Mountain permafrost: transient spatial modelling, model verification and the use of remote sensing

Abstract

Permafrost research has a practical relevance for the construction and maintenance of infrastructure as well as for the assessment or prevention of natural hazards in cold-mountain areas. The occurrence and temperature of permafrost are largely controlled by climatic conditions and, therefore, atmospheric warming leads to corresponding warming or thawing of permafrost in most cases. Because it is invisible at the ground surface, the delineation and thermal characterization of permafrost largely depends on numerical models. The design and validation of such models, however, is complicated by the high spatial variability of surface properties and conditions that are characteristic of mountain areas. Three main challenges can be identified for the development of models simulating transient ground temperatures and mountain permafrost: a) to investigate and to model quantitatively, aiming for the transient simulation of three-dimensional temperature fields; b) to investigate surface types such as bedrock or moraines and to expand knowledge on permafrost beyond coarse blocky surfaces; and c) to provide the quantitative spatial input data required by such models. Several steps in the direction of these challenges were taken and lead to interesting results. A surface energy-balance model partly developed in this dissertation successfully simulates time series of rock surface temperatures in rugged high-mountain terrain. This was validated using the results of a systematic one-year measurement campaign of rock temperatures in the Swiss Alps. Coupled with a one-dimensional ground heat-conduction scheme this validated model was then employed to investigate the evolution and distribution of Alpine rock temperatures in steep terrain using a 22-year forward model run. This combined approach of measurements and modelling was also successful in demonstrating the degradation of permafrost during the hot summer of 2003. A three-dimensional model of heat-transfer in rock forced with surface temperatures based on measurements and model experiments was used to investigate the effects of topography and variable surface conditions on the temperature distribution at depth. Using this approach, possibilities and caveats of the reconstruction of temperature histories from borehole temperaturedepth profiles could be elaborated. Straight-forward methods for the delineation of coarse blocky surfaces from aerial photography and airborne laser-scanning were developed and tested. Airborne hyperspectral data was recorded and used to fit BRDF (bidirectional reflectivity distribution function) models and to derive albedo in rugged terrain. Quantitative transient modelling of mountain permafrost will likely be of increasing importance for research and the assessment of natural or geotechnical hazards in the near future. The methods and results presented in this dissertation contribute to a development into this direction.