Tool to compute the RMSE between the observed and modelled patterns of variability obtained through classification and their relative relative bias in the percentage of occurrence and the persistence of each mode. Modes of variability are first obtained through EOFs and/or k-mean clustering applied to the anomalies for any season or month, where the user can specify the number of clusters to be computed (e.g. 3 for Arctic sea ice or 4 for North Atlantic sea level pressure). The user also has the option to remove linear or second order trends from the data before obtaining the modes.


    Description of user-changeable settings on webpage':'
    Selection of region.
    Selection of variable.
    Selection of temporal period.
    Selection of RCP scenario.
    Selection of models.
    Selection of detrending method.
    Selection of EOF or spatial data.