Evaluating the physical and biogeochemical state of the global oceancomponent of UKESM1 in CMIP6 Historical simulations

2020 
Abstract. The ocean plays a key role in modulating the climate of the Earth system (ES). At the present time it is also a major sink both for the carbon dioxide (CO2) released by human activities as well as for the excess heat driven by the resulting atmospheric greenhouse effect. Understanding the ocean's role in these processes is critical for model projections of future change and its potential impacts on human societies. A necessary first step in assessing the credibility of such future projections is an evaluation of their performance against the present state of the ocean. Here we use a range of observational properties to validate the physical and biogeochemical performance of the ocean component of UKESM1, a new Earth system (ESM) for CMIP6 built upon the HadGEM3 physical climate model. Analysis focuses on the realism of the ocean's physical state and circulation, its key elemental cycles, and its marine productivity. UKESM1 generally performs well across a broad spectrum of properties, but it exhibits a number of notable biases. Physically, these include a global warm bias inherited from model spin-up, excess northern sea-ice but insufficient southern sea-ice, and sluggish interior circulation. Biogeochemical biases found include shallow remineralisation of sinking organic matter, excessive iron stress in regions such as the Equatorial Pacific, and generally lower surface alkalinity that results in decreased surface and interior dissolved inorganic carbon (DIC) concentrations. The mechanisms driving these biases are explored to identify consequences for the behaviour of UKESM1 under future climate scenarios, and avenues for model improvement. Finally, across key biogeochemical properties, UKESM1 improves in performance relative to its CMIP5 precursor, and compares favourably to fellow members of the CMIP6 ensemble.
    • Correction
    • Source
    • Cite
    • Save
    122
    References
    3
    Citations
    NaN
    KQI
    []
    Baidu
    map