A 1/16° eddying simulation of the global NEMO sea-ice-ocean system

Abstract

Analysis of a global eddy-resolving simulation using the NEMO general circulation model is presented. The model has 1/16° horizontal spacing at the Equator, employs two displaced poles in the Northern Hemisphere, and uses 98 vertical levels. The simulation was spun up from rest and integrated for 11 model years, using ERA-Interim reanalysis as surface forcing. Primary intent of this hindcast is to test how the model represents upper ocean characteristics and sea ice properties. Analysis of the zonal averaged temperature and salinity, and the mixed layer depth indicate that the model average state is in good agreement with observed fields and that the model successfully represents the variability in the upper ocean and at intermediate depths. Comparisons against observational estimates of mass transports through key straits indicate that most aspects of the model circulation are realistic. As expected, the simulation exhibits turbulent behaviour and the spatial distribution of the sea surface height (SSH) variability from the model is close to the observed pattern. The distribution and volume of the sea ice are, to a large extent, comparable to observed values. Compared with a corresponding eddy-permitting configuration, the performance of the model is significantly improved: reduced temperature and salinity biases, in particular at intermediate depths, improved mass and heat transports, better representation of fluxes through narrow and shallow straits, and increased global-mean eddy kinetic energy (by 40%). However, relatively minor weaknesses still exist such as a lower than observed magnitude of the SSH variability. We conclude that the model output is suitable for broader analysis to better understand upper ocean dynamics and ocean variability at global scales. This simulation represents a major step forward in the global ocean modelling at the Euro-Mediterranean Centre on Climate Change and constitutes the groundwork for future applications to short-range ocean forecasting.