Derivation and analysis of a high-resolution estimate of global permafrost zonation
- Glaciology, Geomorphodynamics and Geochronology, Department of Geography, University of Zurich, Switzerland
Abstract. Permafrost underlies much of Earth's surface and interacts with climate, eco-systems and human systems. It is a complex phenomenon controlled by climate and (sub-) surface properties and reacts to change with variable delay. Heterogeneity and sparse data challenge the modeling of its spatial distribution. Currently, there is no data set to adequately inform global studies of permafrost. The available data set for the Northern Hemisphere is frequently used for model evaluation, but its quality and consistency are difficult to assess. Here, a global model of permafrost extent and dataset of permafrost zonation are presented and discussed, extending earlier studies by including the Southern Hemisphere, by consistent data and methods, by attention to uncertainty and scaling. Established relationships between air temperature and the occurrence of permafrost are re-formulated into a model that is parametrized using published estimates. It is run with a high-resolution (<1 km) global elevation data and air temperatures based on the NCAR-NCEP reanalysis and CRU TS 2.0. The resulting data provide more spatial detail and a consistent extrapolation to remote regions, while aggregated values resemble previous studies. The estimated uncertainties affect regional patterns and aggregate number, and provide interesting insight. The permafrost area, i.e. the actual surface area underlain by permafrost, north of 60° S is estimated to be 13–18 × 106 km2 or 9–14 % of the exposed land surface. The global permafrost area including Antarctic and sub-sea permafrost is estimated to be 16–21 × 106 km2. The global permafrost region, i.e. the exposed land surface below which some permafrost can be expected, is estimated to be 22 ± 3 × 106 km2. A large proportion of this exhibits considerable topography and spatially-discontinuous permafrost, underscoring the importance of attention to scaling issues and heterogeneity in large-area models.