This menu item enables users to derive their own classification of the model results: As mentioned above classification can be performed on the mean values or on calculated percentiles. The appearing dialog lets you select the basis for further classification. The next step is to specify the number of classes. By default all stretches with a concentration of zero are grouped into one class and the highest class ranges up to the 100-percentile. The classes inbetween are defined by percentile ranges that are no longer related to a single stretch but to the entire catchment, as explained by the entry labels within the dialog.