2015

This study presents a method for using the CAPPI data from a weather radar to verify forecasts of 24 h accumulated precipitation from a numerical weather prediction (NWP) model, during 2010–2012. The radar used in this study consisted of a 2∘ beam width, Doppler and single polarization, S-band radar, located at the Meteorological Research Institute (IPMET) of Sao Paulo State University, Bauru, Sao Paulo, Brazil. A tuned version of the Eta model was used in the verification, though any model could be used with a few minor adaptations. The model, used actively at IPMET, had a horizontal grid spacing of 10 km, and was defined with the lateral boundary conditions from the Global Circulation Model of the Center for Weather Forecasting and Climate Research of the Brazilian Institute for Space Research. A linear correction was applied to the radar data, using selected rain gauges from the state of Sao Paulo’s meteorological observation network, to create a reference series for both radar and NWP quantitative precipitation estimates. The reference data were used to verify the rainfall rates forecasted with the NWP, in terms of both their spatial distribution and the rainfall quantity at ground level. The results agreed well with the specific ranges of rainfall values, but there were situations where the radar data presented limitations for the verification. Ways in which to improve the methodology presented here are discussed. The current study provides an opportunity to use a high-resolution data set to verify predicted rainfall across a large spatial coverage, particularly in places which lack rain observational data.

BASSAN, JOSÉ MARCIO; MARTINS, JOÃO EDUARDO MACHADO PEREA ; SUGAHARA, SHIGETOSHI ; DA SILVEIRA, REINALDO BOMFIM . Application of weather radar CAPPI data to verify NWP rainfall accumulation data. Meteorology and Atmospheric Physics (Print), v. 127, p. 1-19, 2015.