Fine scale distribution of precipitation in the Voss area

Abstract

The precipitation in Western Norway is strongly affected by its position in the mid latitude conveyor belt of synoptic systems. When the warm and moist air impinges on the complex mountainous terrain, large variabilities in local precipitation amounts arise. This master thesis has examined the orographic modifications of precipitation in an area around the city of Voss in Western Norway. To do so, two observational campaigns with HOBO rain gauges have been conducted, resulting in a valuable dataset for the area with high horizontal and temporal resolution. The observed precipitation has been studied and large areal variability was detected. In the period May to November 2014 a synoptic scale difference of 114 % was found between Vasslii (674 mm) and Hisdalen (1443 mm), whereas a smaller mesoscale signal shows a total precipitation difference of up to 40 % on gauges located within a distance of 10 km. Mainly during high wind situations, defined as 10 ms^−1 or above, the fine scale distribution of precipitation has large inhomogeneities. During lower wind situations, below 6 ms^−1, the precipitation pattern seems to be more homogeneous, but with some variability at the coastal stations. Topography importance has been investigated and the effects of up- and downwind barriers in the vicinity are discussed. The dataset has also been used for a model validation and sensitivity test for the period of the major flooding event in late October 2014. Large sensitivity was found for grid resolution whereas the spectral nudging settings and terrain smoothing had less influence on the reproduced precipitation. The best model option for this study, judging by RMSE, was found to be the option with 1 hour relaxation time, with nudging of wavelengths larger than 677 km zonally and 609 km meridionally and a grid resolution of 1 km. The optimized model set up was capable of representing the observed precipitation amounts rather well with respect to absolute values, timing of events and spatial variability, but further optimization is still needed for a satisfactory performance at some complex terrain areas. A closer investigation of the flooding event has been performed, both observational data and model data from WRF have been examined. For precipitation distribution, large scale driving mechanisms as stability, humidity and the seeder feeder effect have been identified as important factors. Furthermore smaller scale features, with a horizontal scale of less than 5 km, e.g. gravity waves, small scale forced ascent and spill over effects were found to have a major influence on the variability of the precipitation.