Ensemble Dynamical Forecast Model Calibration for Seasonal Climate Prediction

The World Climate Service uses a special technique called “calibration” to ensure the forecasts of probabilities of weather events are properly estimated. It’s a fairly complex technique that requires processing a significant amount of retrospective data from dynamical models such as the CFSv2 and the ECMWF.

The subseasonal and seasonal dynamical model calibration methodology used by the World Climate Service has been published in the peer-reviewed literature. It can be found at

“Calibration and combination of dynamical seasonal forecasts to enhance the value of predicted probabilities for managing risk”

or at

https://link.springer.com/article/10.1007%2Fs00382-013-1764-2

The result of calibration is an improved understanding of the risk of adverse (or beneficial) weather or climate events.

In this example, the monthly seasonal forecast depicts the forecast shift in the temperature distribution based on the ECWMF seasonal forecast model ensemble one month in advance.