Brief Overview Of The Integrated Modeling Project

Fifteen investigators from seven institutions are cooperating in a USDA Forest Service funded project to assess forest responses to changing climate, air quality, and land use in thirteen southeastern states of the USA. We are developing information transfer between seven simulators ranging from canopy photosynthesis and soil nutrient dynamics to plantation management. Each selected result passed between two models has a mean and variance from the application of uncertainty analysis which propagates variability of soil and plant inputs through each model. These simulations are generated for combinations of atmospheric CO2, ozone, nitrogen deposition, temperature, and precipitation representative of the whole region. Selected results are stored in response surfaces as normalized values relative to results from calibration sites. The response surface files are used in assessment simulations. GIS databases are used to characterize plant, soil, and climate attributes for each of the 2.2 million square km pixels of the region. Cluster analysis aggregates these attributes, each as a mean and variance, into about 1000 clusters. Forest assessments are simulated with uncertainty analysis for all clusters with the Land Use Change and Analysis System (or LUCAS) model which accesses and interpolates response surface values stored at each node of a parallel processing computer network. Outputs, such as forest production, evapotranspiration, and carbon pools, may be compared statistically for alternative equilibrium or transient scenarios.