Home > climate models, uvic_escm > IPCC AR4 WG1 Earth System Models of Intermediate Complexity

IPCC AR4 WG1 Earth System Models of Intermediate Complexity

2012 February 21

STUDENT: Which climate models should the student seek out?
Which papers are good at outlining the points of failure?
Beside ‘institutional knowledge’, where are the strengths and weaknesses of various models catalogued?

PROFESSOR: All those answers await, and many more, once you deposit $25000 in an educational institution near you.


A first small step can be found in the IPCC AR4 WG1 in its table of Earth Models of Intermediate Complexity (table 8.3). EMICs are the models of my current interest. The UVIC_ESCM is an EMIC.

There are eight models listed with some high level summaries of their atmosphere, ocean, land, ice, and flux characteristics.

More descriptions can be found in Petoukhov et al 2005 EMIC Intercomparison Project (EMIP–CO2): comparative analysis of EMIC simulations of climate, and of equilibrium and transient responses to atmospheric CO2 doubling. Petoukhov examines many modeled climate features. Some of these are displayed in the figures below.

  1. Surface air temperature
  2. Planetary albedo
  3. Outgoing longwave radiation
  4. Precipitation
  5. Evaporation/transpiration
  6. P-E
  7. Latent heat fluxes and cloudiness
  8. CO2 direct radiative forcing
  9. The Atlantic overturning circulation
  10. Sea ice area

In Figure 1 below, we observe “Latitudinal distribution of simulated zonally averaged surface air temperature (SAT, in degC) (a, b), planetary albedo (A, as a fraction of unity) (c, d) and outgoing longwave radiation (OLR, in W/m2) (e, f) in the EMICs for DJF (a, c, e) and JJA (b, d, f).”

In Figure 2 below, we observe “Zonally averaged precipitation (P, in mm/day) (a, b), evaporation (E, in mm/day) (c, d), and precipitation minus evaporation (P-E, in mm/day) (e, f) in the EMICs as a function of latitude for DJF (a, c, e) and JJA (b, d, f).”

There is also an examination of the trends in many of these climate features. See the paper.

The paper’s conclusion begins by describing what the authors believe to be the climate models qualitative and quantitative successes and weaknessed in relation to the full scale AGCMs. It ends as follows:

We thus conclude that EMICs could be successfully employed as a useful and highly efficient, in terms of the running time, tool for the assessment of the long-term surface air temperature, precipitation and sea level changes, under a variety of future and past climate scenarios, as well as for testing and validating different concepts and parameterisation schemes for the individual climate mechanisms and feedbacks. However, a noticeable dispersion is detected in EMIC results as to the latitudinal response to transient and equilibrium CO2 doubling of some important climate characteristics, e.g. the zonally averaged planetary albedo, outgoing longwave radiation, and P-E. A pronounced quantitative discrepancy is revealed in simulating the CO2-driven changes in the maximum Atlantic overtuning streamfunction and the sea ice areas in the NH and SH. This indicates that the sensitivity of the specific climate processes (e.g. the global hydrological cycle) and feedbacks (e.g. cloud and snow/sea ice surface albedo feedbacks), as well as of their relative strength, to the external (e.g. anthropogenic) forcing might differ widely in EMICs. This necessitates further intercomparison of EMICs, in particular, with respect to the efficiency of climate feedbacks operating in these climate models. A serious problem for the majority of EMICs remains the lack of adequate simulation of interannual and interdecadal climate variability. In this context, the implementation of the nonstationary partial differential equations for treating the atmosphere synoptic-scale eddy/wave ensemble fluxes and variances is of specific interest.

Climate Dynamics (2005) 25: 363–385
DOI 10.1007/s00382-005-0042-3