The multi-model is a product that combines different global models into one forecast by dynamic weighting.
A major advantage of the multi-model is that extreme solutions from one model do not bias the forecast in that direction.
The multi-model will therefore never represent an outlier variant and thus can be taken as a good starting point for more detailed considerations.
If, for example, the multi-model predicts extremely high amounts of precipitation for a region, then this is a serious indication that it will come true, since this forecast is supported by several independent weather models and the impact of extreme solutions have been damped.
For some parameters we also offer the maximums and minimums of the model ensemble, from which you can see the upper or lower end of the predicted values and their variation.
Furthermore, we can use the multi-model forecasts to calculate important indices for specific severe weather hazards, such as the flash flood index.
Also, with the multi-model, as in an EPS (Ensemble Prediction System), we can derive a probability of the occurrence of certain threshold values (e.g. how high is the probability that the temperature will fall below 0°C?).
We offer such parameters in the "Probabilities" subgroup.