Vertical Reflectivity Correction for Winter Precipitation in Mountainous Region

Estimating precipitation accurately in regions with complex terrain poses a significant challenge for meteorologists. In Taiwan, for instance, radar-derived quantitative precipitation estimation (QPE) exhibits substantial underestimation in the northeastern coastal areas (YiLan district) because the nearest radar is located on a hilltop of 766 meters above mean sea level. The lowest unblocked radar beam over the low-lying coastal areas is around 1~1.5km above the ground, which overshoots the precipitation processes near the surface. In the wintertime northeasterly precipitation system, the condensational growth of raindrops near the surface contributes significantly to the rainfall amounts, and the lack of radar observations near the ground often results in a significant underestimation in the radar QPE.

To mitigate this problem, we examined the correlation between radar reflectivity observations aloft and the Parsivel disdrometer data on the ground in YiLan district. We explored a vertical reflectivity correction scheme to improve the radar QPE accuracy in the area. Using 14 cases from 2021-2022, an automatic vertical reflectivity correction scheme was developed based on empirical relations between the radar and disdrometer data. The radar QPE with the correction showed a significant improvement for the cool season stratiform rainfalls. Our study also showed large temporal variabilities of the relationship between the reflectivity aloft and at the ground for different rainfall regimes. Therefore, such vertical correction would only benefit the radar QPEs when the empirical relationships are representative.