We used the original Smith and Sandwell (1997) data to interpolate a 3×3 km², Albers Alaska projected grid of the study area. The Smith and Sandwell data are a global coverage, in geographic coordinates, delivered as an ASCII XYZ file. Once the data were projected, we found that we needed to shift the points 3 km N, and 2 km W to align them with a correctly registered 1:63,360 scale State of Alaska coastline file (MMS 2002). The commonly distributed Smith and Sandwell data are shifted as a result of being moved from a point to a raster representation (Marks and Smith 2006). All on land points were removed prior to generating the bathymetric coverage. This included both points whose depth was >= 0, and those that occurred on land based on the Alaska coastline file).
Bathymetry for the 9×9 km² Bering Sea / Gulf of Alaska study area.
We used the ArcGIS TopoRaster function to interpolate the 3×3 km² bathymetric grid for the North Pacific. Given the extents of this region, the process was completed using a mosaic of 4 rasters. The raster was then clipped to the specific regions of interest, using predefined rectangular coordinates.
For the 9×9 km² GOA/BS study area, the data were extracted from the 3×3 km² North Pacific raster and spatially aggregated onto the coarser, 9×9 km² study grid.
A number of revised bathymetric products are now available including an updated global bathymetry fromETOPO2 product. The differences between these data sets are well described by Marks and Smith (2006).
Slope was calculated separately for each of the study area bathymetries using the IDRISI SURFACE slope function, with a scaling factor of 1 (because both the vertical and horizontal units were metres) and units of degrees. Maximum slopes of 19.4° was obtained for the 3×3 km² scale, and 15.5° for the 9×9 km² scale.
The bathymetry and slope files are provided as binary float files (*.flt & *.hdr pair) for each study area.
Marks, K.M., and W.H.F. Smith. 2006. An evaluation of publicly available global bathymetric grids. Marine Geophysical Researches, 27:19-34. DOI 10.1007/s11001-005-2095-4.