A priori confidence index for seasonal forecasts

Work package 4 of C3S2_MF_370 proposes to design a new tool for the interpretation of multi-model ensemble seasonal forecasts, consisting in a confidence index to be included whenever a new real-time forecast is issued (Task 4.3). The aim of the confidence index is to provide an a priori estimation of how correct the forecast will be against the verifying observations. The focus is set on European climate, defined by average near-surface temperature and accumulated precipitation over the next season, and the corresponding atmospheric circulation. This tool is intended to bring additional information in the preparation of the real-time seasonal forecast bulletins at Météo-France, and will be experimented shortly. This deliverable is the final document for Task 4.3. It is a follow-up of the intermediate milestone M4.3.1 issued in June 2023: while M4.3.1 is mostly focusing on confidence in a forecast of a single parameter such as a grid-point value, the present report elaborates on a confidence index for the spatial patterns of forecast anomalies over Europe.