80% smaller than for the vertical. St udies in the field of GPS meteorology may be improved by estimating frontal parameters.