Inventory of rock avalanches in western Glacier Bay National Park and Preserve, Alaska, 1984-2016: a baseline data set for evaluating the impact of climate change on avalanche magnitude, mobility, and frequency

The effects of climate change have the potential to impact slope stability. Negative impacts are expected to be greatest at high northerly latitudes where degradation of permafrost in rock and soil, debuttressing of slopes as a result of glacial retreat, and changes in ocean ice-cover are likely to increase the susceptibility of slopes to landslides. In the United States, the greatest increases in air temperature and precipitation are expected to occur in Alaska. In order to assess the impact that these environmental changes will have on landslide size (magnitude), mobility, and frequency, inventories of historical landslides are needed. These inventories provide baseline data that can be used to identify changes in historical and future landslide magnitude, mobility, and frequency.

This data release presents GIS and attribute data for an inventory of rock avalanches in a 4800 km 2 area of western Glacier Bay National Park and Preserve, Alaska. We created the inventory from 30 m resolution Landsat imagery acquired from June 1984 to September 2016. For each calendar year, we visually examined a minimum of one Landsat image obtained between the months of May and October. We examined a total of 104 Landsat images. The contrast between the spectral signatures of freshly exposed rock avalanche source areas and deposits and surrounding undisturbed snow and ice was typically significant enough to detect surficial changes. We identified and mapped rock avalanches by locating areas with 1) high contrast compared to surrounding snow and ice, 2) different spectral signatures between successive Landsat images, and 3) lobate forms typical of rock-avalanche deposits. Using these criteria, we mapped a total of 24 rock avalanches ranging in size from 0.1 to 22 km 2.

Attribute data for each rock avalanche includes: a date, or range in possible dates, of occurrence; the name of the Landsat image(s) used to identify and map the avalanche; the total area covered by the rock avalanche (including the source area and deposit); the maximum travel distance measured along a curvilinear centerline (L); and the change in elevation between the start and end points of the centerline (H). We also include a table containing a list of all the Landsat images examined. We acknowledge that our mapped polygons will be different, and less accurate, than polygons that could be mapped from higher-resolution satellite, aerial, and hand-held imagery. We specifically chose not to use high-resolution imagery because we desired a long-term historical inventory that was unbiased by changes in image resolution. Eventually, new mapping should be done to create an inventory that fully utilizes recently available high-resolution imagery.

Data included in this release form the basis of an interpretive paper available in the conference proceedings of the 3 rd North American Symposium on Landslides held in Roanoke, Virginia in June, 2017.