Home

Climate Model Metrics

We are developing new data/model metrics to quantify the success of simulations based on their ability to reproduce observationally-measured and/or derived fields. This will allow for demonstrable progress and the reduction of model uncertainty in the projections of future climate.

Metrics include:

Atmospheric variables and surface forcing: (winds, clouds, heat and carbon fluxes, evaporation and precipitation, and sea ice)

Hydrography and circulation variables: (sea surface height, temperature, salinity, heat content, mixed layer depth, isopycnal slopes, transport in the ACC, water masses)

Lateral exchanges across 30°S: (volume, heat, fresh water, carbon, nutrients)

Biogeochemistry: (pH, dissolved inorganic carbon, nitrate/phosphate)

Climate Model Metrics figure 1
Figure after Talley/Olbers, 2014

 

Observational data comes from the World Ocean Atlas, the Carbon Dioxide Information Analysis Center (CDIAC), ARGO, AVISO, GLODAP, the National Snow and Ice Data Center (NSIDC) and others. Reanalysis data come from the National Center for Environmental Prediction (NCEP & NCEP2) and the Climate Forecast System Reanalysis (CFSR).

 

Climate Model Metrics figure 2

Westerly winds:

  • source of virtually all of the momentum in the Southern Ocean (ACC)
  • effects depends critically on the latitudinal structure
  • drives upwelling/downwelling
  • play a key rome in balancing density structure: act to weaken stratification
  • models continue to exhibit a northward bias in the location of the maximum of the wind stress.

Climate Model Metrics figure 3

Sea Ice Extent (max/min):

  • forms a physical barrier to the air/sea exchanges of heat and carbon
  • brine rejection believed to be an essential part of Antarctic Bottom Water formation
  • biases in the sea-ice cover most likely caused by errors in the simulated heat budgets of the upper ocean