Improving radar precipitation estimation and 3D radar mosaic analysis using dynamic vertical profile of reflectivity

Radar based rainfall estimates are usually derived from reflectivity observations in the lowest unobstructed radar beams. Due to the earth's curvature and positive scan elevation angles, radar observations at far ranges can be as high as several kilometers above the ground. As a result, large data voids exist below the lowest radar beams when volume scans of radar data are transformed onto a 3-D Cartesian grid. Further, inferring precipitation at the ground from radar data can involve uncertainties because of non-uniform vertical profiles of reflectivity (VPR) between the radar beam and the ground. Previous studies have shown that radar rainfall products were poor at the far ranges due to non-uniform VPR, especially for winter precipitation events. Techniques have been developed to correct for range-dependent bias in the radar rainfall estimates due to non-uniform VPR. These VPR correction techniques have potential of significantly improving radar rainfall estimates for large-scale stratiform precipitation, especially in complex terrain.

In the current study, VPR of different space and time scales in addition to storm types are derived from radar data at close ranges and are applied at far ranges for the adjustment of radar-based precipitation and for gap-filling in the 3D reflectivity mosaic analysis. Initially a volume scan of reflectivity data are segregated into convective and stratiform precipitation according to the vertical structure of the reflectivity and the environmental thermodynamic field. Then the reflectivity observations within a range zone close to the radar are used to estimate two VPRs, one for convective precipitation region and another for stratiform precipitation region. The convective and stratiform VPRs are then fittingly applied at far ranges to fill in voids between the bottom of the lowest radar beams and the ground for the 3-D radar reflectivity mosaic analysis. The VPR corrected reflectivities are then used to estimate precipitation in addition to providing a more accurate severe weather attributes. Rain gage observations will be used to evaluate the performance of the VPR adjustment.