Paper Reviewed
Trenberth, K.E., Zhang, Y. and Fasullo, J.T. 2015. Relationships among top-of-atmosphere radiation and atmospheric state variables in observations and CESM. Journal of Geophysical Research: Atmospheres 120: 10,074-10,090.

In a revealing study, Trenberth et al. (2015) conducted a detailed examination of the ability of the Community Earth System Model (CESM) to accurately describe the relationships that existed among top-of-the-atmosphere radiation, water vapor, temperature and precipitation over the period 2000-2014. And what did they thereby learn? Let us count the many ways in which they found that the projections of the model fell far far short of reality.

For starters, the three U.S. researchers report that (1) "the vertical structure of the CESM temperature profile tends to be top heavy in the model, with [2] too much deep convection and [3] not enough lower stratospheric cooling," that (4) "there is too much absorbed solar radiation over the Southern Oceans and [5] not enough in the tropics," which (6) "also biases meridional gradients," and (7) "alters ocean heat transports."

Simultaneously, they say that (8) the "El NiÑo-Southern Oscillation (ENSO) is too large in amplitude." And over the Warm Pool in the tropical western Pacific and Indian Oceans, they report that several related biases emerge, including the facts that in response to high sea surface temperature anomalies, there is (9-11) "more precipitation, water vapor, and cloud," as well as (12) "less absorbed solar radiation and [13] outgoing longwave radiation in the model than observed."

Trenberth et al. additionally note that (14) "of particular concern is the existence of a double ITCZ [intertropical convergence zone] bias." And they go on to add that (15) "the vertical structure of the temperature profile is shown to be top heavy in the model, with [16] too much deep convection and [17] not enough lower stratospheric cooling." In addition, they report that (18, 19) "both ASR [absorbed solar radiation] and OLR [outgoing longwave radiation] are smaller in the model than in the real world."

All things considered, therefore, this was clearly not a good showing for a highly-hyped model.