Spatial Probabilistic Assessment Of Near-Surface Permafrost Parameters For The North Slope Of Alaska

Although the ground thermal regime is controlled primarily by climate, details are affected significantly by ecosystem factors, which are responsible for the large spatial heterogeneity in the active layer and near surface permafrost parameters. This variability presents a significant challenge in representing near-surface permafrost on small-scale (large area) maps. We have applied a stochastic approach, in combination with an equilibrium permafrost model, to map the spatial distribution of active-layer thickness and near-surface permafrost temperature on the North Slope of Alaska. A Monte Carlo technique was used to simulate the spatial variability of ecosystem factors controlling the ground thermal regime within each grid cell of the permafrost model. The required edaphic input parameters of the model (vegetation and soil properties) were described by their probability density function, derived from analysis of major landscapes within each grid box. To incorporate the uncertainty associated with the climate data, the model was forced by four sets of gridded air temperature and precipitation fields at 625 km2 resolution, which is used widely for spatial modeling applications. The entire model runs a large number of times with different combinations of the input parameters. As a result of a large series of calculations, a whole spectrum of possible solutions is obtained and statistical estimates of the expected value and dispersion, as well as a distribution of probabilities, is estimated for each grid cell. The results are presented as a series of probability maps of active-layer thickness and near-surface permafrost temperature for the North Slope of Alaska. Aside from portraying the level of uncertainty, these maps depict the probability of active layer and near-surface permafrost temperature exceeding specified thresholds.