Improvements on quantitative precipitation estimation by radar applying a VPR correction method based on separating precipitation types

Errors in quantitative precipitation estimation (QPE) by radar due to the fact that radar does not scan the precipitation at ground level, but a certain height, strongly depend on the type of precipitation. The reason is that different types of precipitation present different characteristic vertical profiles of reflectivity (VPR). In fact, only in case of stratiform rain it is possible to talk about systematic errors: the overestimation produced when radar measurements intercept the bright band, and the underestimation produced when radar scans precipitation in the form of snow. Several techniques have been previously applied to quantitatively estimate the precipitation at the ground by applying a representative VPR, which is usually the same for the whole radar volume. However, it is very common to register different types of precipitation (with different characteristic VPR) in the same radar scan. So, an improvement in the QPE is expected if the radar volume is previously classified in zones with different types of precipitation, and the VPR correction is applied separately to these zones. Accordingly, a method to correct the error related to the vertical variation of the reflectivity, based on separating convective and stratiform precipitation, is presented and tested. The proposed method is first evaluated by using the first PPI measurements as reference, and the improvements achieved are verified by comparing the obtained estimates against data from a dense raingauge network.