A Study on Blending Radar and Numerical Weather Prediction Model Products in Very Short-range Forecast and Nowcasting

摘要

This paper presents a new blending approach: Using meso-scale Numerical Weather Prediction (NWP) products as environment field, coordinating with weather radar information, blending radar data and NWP model products together, and finally being applied to short-range forecast and nowcasting. Considering the fact that the weight of radar extrapolation and NWP model products should be adjusted because they will change with time, three solutions are offered to calculate NWP weight to blend radar extrapolation and NWP model products in a certain area. At last, RMSF method is used to verify and examine the example. After qualitative comparison with figures, the results showed: blending with equal weights and blending with sin~2(at+b) as NWP weight improved radar extrapolation and NWP results. A new effective way has been discussed preliminarily. Hoping it is of reference value to short-range forecast and nowcasting in further research.