How does rising atmospheric CO2 affect marine organisms?

Click to locate material archived on our website by topic


Near Surface Atmospheric Variables in CMIP5 Models

Paper Reviewed
Svensson, G. and Lindvall, J. 2015. Evaluation of near-surface variables and the vertical structure of the boundary layer in CMIP5 models. Journal of Climate 28: 5233-5253.

Introducing their recent publication in the Journal of Climate, Svensson and Lindvall (2015) described how they evaluated the diurnal cycles of near-surface atmospheric variables and turbulent heat fluxes in 16 CMIP5 models and compared the results with real-world observations made at a total of 26 flux tower sites, composed of 7 from mid-latitude grassland/cropland sites, 8 from mid-latitude forest sites, 2 from boreal forest sites, 4 from Arctic tundra/wetland sites and 5 from tropical forest sites, four of which sites were evaluated in more detail, "including the vertical structure of the atmospheric boundary layer at the Atmospheric Radiation Measurement (ARM) Southern Great Plains site in Oklahoma (USA). And what did this investigation reveal?

The two Swedish scientists report that the amplitude of the ensemble model median diurnal cycle of 2-m temperature was (1,2) "too weak over mid-latitude forest sites in summer and grassland in winter," but they add that (3,4) "a large intermodal spread and substantial biases" were found "at all locations and times."

Moving up in the world (literally), they additionally report that (5,6) "the diurnal cycle of 10-m wind speed is underestimated at mid-latitude grassland sites and slightly overestimated at mid-latitude forest sites," while (7) "some models even have diurnal cycles out of phase with the observations." And they note that all of the models have (8) "a too negative net radiation," as well as the problems of (9,10) "the sensible heat flux being too negative and latent heat flux too large and positive."

Focusing on the four models that were employed in the ARM study conducted on the Southern Great Plains of Oklahoma, Svensson and Lindvall state that (11) "all four models generally overestimate the boundary layer depth in all seasons," and that (12) this phenomenon "is most pronounced in summer." In addition, they report that (13) "the models show a substantial daytime mean warm bias in summer [of] up to 10°C." And last of all, they report that (14) "the modeled mean surface stress opposing the free flow wind is twice as large as the observed values."

Consequently, the host of model failures to adequately portray the real-world weather parameters the Swedish scientists investigated does not bode well for the current crop of climate models.

Posted 26 October 2015