Framework for Describing Groundfish Habitat Information for the Gulf of Alaska and Aleutian Islands

Using only existing data, the NMFS and USGS are collaborating to provide broad and generalized seafloor habitat information in support of fisheries research in Alaskan waters.  Diverse sediment data, such as verbal descriptions and quantified grain size analyses, are combined into a unified database using fuzzy set theory (see http://walrus.wr.usgs.gov/usseabed/).  A central concept behind the database is that verbal descriptions of sediment samples can be translated into numerical granulometric values by placing them in a mathematical-type format which are then are brought into conformance with lab analyses using known or empirical relationships. These descriptive data can also be mined for constituent biological, geological, and feature components.  The Alaskan portion of this national dataset contains original data primarily from NOS charts, USGS surveys, and OCM studies.

Bathymetric readings, mostly from historical NOS charting surveys, are available from the National Geophysical Data Center (NGDC) as text files and as map images. In a GIS, the soundings are digitized and datum shifted as needed, edited, and processed into triangulated irregular networks (TINs) or grids of seafloor depth. Bathymetric derivatives such as slope (steepness) and rugosity (roughness) can be produced as separate surfaces that reveal abrupt changes in depth that are important in defining different fish habitats. 

The draping of smoothed, gridded, sediment surface over bathymetry provides a simple and effective way to show large areas that may contain various seafloor features such as rocky reefs, gravel piles, areas of flat and (or) soft sediment, banks, gullies, possible relict glacial moraines, and earthquake fault lines.  Additionally the flexibility of the usSEABED database allows the user to estimate various sediment parameters such as individual gravel, sand or mud surfaces, or geological or biological constituents.  Each seafloor location can be also described according to its unique characteristics, rather than as a generalized region. This facilitates the statistical modeling of relationships between fish and habitat so that biologically meaningful habitat boundaries can be derived.