Radar Nowcasting Blending with Numerical Weather Prediction for Orographic Rainfall Forecasting in Mountainous Areas

During summertime, heavy rainfall induced by mesoscale convective systems and typhoons often occurs on the Korean Peninsula. Rainfall distributions are highly variable because mountain areas cover more than 70% of the Korean territory. Heavy rainfall in mountain areas tends to be caused by a low-level warm and humid air that is lifted to the level of free convection (LFC) by a forcing driven in mountains. To mitigate and prevent disasters in mountain areas, it is essential to develop short term weather forecasting technique considering topographic effect.

Radar-based nowcasting systems are widely used for quantitative precipitation forecasting (QPF) within 6 hours using the extrapolation. It performs high prediction accuracy for short term weather forecasting (within 3 hours), but the accuracy gradually decreases with increase in the duration of prediction. It does not fully reflect the radar-based nowcasting for the development and dissipation of precipitation system over time. Therefore, the method for radar short-term precipitation forecasting blending with numerical weather prediction(NWP) is needed by considering topographic effect.

In order to consider the characteristics of orographic precipitation-producing mechanisms, atmospheric factors such as wind direction, speed, mixing ratio of water vapor were investigated. The dynamic aspects depend on the upstream characteristics of the flow and the terrain configuration were considered in high-resolution NWP model using temporal weighting function. Their blending results were validated using radar observation data.

Acknowledgements

This research was supported by the Korea Meteorological Institute under Grant KMI 2018-06210.