Page - 389 - in Book of Full Papers - Symposium Hydro Engineering

It was found that the quantile mapping method does not produce results that are usable for this study. The reason is that quantile mapping does not take the temporal sequence of values into account, but only adjusts the distribution of values. If, as is the case here, the input time series already contain large differences regarding the temporal sequence of values, quantile mapping can even amplify these differences and produce time series whose goodness of fit parameters are in some cases worse than those of the original time series (see Table 1 and Fig. 3). Fig. 3 Comparison of NOAA forecast data before / after qmap correction, example eastern Germany Performing the bias correction using monthly factors consistently produced better goodness of fit parameters (see Table 1). Tests using quantile mapping and monthly factors in sequence also produced worse results than using monthly factors alone (see Fig. 4). 389