engleski

Bias correction of monthly temperature and precipitation from regional climate model over Croatia – calibration and validation

Application of two bias correction methods on regional climate model data was examined. Temperature and precipitation were corrected. Mean monthly temperature and precipitation sum on spatial resolution of 1 km (CroMonthlyGrid) was used as observations. Quantile mapping is used to correct higher moments of a variable, thus transfer functions were developed for each variable separately for every grid point for the period 1951-1980 and applied to regional model data for the period 1981-2010. For the same period bivariate quantile mapping bias correction method with gamma and Gaussian distribution was used to correct correlation between the two variables. Pearsons correlation coefficient was analyzed for row model data, observational data, univariate quantile mapping bias correction method and bivariate quantile mapping bias correction method and validation of the methods for the regional model was made.

bivariate bias correction, climate models

nije evidentirano