A radar data based short-term rainfall prediction model for urban areas — a simulation using meso-scale meteorological modelling

A conceptual short-term rainfall prediction model for urban catchments is presented. The only input variables are surface rainfall and vertically integrated liquid water content (VIL), both measurable by radar on site. The forecast is based on simple mass balancing of water within air columns and the spatial advection of the variables using information from consecutive time steps. This paper describes a qualitative study, in which the conceptual model is tested within an idealised numerical framework: instead of using observed and potentially noisy radar data, it is initialised with output from a three-dimensional physically based meso-scale meteorological model. The meteorological model provides the microphysical data as well as detailed information on the dynamics and structure of the atmosphere, which are generally not available with observational data. The performance of the prediction model is assessed with regard to different types of rainfalls as well as respond time of the catchments. First results suggest that the conceptual model is capable of qualitatively predicting future surface rainfall development, including formation of new cells, cell splitting and decay. There is also indication that the conceptual model performs better than simple advection routines for lead times that correspond roughly to the respond times of the catchments. The results lead to the conclusion that the information of VIL may be useful for quantitative rainfall prediction, and that the conceptual model should be further developed and tested with real radar data.