Abstract
We present a non-linear AI-model designed to reconstruct monthly meananomalies of the European temperature and precipitation based on theEuro-Atlantic Weather regimes (WR) indices. WR represent recurrent,quasi-stationary, and persistent states of the atmospheric circulation thatexert considerable influence over the European weather, therefore offering anopportunity for sub-seasonal to seasonal forecasting. While much research hasfocused on studying the correlation and impacts of the WR on European weather,the estimation of ground-level climate variables, such as temperature andprecipitation, from Euro-Atlantic WR remains largely unexplored and iscurrently limited to linear methods. The presented AI model can capture andintroduce complex non-linearities in the relation between the WR indices,describing the state of the Euro-Atlantic atmospheric circulation and thecorresponding surface temperature and precipitation anomalies in Europe. Wediscuss the AI-model performance in reconstructing the monthly mean two-metertemperature and total precipitation anomalies in the European winter andsummer, also varying the number of WR used to describe the monthly atmosphericcirculation. We assess the impact of errors on the WR indices in thereconstruction and show that a mean absolute relative error below 80% yieldsimproved seasonal reconstruction compared to the ECMWF operational seasonalforecast system, SEAS5. As a demonstration of practical applicability, weevaluate the model using WR indices predicted by SEAS5, finding slightly betteror comparable skill relative to the SEAS5 forecast itself. Our findingsdemonstrate that WR-based anomaly reconstruction, powered by AI tools, offers apromising pathway for sub-seasonal and seasonal forecasting.