bution of drifting snow and subsequent melt pattern and timing (Liston et $al., 2002).
These forcings are not yet well understood and are the subject of active research (Cox et al, 2000; Betts, 2001; Friedlingstein et al., 2001). They will likely be associated with multiple types of climate responses and are not expected to be additive to the traditionally defined forcings. Complex interactions among these forcings make it difficult to determine their net climate effects (Eastman et al., 2001b; Narisma et al., 2003; Raddatz, 2003). Eastman et al. (2001b), for example, found that with doubled CO2 the grasslands of the central United States were more water efficient on an individual stoma level (biophysical forcing), but grew more biomass (biogeochemical forcing). The net effect was cooler daytime temperatures during the growing season.
There are no widely accepted metrics for quantifying regional nonradiative forcing. Indeed, because nonradiative forcings affect multiple climate variables, there is no single metric that can be applied to characterize all nonradiative forcings (Marland et al., 2003; Kabat et al., 2004). Nonradiative forcings generally have significant regional variation, making it important that any new metrics be able to characterize the regional structure in forcing and climate response—whether the response occurs in the region, in a distant region through teleconnections, or globally. As is the case for regional radiative forcing, further work is needed to quantify links between regional nonradiative forcing and climate response. Another consideration in devising metrics for nonradiative forcings is enabling direct comparison with radiative forcings, computed in units of watts per square meter. However, not all nonradiative forcings are easily quantified in these units.
A metric that could prove useful for quantifying impacts on the hydrological cycle is changes in surface sensible and latent turbulent heat fluxes. For example, Pielke et al. (2002) proposed the surface regional climate change potential (RCCP), which is calculated by summing and weighting globally the absolute values of changes in the surface sensible and latent turbulent heat fluxes. In their study, land-use change from the natural to the current global landscape produced a global average RCCP of 0.7 W m−2 when teleconnection effects were not included, and 8.9-9.5 W m−2 when teleconnections were included. Such a scaling of the land surface forcing provides a metric that can be expressed in the same units as radiative forcing. Extending this concept to the global water cycle, Pielke and Chase (2003) quantified landscape forcing in terms of precipitation and moisture flux changes. They found globally averaged differences between the current and the natural landscape of 1.2 mm day−1 for precipitation and 0.6 mm day−1 for moisture flux. However, such metrics do not provide a