CELL-BASED SIMULATION OF PEAT ACCUMULATION IN NORTHERN PEATLANDS

Northern peatlands, such as the Glacial Lake Agassiz Peatlands, contain nearly one third of the carbon stored in soils. Depending on the response of these peatlands to temperature and hydrologic change, they could function as either important atmospheric carbon sources or sinks. To evaluate peatland response to changes in climate, we have prepared a computer model that simulates peat accumulation using stochastic cellular automata within a rectangular lattice. In contrast, other research groups have simulated peat accumulation based on analytical and numerical solutions to differential equations. Starting with a basin of geologic sediment, 10 cm thick peat blocks are randomly added at the sediment-atmosphere interface based on a probability consistent with the average peat production rate of 100 cm per 88 years. Decay is managed in a similar manner with cells randomly transforming through ten different degrees of humification, becoming more recalcitrant and less likely to decay with increasing humification.

When cells of peat are deposited they have a decay state of 10 indicating that the cells have undergone no decay. When cells undergo decay, the decay state of that cell is reduced by 1. When a cell reaches a decay state of zero it is assigned a value of 0.001, a negligible decay value that will prevent it from being removed from the model, yet indicates that the cell has undergone significant decay. To simulate decreasing volume of peat due to decay, when a cell of decay state 10 to 5 or <1 is marked for decay, the decay state of that cell is reduced, and the attributes of that cell (decay state, age of deposition) are averaged with the cell above it. The two cells are then merged into one. Preliminary runs of the peat accumulation model produce peatlands with net accumulation rates averaging 1.0 x 10^-1 cm year^-1, which is consistent with previously reported accumulation values for the Glacial Lake Agassiz Peatland. This peat accumulation model will be linked to a hydrologic model to evaluate feedbacks between peat growth, decay, and groundwater flow processes.