Assessment of WRF for the year 2009

Abstract

In this present work the WRF model is tested over south Brazil. The assessment is done comparing model output variables to variables collected at INMET weather stations located in Curitiba-PR, Paranaguá-PR, São José-SC, Idaial-SC, Passo Fundo-RS and Rio Grande-RS. Conclusions based on Root Mean Square Error show that the variable which has greater predictability is Sea Level Pressure, followed by Surface Temperature, Relative Humidity and Wind Vector at Surface. Bias calculations suggest that tendencies of the model must be investigated by running WRF with different parameterizations schemes. The Bias range for Sea Level Pressure, Surface Temperature, Relative Humidity and Surface Wind are respectively -6,5 e 1,7 hPa; -5,0 e 5.0 ºC; -37.0 e 18.0 %; -13,0 e 12,0 ms-1.