Prediction skill of ENSO and ENSO teleconnections in NorCPM: A comparative analysis with C3S

In recent decades, high demand from stakeholders and policymakers has intensified research efforts toward operational seasonal forecasts. This is particularly relevant when it comes to El Niño-Southern Oscillation (ENSO), the primary driver of year-to-year climate variability with global socioeconomic impacts. While various state-of-the-art systems effectively target ENSO forecasting, the relative influence of different modeling approaches on prediction skill remains incompletely understood. Here, we evaluate an upgraded version of the Norwegian Climate Prediction Model (NorCPM), which employs anomaly initialization and ocean-only data assimilation (DA), against Copernicus Climate Change Service (C3S) systems using full-field DA in both atmosphere and ocean components. Through systematic comparison, we examine the spatio-temporal prediction skill of sea surface temperatures (SST) in the tropical Pacific and associated teleconnection patterns. Despite its ocean-only constraint and lower resolution, NorCPM demonstrates competitive performance for ENSO’s SST signature. However, the system exhibits systematic amplitude errors in representing mid-latitude atmospheric circulation and precipitation patterns. Our comparative analysis reveals three key findings: (1) ocean-only initialization achieves comparable skill to comprehensive approaches during the mature phase of ENSO; (2) model resolution impacts forecast quality in the eastern Pacific boundary region, where fine-scale ocean-atmosphere coupling processes are crucial, particularly during transition seasons; and (3) initialization strategy more strongly influences the representation of global teleconnection patterns than local ENSO dynamics. While targeted experiments are encouraged to establish causal links between specific model features and prediction capabilities, these results provide practical insights for optimizing prediction systems, particularly for applications requiring accurate precipitation predictions in regions affected by ENSO.