The Influence of Topography and Vegetation on the Snow Cover in Tundra: Case Study from the Southern Spitsbergen Area
Kępski, Daniel Dawid
MetadataShow full item record
The aim of the study was to identify the snow cover distribution in the southern Spitsbergen tundra environment and to quantify its relationship with the topography and land cover. This was achieved by correlating vectorized cartographic materials, modeled climatic parameters and calculated topographic indices with snow cover properties measured in the field and obtained using remote sensing techniques. An important source of information used in the analyzes was time-lapse photography, which, thanks to the developed methodology and created tools, allowed to obtain snow cover extent data characterized by high temporal and spatial resolution. Snow cover distribution in the immediate vicinity of the Polish Polar Station obtained from time-lapse material was related to the results of satellite image classification from the 2014 ablation season. The field data was used to validate the SNOWPACK (predicting snow cover structure) and Alpine3D (predicting snow spatial distribution) models developed by SLFWSL. They were implemented for the first time in the tundra environment of Svalbard. The models were also used to simulate snow conditions in the vicinity of the Polish Polar Station at the end of the twenty-first century. For this purpose, data from climate projections of the Polar CORDEX initiative were adopted. The worst-case climate change scenario (RCP8.5) was assumed. The obtained results point to an increased snow deposition during the winter season in the western parts of the valleys, related to the snow redistribution by the dominant eastern wind. However, on a macroscale, the snow cover duration and depth increase eastward with growing distance to the open Greenland Sea. The snow cover extent in the ablation phase shows a relatively strong correlation with the modeled average annual air temperature (r = –0.78), precipitation total (r = 0.57) and, to a much lesser extent, with the potential insolation (r = –0.24). Topographic indices turned out to be important primarily on a local scale. In the Fuglebekken catchment, the strongest impact on the snow cover duration was found for the Terrain Ruggedness Index and Wind Exposition Index, especially for the NE wind. This proves the dominant influence of wind activity on the local snow cover distribution, which is less visible on a larger scale of the entire fiord. Significant relationships were found between the duration and thickness of the snow cover with the land cover. The longest snow cover persistence was in places devoid of vegetation. On the opposite side there was a plant formation, the tallest of the local vascular plants being the polar willow (Salix polaris). The observed difference in snow ablation time between these two land-cover formations, rock debris and lichen-herb-heath tundra, was approximately two weeks. On the other hand, a greater snow cover thickness during the winter season was found on wet moss tundra. This class is characterized by the highest value of the normalized differential vegetation index (NDVI), which is an indicator of biological productivity. This means complex relationhips between the snow cover and the vegetation, where the development of flora is hindered in accumulation places, with prolonged snow persistence. However, some plant species are well adapted to living under a thicker layer of snow. The presented results of snow cover modeling did not take into account its redistribution under the influence of wind. Consequently, the snow thickness in the Fuglebekken catchment area was overestimated by an average of 50%. This is the potential amount of snow removed from the tundra by wind activity and associated sublimation. Under changed climatic conditions, with an air temperature increase of about 6.5 °C at the end of the 21st century, the models show a significant reduction of the snow cover period in the tundra. In the projections for 2089- 2100, snowfall occurs only in November–May, and in the winter season thaws strong enough to melt completely the snow cover in the middle of winter take place. Additionally, climate projections indicate development of thick ice layers in the future tundra snowpack, which could cause severe environmental effects on both plants and animals.