How does rising atmospheric CO2 affect marine organisms?

Click to locate material archived on our website by topic

Seemingly Unresolvable Biases of CMIP3 and 5 Climate Models

Paper Reviewed

Zhang, X., Liu, H. and Zhang, M. 2015. Double ITCZ in Coupled Ocean-Atmosphere Models: From CMIP3 to CMIP5. Geophysical Research Letters 42: 8651-8659.

Introducing their work, Zhang et al. (2015) write that "coupled ocean-atmospheric climate models tend to simulate a double Intertropical Convergence Zone (ITCZ) that is characterized by two zonal bands of annual precipitation in the equatorial central Pacific that is not present in observations." And they further note that this problem "not only creates biases of latent heating in the tropics that can impact mid-latitude weather and climate through atmospheric teleconnections (Schneider et al., 2009; Manganello and Huang, 2009) but also affects the intensity of the Hadley circulation and the distribution of the trade winds that are directly related with the simulation of El Niño events."

In light of these disturbing facts, the three researchers decided "to examine the collective progress of climate models in reducing the double ITCZ from the time of Coupled Model Intercomparison Project Phase 3 (CMIP3) to Phase 5 (CMIP5). And what did they learn as a result of this undertaking?

For starters, Zhang et al. report that biases common to both CMIP3 and CMIP5 models include (1) a spurious precipitation maximum in the southeastern Pacific, (2) warmer sea surface temperatures, (3) weaker easterly and (4) stronger meridional wind divergences away from the equator relative to observations. They also write that (5) "there is virtually no improvement in all these measures from the CMIP3 ensemble to the CMIP5 ensemble models." And they say that (6) "no progress can be identified in the sub-ensembles of the five best models from CMIP3 to CMIP5 even though more models participated in CMIP5." In fact, they report that (7,8) "the biases of excessive precipitation and overestimated sea surface temperature in the southeastern Pacific are even worse in the CMIP5 models."

The lack of progress in these several areas is said by Zhang et al. to likely be due to (9) "the inadequate simulations of stratocumulus clouds in the southeastern Pacific" and (10) "the stratocumulus to cumulus transition away from the coast," as well as (11) "the triggering" and (12) "the entrainment parameterizations of deep convection," plus (13) "insufficient resolution of the models in resolving mesoscale eddy transport in the ocean," and (14) "upwelling along the coast."

So what's new in this exciting field of research? Apparently, it's the discovery of ever more things that the most up-to-date CMIP5 models fail to do correctly.

Manganello, J.V. and Huang, B. 2009. The influence of systematic errors in the Southeast Pacific on ENSO variability and prediction in a coupled GCM. Climate Dynamics 32: 1015-1034.

Schneider, E.K., Fennessy, M.J. and Kinter, J.L.I. 2009. A statistical-dynamical estimate of winter ENSO teleconnections in a future climate. Journal of Climate 22: 6624-6638.

Posted 25 March 2016