Towards a permafrost map of the Dry Central Andes: A comparison of mapping and modelling techniques for creeping mountain permafrost distribution in the Cordón del Plata, Cordillera Frontal, Mendoza, Argentina

TROMBOTTO, DARIO; BRENNING, ALEXANDER

Potsdam, Alemania

Congreso; 2nd European Conference On Permafrost; 2005

European Permafrost Association

Creeping permafrost as expressed by rock glaciers is of particular importance in the Central Andes due to its ice content and cryogenic indicator function. This study presents an inventory of the periglacial and glacial environment of the Cordón del Plata (Andes of Mendoza, Argentina; 32° 40´ - 33° 24´ S, 69° 45´ - 69° 12´ W), and compares traditional mapping with statistical modelling techniques at a regional scale. The periglacial environment of the study area comprises 1658 km2 between the lower limit of 3450 m a.s.l. and the highest summit elevation of 6310 m a.s.l. at Cerro El Plata. This area includes the creeping permafrost with over 380 principal bodies of approximately 119 km2. Creeping and high-elevation „quasi-continuous“ permafrost are distinguished since these types adapt best to regional scales and visualization, and since technical limitations make a discrimination of quasi-continuous and other types of high-mountain permafrost very difficult. Degraded permafrost may be identified through the activation and enlargement of old thermokarst forms. Permafrost with very poor ice-content is also present in the study area. Rock glaciers, thermokarst and glaciers were mapped from Landsat imagery, and regional permafrost distribution inferred from direct and indirect field observations including the use of geomorphological and geophysical methods. Furthermore, a predictive statistical model of rock glacier and glacier distribution that has been fitted to the entire Andes of Mendoza and Santiago (Argentina/Chile) was applied to the study area and is compared to the mapping results. The model reproduces very well the „rock glacier pattern“ found in the Cordón del Plata, although this area is just a marginal part of the training area of the model. A comparison with the regional glacier inventory of 1981 also shows that predictive statistical modelling can efficiently produce satisfactory results in the assessment of permafrost distribution.