The Arctic Tundra Simulator (ARTUS) is a computer-based simulation model of Eriophorum vaginatum tussock tundra ecosystems found in north central Alaska. ARTUS simulates the annual patterns of heat and water balance, carbon fixation, plant growth, and nitrogen and phosphorus cycling. ARTUS runs in 1-d time steps for a growing season from 1 May to 17 September and is intended to run for several years. The abiotic section of ARTUS encodes the seasonal input of the environmental driving variables and calculates the resultant thermal and water regimes to define the heat and water environments for the tussock tundra system. The primary driving variables are daily total solar radiation, air temperature, precipitation, surface albedo, wind, and sky conditions. The soil compartment contains three organic horizons, which are recognized by their state of physical and chemical decomposition, and one mineral horizon. Six vascular plant species and four moss species are simulated. The model has seven compartments for each vascular plant species: total nonstructural carbohydrates, total nitrogen, total phosphorus, leaves grown in the current season, leaves grown in previous years, conducting and storage stems plus roots, and absorbing roots. In ARTUS the functional unit of the plant is the shoot system or ramet. Each shoot system consists of leaves, stems, fine roots (which do not have secondary growth and have a limited life-span), and larger roots, which have secondary growth and an extended life-span. Although plant processes are based on individual shoots, the ARTUS model as a whole is based on a square meter of ground. Values per square meter are calculated from the values per shoot by multiplying by the shoot density of each species.

The model was validated by comparing calculated and measured peak season biomasses and nutrient contents, and the seasonal progression of environmental processes, biomass, carbohydrate contents, and nutrient contents. ARTUS successfully simulated the seasonality of the physical environment, but simulated thaw depths were deeper than those measured at all sites. The simulated value for total vascular plant production was 77% of the measured value. The simulated values for ecosystem respiration for Eagle Creek were within the range of measured values. Simulations with ARTUS indicated different patterns of growth and different storage carbohydrate levels in deciduous shrubs, evergreen shrubs, and graminoids. The simulated seasonal course of net primary production of vascular plants and mosses was similar to the pattern measured at Eagle Creek.

Sensitivity analysis using ARTUS indicated that the tussock tundra is more sensitive to external environmental factors, such as increased temperature, than to internal ecosystem variables. The development of ARTUS was limited by the unavailability of data on whole plant carbon balance including root and stem respiration. More data are also needed on decomposition processes and nitrogen and phosphorus cycling. Adequate climatological data for northern Alaska are needed for extensive validations of the model. While caution should be used in basing managerial decisions on model simulations, ARTUS can be used to identify and quantify the magnitude and direction of plant responses to changes in state variables in the model.