Detection of Vehicle Tracks and Vegetation Damages Caused by use of Snowmobiles in the Longyearbyen Area on Svalbard using Unmanned Aircraft

The study area in the surroundings of Longyearbyen on Svalbard, Arctic Norway, located at 71.2°N and 16°E is characterized by dry Arctic climate with a snow season of more than eight months, annual precipitation of less than 200mm, and a mean July temperature of about 6°C. Longyearbyen is the main settlement on Svalbard, with about 2000 inhabitants. During the last two decades the number of snowmobiles have increased from a few hundred to a number almost equals the number of inhabitants, and snowmobile trips are today the one of the main leisure activities. In addition, thousands of tourist visits every spring, and many of these go on organized snowmobile trips. Due to the often thin snow cover, and use of snowmobile even during the spring snow melt in May and early June, the rapid growth in use of snowmobile has made some damage to the vegetation. Damage on the fragile vegetation caused by the skids and belts of the snowmobile can be observed in most parts of the Adventdalen valley, close to Longyearbyen. The main aim of this study is to explore the feasibility and accuracy of using data from Unmanned Aircraft Systems (UAS) to identify vehicle tracks and damages on vegetation caused by the use of snowmobiles. Use of UAS give the opportunity to carry out research in a manner that minimizes the environmental footprint of the research activities. Small unmanned aircraft, combining both fixed wing multi rotor types allow us to collect image data for vegetation mapping without having any personnel walking into the field disturbing the sensitive High Arctic ecosystems. UAS used here are inexpensive and simple to operate. They are being developed with the goal of providing airborne capabilities for scientists at an affordable cost. The aircraft were instrumented with a normal Canon Powershot S100 RGB compact camera and a modified Canon Powershot SX230 NDVI camera. The fixed wing aircraft was taking pictures from 100 meters altitude with ground resolution of 2.5 cm mapping 2-3 sq.km per flight. The multirotor helicopter were mapping areas of a few hundred square meters with ground resolution as high as 1 mm. An automated technique using HSV (Hue, Saturation and Value) was used instead of RGB color space to automatic detect tracks and quantify area affected. This may be used to monitor future changes and effect of regulatory actions. The 2.5 cm resolution data easily detected tracks on the flat valley floor. These areas have mixed vegetation of mires and dry areas. The dry areas have silty substrate, which is easily compressed by the skids and belts of the snow scooters. The vegetation in these areas is scattered, but rather species rich. Most common is the small Arctic Willow (Salix polaris), several bryophytes and graminoides, and the small shrubs White Arctic bell heather (Cassiope tetragona) and Mountain Avens (Dryas ocopetala). Among these species Mountain Avens seems to be most affected by the scooter activity. The mires seem to be less affected by the snow scooter activity. The slopes of the valley are dominated by Mountain Avens ridges, heaths, and spots with moss tundra. However, tracks were only detected on the ridges of the valley slopes.