Rainfall estimation using radar reflectivity during TiMREX

A preliminary study examining the relationship between radar reflectivity and rainfall estimates has been conducted in Taiwan during the Terrain-influenced Monsoon Rainfall Experiment (TiMREX) for the summer months of 2008. Accurate rainfall estimations using radar data is desired due to the high spatial and temporal observations. Estimations for rainfall were made for both SPOL and Chi-Gu radars located on the southeast coast. A 40 km square box located between the two radars was used for comparing rainfall estimations to rain gauge observations. Within this box, the radar beam was approximately 1 km above the ground and much of the rainfall was synoptically forced. Using the Z-R relationships used for Taiwan radar operations (Z=32.5R1.65) and a disdrometer located at the supersite about 30 km northeast of SPOL (Z=280R1.38), average rainfall estimation over the box was determined. On average, the two radars were comparable, although SPOL tended to average slightly higher. When compared to the rain gauge averages over the area, this relationship fit fairly well, with slight over-estimations for both radars. The main exception was with heavy rainfalls (>100mm) where it was overestimated. On the other hand, when using the disdrometer relationship, the rainfall was underestimated substantially. The reason for these large errors will be explored, with suspicions of the difference in height being the main contributor. Using disdrometer data measured within 10km of SPOL, the difference in height will be reduced, resulting in a more agreeable height comparison. Time height plots over the disdrometer locations of reflectivity (dBZ) and differential reflectivity (ZDR) will also be explored for investigating the nature of any differences in the vertical distribution.