Catchment-Scale Permafrost Mapping using Spring Water Characteristics

This study presents a method for mapping the spatial distribution of mountain permafrost based on the chemical-physical characterisation of spring water in a 36km(2) high-elevation catchment in the Eastern Italian Alps. Water temperature, electrical conductivity and isotopic composition (H-2 and O-18) were measured in 45 springs during summer 2007, 2010 and 2012. Existing evidence of permafrost enabled the areas upslope of springs to be classified into two categories of permafrost occurrence (probable permafrost and no permafrost) and used to determine the most suitable tracer for permafrost mapping. Springs from probable permafrost areas have a specific water temperature signature. Spring water temperature was therefore used as a response variable in multiple linear regression, and mean elevation and mean clear sky radiation of spring upslope areas were used as predictors. The multiple regression models were statistically significant and used to map the potential spatial distribution of spring water temperature, which was reclassified into three permafrost categories (probable, possible and improbable). Cross-validation and independent validation by ground surface temperature data provided evidence that the spring water temperature can be used alone for easy and low-cost assessment of the catchment-scale permafrost distribution in similar alpine catchments. Copyright (c) 2015 John Wiley & Sons, Ltd.