Anomalous Atmospheric Absorption? Or Aerosols.
Climate models — the computer simulations used to study problems such as global warming — seem to be under perpetual attack. Scientifically, this is healthy as it leads to continued testing, verification and improvement of the models. The prominent issue in the 1990s has been the "missing atmospheric absorption" — a claim that the models underestimate sunlight absorbed in the atmosphere by 20-40 W/m2 and overestimate solar radiation absorbed at the planetary surface by a similar amount. Such errors could generate inaccurate atmospheric and oceanic circulations in the models. Thus, this issue has set off a search for possible exotic mechanisms of absorption that are not included in present models.
But how is atmospheric absorption, A, observed? Actually, it's not. A is a tertiary quantity obtained via a round-about calculation. Instead, the primary quantity observed, at several hundred stations around the world, is the solar radiation incident on Earth's surface, S. The energy absorbed by Earth's surface, an important quantity in climate models, is a secondary quantity in the sense that it must be obtained by multiplying S by the surface co-albedo (i.e., 1 minus the albedo). Unfortunately, the surface albedo varies on small scales and is not well measured globally, so that the uncertainty in energy absorbed at the ground includes the errors in both S and surface albedo. The atmospheric absorption is a tertiary quantity because its calculation requires still additional data, such as Earth's top-of-the-atmosphere albedo, that adds further error.
A comparison of climate model with observations for a primary variable is shown in Figure 1. The model is the GISS SI99 climate model. Observations are the GEBA (Global Energy Balance Archive) data for solar radiation at Earth's surface measured at 700 stations world-wide. The model and observations are in good agreement, within a few Watts per square meter, which is less than the observational uncertainty. How can this result be reconciled with the common perception that models underestimate atmospheric absorption?
Much of the answer is that the current GISS model includes a substantial amount of atmospheric aerosols, i.e., fine airborne particles, including sulfates, organics, black carbon, soil dust, and sea salt. It is also important that models accurately calculate water vapor absorption, but that seems to be easier than accurate specification of atmospheric aerosols. Figure 2 summarizes the properties of aerosols used in the GISS model, specifically their total optical depth and single-scattering albedo. The most notable feature of the aerosols is that they absorb substantially, with single-scattering albedo typically between 0.9 and 0.95. Much of this absorption comes from the black carbon aerosols, a significant amount comes from soil dust, and a very small amount comes from organic aerosols.
The issue is whether the aerosol properties are accurate. Most of the aerosol distributions are obtained from aerosol transport models, with little if any verification against local data or satellite measurements. Qualitatively, the low single-scattering albedos are consistent with field measurements near the East Coast of the United States and in the Indian Ocean, but much more comprehensive data are needed to develop aerosol distributions for all constituents. The real problem is that there are many aerosol types, which are distributed heterogeneously and vary in time.
Measurements from space are probably the only way to obtain the needed quantitative global data on all these aerosols. NASA plans for satellite measurements within the next few years include a focus on aerosols. Stay tuned. If all goes well, we should be able to determine whether there really is any anomalous atmospheric absorption or whether the "anomaly" is just plain old aerosols.
Hansen, J., R. Ruedy, A. Lacis, M. Sato, L. Nazarenko, N. Tausnev, I. Tegen and D. Koch 2000. Climate modeling in the global warming debate. In Climate Modeling: Past, Present and Future (D. Randall, Ed.), pp. 127-164. Academic Press, San Diego.
Ohmura, A., et al. 1998. Baseline Surface Radiation Network. Bull. Amer. Meteorol. Soc. 79, 2115-2136.