GIOTTO: A coupled atmosphere‐ocean general‐circulation model: The tropical Pacific

• Published in: Quarterly Journal of the Royal Meteorological Society
• Main Authors: Fischer, Martin, Navarra, Antonio
• Format: Article in Journal/Newspaper
• Language: English
• Published: Wiley 2000
• Subjects: Pacific; Sea ice
• Online Access: http://dx.doi.org/10.1002/qj.49712656702; https://api.wiley.com/onlinelibrary/tdm/v1/articles/10.1002%2Fqj.49712656702; https://rmets.onlinelibrary.wiley.com/doi/pdf/10.1002/qj.49712656702

Abstract A new coupled general‐circulation model (GIOTTO) has been developed. The individual components are composed of the atmosphere model, ECHAM‐4, and the ocean model, MOM (Modular Ocean Model)‐1.2. The model domain is global, and no flux correction is applied. The coupling is active between 60°N and 60° S. Poleward of 60° the atmosphere is forced by the climatological sea surface temperature (SST), and the ocean is relaxed towards the climatological SST and sea surface salinity. Further, the sea‐ice coverage is prescribed. The coupling interval is set to two hours to resolve the diurnal cycle. In this paper we describe the design of the model, and discuss results of a coupled 20‐year integration. The representation of the mean state is realistic, although there is an overall cold SST bias of about one degree centigrade in the tropics, and a tendency to simulate a double Inter Tropical Convergence Zone. The annual cycle, as simulated in the equatorial Pacific, is too weak in the east Pacific and too strong in the warm‐pool region. The phase, however, is well captured. The SST variability in the equatorial Pacific is underestimated by about 30%, and the anomalies are too confined to the equator. The main features of El Niño‐Southern Oscillation (ENSO) dynamics, like propagation of heat‐content anomalies, reflection of equatorial Kelvin and Rossby waves, and westerly wind bursts, however, are correctly represented by the model. A variability analysis based on empirical orthogonal functions indicates that the ENSO mechanisms are simulated correctly. The model also appears to be well balanced with a remarkably low SST drift (0.5 degC decade −1 ), and a realistic equatorial thermal structure. We are, therefore, confident that the model can be used for experimental seasonal predictions.