diff --git a/mlair/plotting/postprocessing_plotting.py b/mlair/plotting/postprocessing_plotting.py
index d25736f1a9ca47984ac513805a9b458ff09ff667..b23807ae30f52f19581b188d19213fdc81c9258e 100644
--- a/mlair/plotting/postprocessing_plotting.py
+++ b/mlair/plotting/postprocessing_plotting.py
@@ -1253,6 +1253,7 @@ class PlotTimeEvolutionMetric(AbstractPlotClass):
         vmax = int(data.quantile(0.95))
         data = self._prepare_data(data, time_dim, model_type_dim, model_indicator, model_name)
 
+        # detailed plot for each model type
         for t in data[model_type_dim]:
             # note: could be expanded to create plot per ahead step
             plot_data = data.sel({model_type_dim: t}).mean(ahead_dim).to_pandas()
@@ -1262,6 +1263,18 @@ class PlotTimeEvolutionMetric(AbstractPlotClass):
             self.plot_name = f"{plot_name}_{t.values}"
             self._plot(plot_data, years, months, vmin, vmax, str(t.values))
 
+        # aggregated version with all model types
+        remaining_dim = set(data.dims).difference((model_type_dim, time_dim))
+        _data = data.mean(remaining_dim, skipna=True).transpose(model_type_dim, time_dim)
+        vmin = int(_data.quantile(0.05))
+        vmax = int(_data.quantile(0.95))
+        plot_data = _data.to_pandas()
+        years = plot_data.columns.strftime("%Y").to_list()
+        months = plot_data.columns.strftime("%b").to_list()
+        plot_data.columns = plot_data.columns.strftime("%b %Y")
+        self.plot_name = f"{plot_name}_summary"
+        self._plot(plot_data, years, months, vmin, vmax, None)
+
     @staticmethod
     def _find_nan_edge(data, time_dim):
         coll = []
@@ -1304,7 +1317,7 @@ class PlotTimeEvolutionMetric(AbstractPlotClass):
 
     @staticmethod
     def _aspect_cbar(val):
-        return min(max(1.25 * val + 7.5, 10), 30)
+        return min(max(1.25 * val + 7.5, 5), 30)
 
     def _plot(self, data, years, months, vmin=None, vmax=None, subtitle=None):
         fig, ax = plt.subplots(figsize=(max(data.shape[1] / 6, 12), max(data.shape[0] / 3.5, 2)))