Abstract : N.3
High resolution temperature analysis in mountainous terrain

Thomas Haiden
Central Institute for Meteorology and Geodynamics

A new approach to high-resolution temperature analysis in mountainous terrain is presented. Starting with a short-range numerical weather prediction (NWP) forecast as first guess, a three-dimensional correction field is constructed from differences at station locations between observed temperature and first guess. The spatial interpolation of the correction is based on distance weighting in the horizontal and distance weighting in potential temperature in the vertical. With this method, the vertical range of influence of a station adapts in a physically meaningful way to variations in stratification. However, in situations where cold air pools are present in valleys and basins, potential temperature weighting by itself is not sufficient to adequately resolve the temperature distribution. In such cases the analysis can be further improved by including in the interpolation a topographic parameter which represents the degree to which the local topography supports the buildup of cold air. Analysis fields obtained by this method, which has been implemented into the operational analysis and nowcasting system INCA (Integrated Nowcasting through Comprehensive Analysis) are presented and verification results are shown. The sensitivity of the analysis to the choice of interpolation parameters is discussed.