greenhouse gases. This offset could even be significant on the global scale (e.g., Anderson et al., 2003a).
The degree of spatial heterogeneity can be seen by considering the aerosol optical depth for a number of aerosol species as shown by model results in Figure 4-1. The large optical depths off of the Sahara are due to mineral dust, while the large optical depths in South America and Africa are related to biomass burning. Large aerosol optical depths due to sulfur emissions occur in Northern Hemisphere industrial regions. These optical depths can be used in conjunction with assumptions about aerosol radiative properties to calculate the direct forcing. Results are shown in Figure 4-2 for TOA, surface, and atmospheric radiative forcings. Note the significant difference between the TOA forcing and the effect at the surface due to the absorptive properties of the aerosols. Regional forcing values at the surface can be as large as −20 to −30 W m−2.
The consequences of regional radiative forcing on the climate system for a region, for other regions, and globally must be better understood. Because global forcings can also have regionally specific responses, it is difficult to attribute regional climate changes to a particular regional or global forcing. A further complication is that regional diabatic heating results in nonlinear long-distance communication of convergence and divergence fields, often referred to as teleconnections. For example, Chase et al. (2000a) found that regional land-use change can cause significant climate effects in other regions through teleconnections, even with a near-zero change in global averaged radiative flux. Chen and Ramaswamy (1996) and Ramaswamy and Chen (1997) showed that significant responses in precipitation patterns can arise in the presence of a near-zero global change in radiative forcing. Regional radiative forcing may provide a better measure of regional climate response than global radiative forcing, but further work is needed to quantify the links of regional radiative forcing to regional and global climate response.
Some forcings affect the climate system in nonradiative ways, in particular by modifying the hydrological cycle or vegetation dynamics. These nonradiative forcings generally have radiative impacts, but describing them only in terms of this radiative impact does not convey fully their influence on climate variables of societal relevance. For example, aerosol-induced changes in precipitation may have a small net effect on TOA radiative forcing, but could have significant impacts on the amount of rainfall a region receives, with consequences for agriculture, flood control, and municipal water supply. Furthermore, quantifying nonradiative forcings in