Grid-Based Bias Correction for Mesoscale Weather Forecasting: a statistical insight
Department of Statistics Univ. of Waterloo,Ontario
We consider the comparative analysis of three methods for objective grid-based bias removal in mesoscale numerical weather prediction models. The first technique is the "local-observation-based" (LOB) method that extends further the approaches of Mass (2004) and Tebaldi (2002) and is focused on utilizing the information obtained from meteorological stations or neighbor grids in the proximity of a site of interest. The bias at a site of interest might then be considered as a spatio-temporal function of weighted information on the past biases observed in the cluster of neighbors during a certain time window. The second method is an elaboration of model output statistics (MOS), combining several modern multiple regression techniques such as the classification and regression trees (CART) and the alternative conditional expectation (ACE), and therefore is named CART-ACE method. The CART-ACE method allows to represent nonlinear aspects of the bias in a parsimonious linearized statistical model. In final, the third considered method is the natural combination of the LOB and CART-ACE methods where the information provided by the LOB method is interpreted as an extra predictor in the regression model of the CART-ACE method. The proposed methods are illustrated by a case study of an observation-based verification and bias correction of MM5 48-hour surface temperature forecasts over the US Pacific Northwest.