Geoscientific Model Development (Jan 2019)

Description and evaluation of NorESM1-F: a fast version of the Norwegian Earth System Model (NorESM)

  • C. Guo,
  • M. Bentsen,
  • I. Bethke,
  • M. Ilicak,
  • M. Ilicak,
  • J. Tjiputra,
  • T. Toniazzo,
  • J. Schwinger,
  • O. H. Otterå

DOI
https://doi.org/10.5194/gmd-12-343-2019
Journal volume & issue
Vol. 12
pp. 343 – 362

Abstract

Read online

A new computationally efficient version of the Norwegian Earth System Model (NorESM) is presented. This new version (here termed NorESM1-F) runs about 2.5 times faster (e.g., 90 model years per day on current hardware) than the version that contributed to the fifth phase of the Coupled Model Intercomparison project (CMIP5), i.e., NorESM1-M, and is therefore particularly suitable for multimillennial paleoclimate and carbon cycle simulations or large ensemble simulations. The speed-up is primarily a result of using a prescribed atmosphere aerosol chemistry and a tripolar ocean–sea ice horizontal grid configuration that allows an increase of the ocean–sea ice component time steps. Ocean biogeochemistry can be activated for fully coupled and semi-coupled carbon cycle applications. This paper describes the model and evaluates its performance using observations and NorESM1-M as benchmarks. The evaluation emphasizes model stability, important large-scale features in the ocean and sea ice components, internal variability in the coupled system, and climate sensitivity. Simulation results from NorESM1-F in general agree well with observational estimates and show evident improvements over NorESM1-M, for example, in the strength of the meridional overturning circulation and sea ice simulation, both important metrics in simulating past and future climates. Whereas NorESM1-M showed a slight global cool bias in the upper oceans, NorESM1-F exhibits a global warm bias. In general, however, NorESM1-F has more similarities than dissimilarities compared to NorESM1-M, and some biases and deficiencies known in NorESM1-M remain.