2,- 1PhD Candidate Hassania School for Public Works (EHTP), Casablanca, Morocco (karima.moutachaouiq@gmail.com)
- 2Researcher, General Directorate of Meteorology, Casablanca, Morocco (bari.driss@gmail.com)
- 3Researcher, Geophysical Institute, University of Bergen & Bjerknes Centre for Climate Research, Bergen, Norway (Noureddine.Omrani@uib.no)
- 4Professor, Department of Mathematics, Computer Sciences and Geomatics, Hassania School for Public Works (EHTP), Casablanca, Morocco (nafiri@ehtp.ac.ma)
A realistic representation of sea surface temperature (SST) variability in climate models is essential for seasonal-to-interannual forecasting and for understanding large-scale climate oscillations. This study evaluates the impact of three initialization strategies in the NorCPM coupled climate model on the structure and temporal evolution of the leading modes of global SST variability over the 1980–2010 period. The analyzed strategies include a free-running simulation (FREE), ocean data assimilation using an ensemble Kalman filter (ODA), and atmospheric nudging of wind and temperature anomalies (NUDA_UVT). Model results are evaluated against the HadISST observational dataset.
Empirical Orthogonal Function (EOF) analysis is applied to monthly SST anomalies computed over the full globe and using all calendar months, without regional restriction or seasonal stratification. This framework enables a consistent comparison of the dominant large-scale SST variability modes across all datasets. The results indicate that ocean data assimilation (ODA) best reproduces the leading ENSO-related mode, achieving a spatial correlation of 0.98 and the lowest root mean square error in its principal component (RMSE = 1.70). For the second mode, associated with lower-frequency variability, atmospheric nudging (NUDA_UVT) shows improved spatial agreement (correlation = 0.90) compared to ODA. The free-running simulation captures the main spatial structures but displays systematically larger temporal errors.
These findings demonstrate that ocean data assimilation is the most effective strategy for representing ENSO-like variability in NorCPM, while atmospheric nudging provides added value for lower-frequency modes. As a perspective, this work will be extended to investigate the impact of initialization strategies on atmospheric fields, as well as to explore SST variability at specific seasonal and regional scales.
How to cite: Moutachaouiq, K., Bari, D., Omrani, N.-E., and Nafiri, S.: Impact of Different Initialization Strategies on the Representation of Dominant SST Variability Modes in the NorCPM Coupled Climate Model, EGU General Assembly 2026, Vienna, Austria, 3–8 May 2026, EGU26-17596, https://doi.org/10.5194/egusphere-egu26-17596, 2026.
