These data are part of a long term monitoring effort. The entire data set is comprised of the following files: yearly inventory files from 1989 through 2004, excluding 1993 (kelpYYYY.shp), combined inventory from all years (kelp1989_2004.shp), and a geodatabase (map_index.mdb) containing spatial index (map_index.shp) and tabular data describing kelp characteristics for each year that link to the spatial index (kelp_mon.dbf).
Extreme care was used during the compilation of these data to ensure accuracy. However, the Department of Natural Resources cannot accept responsibility for errors or omissions, and, therefore, there are no warranties which accompany this material.
- Canopy mapping was completed on USGS 7.5 quadrangle maps that were enlarged to 1:12,000 on a calibrated photocopier. Film slides were projected onto the maps, and the surface canopy was delineated by hand on the basemaps from the center portion of the slide.
- Kelp canopy species composition was determined by combining visual observations from two low-altitude over-flights and quantitative data from the vertical imagery. In mixed beds, species composition was determined to the nearest 10% through species identification using a random dot grid.
- Bed planimeter maps were created by establishing a computer derived perimeter polygon around the kelp within each species class (25 meter distance threshold).
- Relative density index was produced by dividing the kelp canopy area by the planimeter area for each kelp bed.
- Canopy area and bed planimeter area were determined by scanning the maps.
The following paragraphs describe processing steps in more detail.
The 1989 kelp polygon coverages were projected from Washington State Plane South, NAD-27 to the final coordinate system. These data had a single attribute, dom.type, which contained the values: KELP, H2O, or Null. Species identification was not available for the 1989 data set, therefore, when dom.type was indicated as 'KELP', the species was designated as 'unspecified.' Polygons with dom.type = H2O or Null were internal holes in the kelp polygons, and were assigned a species name of 'none.' Because the data was one single coverage for the year, the overlap flag for all polygons was set to 'N' (no overlap), and the overlap species were set to 0 (none).
The 1990 through 1992 kelp polygon coverages were projected from Washington State Plane South, NAD-27 to the final coordinate system. These data had a single attribute, minor1, that was derived from the original DXF files. The final species codes were assigned based on 'read me' files provided by WDNR. As with the 1989 data, the data were provided as a single coverage for each year, so the overlap flags were set to 'N' (no overlap), and the overlap species were set to 0 (none).
In the data from 1990, 1991, and 1992, a group of polygons appeared to be incorrectly located in bed 25.1, east of their correct location in bed 25.2. This error was validated by the tabular data which lists no floating kelp in bed 25.1 for any year, as well as the similar shape of polygons for all other years in bed 25.2. All other floating kelp polygons for 1990, 1991, and 1992 appeared to be correctly located relative to kelp polygons for all other years. Most likely, there was a registration error on a single map sheet for each of these years. To correct this error, these polygons were selected and shifted 9904.375 feet westward and 466.5 feet northward to match the general location of the bed 25.2 polygons from all other years.
For 1994 through 2000, each annual kelp planimeter grid was converted to a polygon coverage using the default ArcInfo GRIDPOLY command. The mixed areas created through overlapping .img tiles were also converted to polygons and overlaid with the annual planimeter kelp polygon coverage. For these areas, the overlap flag was set (overlap = 'Y'), and the species codes for the two source species were indicated in the overlap_sp1 and overlap_sp2 attributes. Each annual coverage was then projected from the source UTM coordinate system to Washington State Plane South.
In the 1996 source images, the map registration tic marks in bed 25.1 were colored red in the source data, rather than magenta as in all other source image files. Therefore, these tics were converted and erroneously coded as Nereocystis polygons. The error was validated by the tabular data which indicates no floating kelp in bed 25.1 for 1996. These erroneous polygons were selected and removed from the 1996 annual kelp polygon data polygon.
For all years of data, kelp bed index numbers were assigned to all kelp planimeter areas by overlaying the kelp bed spatial index polygons with the kelp planimeter polygons using the ArcInfo IDENTITY command. The relative density index (RDI) attribute was assigned with a tabular join on the kelp bed index number.