Comparing Small Catchment Hydrological Responses to QPE Derived from a Locally Sited X-band Dual-Polarization Radar and that from NWS Radars at Long Range

The MZZU dual-polarization X-band radar was installed in the vicinity of Columbia, MO in July 2015 with an overarching objective of  accurately estimating rainfall over the hydrologically impaired Hinkson Creek watershed. The catchment is approximately equidistant between two National Weather Service radars: St. Louis (KLSX) and Pleasant Hill (KEAX) located, approximately, 130 and 150 km from Columbia, MO, respectively.  The work presented compares catchment rainfall accumulations derived from KLSX and KEAX in addition to  the local MZZU radar using dual-polarization radar rainfall algorithms. Subsequently, the impact of the application of the different catchment total rainfall estimates on streamflow forecasts was examined. A moderately dense network of terrestrial-based raingauges and a flow gauging station were used to validate the different rainfall and flow predictions.

It was observed that the rainfall estimates from the NWS radars suffer from inaccuracies as a result of being further from the gauges. The Quantitative Precipitation Estimation (QPE) from the MZZU radar was superior in moderate rain, but suffers from attenuation when  convective  precipitation was present, particularly when the radome was very wet. However,  implementation of R(KDP) algorithms in heavy rain and applying attenuation correction in moderate precipitation improved the QPE and derived streamflow estimates.