Volume 9, Number 3: 18 January 2006
It is surely no secret that we have long been critical of climate models and what they suggest about earth's response to anthropogenic CO2 emissions, but that is only because we believe they do not yet adequately incorporate all factors of significance to the chief purpose for which they were created, i.e., the providing of realistic insight into the long-term future consequences of our past, current and projected activities; and in this regard we have primarily complained about their neglect of biological factors that may be key to achieving this ultimate objective. Now, in the recent publication of Friend and Kiang (2005), we have a good example of what can happen when biology is given a chance to show what it can do.
In the very first sentence of their insightful paper, Friend and Kiang state the general truth that "vegetation processes determine the partitioning of energy and moisture over much of the global land surface and therefore require realistic representation in models designed to simulate climate dynamics over decadal to century time scales." They then state the more specific truth that the "Goddard Institute for Space Studies (GISS) GCM contains state-of-the-art representations of many components of the climate system but has lacked a sophisticated representation of vegetation." Their ultimate purpose is to remedy the GISS GCM in this regard, and their current paper is described by them as "the first stage [our italics] in addressing this deficiency through the introduction of a physiology-based canopy stomatal conductance and photosynthesis scheme."
The older version of the GISS GCM's representation of canopy stomatal conductance, according to Friend and Kiang, "omits responses to atmospheric humidity and CO2" and "lacks mechanistic responses to light, leaf N [nitrogen], and temperature." The bulk of their paper thus describes how they rectified this situation by (1) developing a new stomatal conductance submodel that includes these responses and (2) testing the old and new submodels using flux measurements made over four different vegetation types. Then, after calibrating the new submodel to account for vegetation-specific physiology and extending the calibration to all vegetation types in the GISS GCM, they performed a global climate integration with the new submodel embedded in the GCM and compared its results with an integration performed with the old conductance scheme.
In reporting their findings, Friend and Kiang say "the new submodel results in surface cooling over many regions previously too warm," noting that "some warm biases of over 2°C are cooled by more than 0.5°C, including over central Eurasia, South America, the western United States, and Australia." In addition they report that "some regions that were previously too cool are warmed, such as northern Eurasia and the Tibetan Plateau." Likewise, they say that "a number of precipitation biases are also reduced, particularly over South America (by up to 1 mm day-1) and the oceanic ITCZs (by over ±1 mm day-1)," while "coastal west Africa becomes significantly wetter." They also say that "cloud cover increases over many land areas previously too clear," and they report that changes in high-latitude climate are brought about by "remote effects of increased tropical latent heating, resulting directly from improved characterization of tropical forest canopy conductance."
In concluding, Friend and Kiang say that "realistic representation of the stomatal control on land evaporation is critical for accurate simulation of atmospheric dynamics," and that "greater realism in canopy conductance behavior increases the value of the GISS GCM as a tool for assessing impacts of greenhouse gases on climate."
We agree with this assessment; but we also agree that this improvement is indeed but the first stage of what needs to be done before we can expect truly realistic results out of the GISS or any other GCM. Prominent in this regard is the need to determine and properly model the many "long and winding roads" that ultimately lead to the complex and multifaceted climatic impacts of biologically-produced aerosols of both aquatic and terrestrial origin, as well as the similar impacts of biologically-produced trace gases of the atmosphere, such as isoprene and methane, and how all of these systems and the numerous interactions and transformations that comprise them are impacted by both rising air temperatures and CO2 concentrations. Until these goals are achieved - and they will not be reached anytime soon - no climate model should be used as a basis for mandating anything about anthropogenic CO2 emissions. There are simply too many unknowns involved; and as the study of Friend and Kiang demonstrates, straightforward changes in the modeling of just one biological parameter (stomatal conductance) can produce substantial changes in state-of-the-art simulations of future climate.
Sherwood, Keith and Craig Idso
Reference
Friend, A.D. and Kiang, N.Y. 2005. Land surface model development for the GISS GCM: Effects of improved canopy physiology on simulated climate. Journal of Climate 18: 2883-2902.