This project resulted in an extensive mapping of coastal change along the entire coastline of the Western Alaska Landscape Conservation Cooperative (LCC). The work provides important baseline information on the distribution and magnitude of landscape changes over the past 41 years. The extent of change to the coastline and to coastal features, such as spits, barrier islands, estuaries, tidal guts and lagoons, was known to be substantial in some areas along the coast (e.g., portions of the Yukon–Kuskokwim Delta), although the extent of change along the full Bering Sea coast was not well documented. With this analysis, changes can be summarized for different land ownerships or other units to assess the extent of recent habitat loss. Hotspots of change identified by the wide coverage of the mapping are available to provide guidance on selecting areas for monitoring or for more intensive research (e.g., terrain sensitivity analysis).
Annual and multiyear Landsat composites for the coastal zone of Western Alaska created as part of the project Extensive Mapping of Bering Sea and Gulf of Alaska coastal change by Landsat time-series analysis, 1972–2013. Individual image tiles (clipped to the study area) were reviewed to exclude bad data (clouds, shadows, surf, sensor noise). The available image tiles for each year were then composited (combined) using the data closest to a midsummer target date, typically mid- to late-July. Annual composite tiles at 60 m resolution were produced for the entire time-series (1972–2013), and included red, green, and near-infrared bands from all sensors with the TM, ETM+ and OLI data aggregated to 60-m resolution. A fourth band with the day-of-year of the source imagery was included in each composite. Annual composite tiles at 30 m resolution were produced from all of the 30 m imagery (TM, ETM+ and OLI). These covered the period 1985–2013. Multiyear composites covering non-overlapping 3–6 year intervals were also produced. These were convenient to use when reviewing results of the time-series analysis results, because they had fairly complete coverage; each individual year had many gaps in spatial coverage. The multiyear composites were not used for analysis, only for visualization.
