Challenges in modeling disturbances' effects on terrestrial carbon cycling

Ecosystem disturbance presents modelers with a number of problems, of which the disturbance event itself may be the least problematic. First, most ecosystem models depend on reaching a steady state before the main computation begins; this already-tricky initial value problem is complicated by disturbances during spinup, which may prevent steady state from being reached. Second, vegetation dynamics becomes more important in disturbed ecosystems, with nutrient cycling, albedo, and evapotranspiration all changing markedly. Succession is handled crudely (or not at all) by many large-scale models, however. Third, disturbances tend to export the carbon--and its climate forcing potential--from the system under study, complicating and expanding the scope of the problem at hand, whether the modeler likes it or not. And finally, humans tend to be intimately involved in disturbances in many ecosystems in some surprising ways. This talk looks at each of these factors in turn, with reference to specific modeling examples or problems, and considers their consequences for the modeling of the North American carbon cycle and the increasing disturbances that affect it.