Quantification of Aerosol Solar Radiative Forcing and Its Impacts on the Biosphere

We used the derived global aerosol optical depths and GOCART model simulations of other parameters (courtesy of Mian Chin of GSFC) to determine aerosol direct solar perturbations at the top of atmosphere (TOA), surface, and in the atmosphere [Figure to the left]. The spectral variation and anisotropy of land surface reflection were characterized using the MODIS retrievals. Large surface cooling and substantial atmospheric heating coexist over land and adjacent oceans, due to absorption by biomass burning aerosol, pollution, and mineral dust. Globally, about a half of surface cooling is compensated by the atmospheric heating. The so-induced more stable lower atmosphere should affect atmospheric circulation and climate. The annual global clear-sky solar perturbation at the TOA by all aerosols is determined to be -4.5Wm^-2, without significant land-ocean contrast because of larger aerosol absorption and higher surface albedo over land.

Inclusion of anisotropic reflection of land surface is necessary to better characterize the diurnal variation of aerosol forcing, because aerosols change the fraction of direct beam and visible radiation and hence the effective reflection from the surface. Such secondary effects of aerosol-albedo interactions lead to a cooling of the earth-atmosphere system at high sun but a heating at low sun [Figure to the right]. Aerosol direct forcing is particularly sensitive to the surface albedo over bright surfaces (e.g., snow and desert) and the MODIS retrieval indicates that the albedo over the Saharan deserts and Arabian Peninsula is highly heterogeneous, e.g., differing by as large as 0.4. The MODIS albedo, with resolutions as fine as 0.05 degree, is being employed to examine the resultant heterogeneity of aerosol forcing in such regions. We are also exploring how to parameterize the effect of turbulence-generated relative humidity fluctuations on aerosol radiative forcing in GCMs.

We also coupled the CLM canopy radiative transfer scheme and Guenther’s isoprene emission model with a delta-four stream atmospheric solar transfer model to examine the impacts of aerosol on the photosynthetically active radiation (PAR) absorbed by sunlit leaves and shaded leaves. The MODIS retrieved leaf-area-index (LAI) and albedo were used. The PAR absorbed by sunlit leaves decreases while that by shaded leaves increases, with a net decrease of total PAR. Isoprene emissions depend on light intensity and leaf temperature. A separate treatment of sunlit PAR and shaded PAR shows that aerosols increase isoprene emission by as large as 10-20% in regions such as Amazon basin, tropical Africa, southeastern US, and southeastern China, because the emission rate increases nonlinearly with PAR and reaches the saturation at a PAR level of 1000 umol/m²/s. Such changes would have important implications for regulating the biosphere-climate-chemistry feedbacks. We are also using the developed aerosol data sets and MODIS LAI and albedo to calculate FPAR under realistic solar conditions (in terms of direct and diffuse fraction of PAR) and then to examine and probably correct the MODIS retrieved FPAR in which the PAR is assumed to be 100% direct beam.