NASA NTRS1997
Simulation modeling can play a number of important roles in large-scale ecosystem studies, including synthesis of patterns and changes in carbon and nutrient cycling dynamics, scaling up to regional estimates, and formulation of testable hypotheses for process studies. Recent comparative studies have shown that ecosystem models of soil trace gas exchange with the atmosphere are evolving into several distinct simulation approaches. Different levels of detail exist among process models in the treatment of physical controls on ecosystem nutrient fluxes and organic substrate transformations leading to gas emissions. These differences are is in part from distinct objectives of scaling and extrapolation. Parameter requirements for initialization scalings, boundary conditions, and time-series driven therefore vary among ecosystem simulation models, such that the design of field experiments for integration with modeling should consider a consolidated series of measurements that will satisfy most of the various model requirements. For example, variables that provide information on soil moisture holding capacity, moisture retention characteristics, potential evapotranspiration and drainage rates, and rooting depth appear to be of the first order in model evaluation trials for tropical moist forest ecosystems. The amount and nutrient content of labile organic matter in the soil, based on accurate plant production estimates, are also key parameters that determine emission model response. Based on comparative model results, it is possible to construct a preliminary evaluation matrix along categories of key diagnostic parameters and temporal domains. Nevertheless, as large-scale studied are planned, it is notable that few existing models age designed to simulate transient states of ecosystem change, a feature which will be essential for assessment of anthropogenic disturbance on regional gas budgets, and effects of long-term climate variability on biosphere-atmosphere exchange.