This page shows a Samish Basin Boundary drawn from A LiDAR DEM and the 2006 NAIP 18-inch resolution orthophotos. Suggestions are welcome. The basin is subject to change if more culverts,dikes and ditches are identified, and if the Nooksack gets feisty. If more of the Samish Bay watershed needs to be included, it must be modeled as one or more additional basins.

--------------------------------------------Comparison with other interpretations. Other boundaries need to be compared.----------------------------------------------Basin boundary over 2006 LiDAR, with 10-meter shaded relief.--------------------------------------------NAIP orthophotography--------------------------------------------Department of Health data-----------------------------------------This is the proposed mask for processing the basin. Cell size is 150 meters. The projection is UTM 10, NAD27.

The creation of the basin model was a painstaking process based on a LiDAR DEM gridded at six-foot postings, the 2006 NAIP orthophoto series at 18-inch pixel size, previously mapped rivers and basin boundaries, and mapped tide gates. The first step was to reproject the North Puget Sound LiDAR Survey from stateplane coordinates to UTM NAD27 at 2-meter resolution. The DEM was filled, flow directions were calculated, and a basin boundary was determined. The flows that cross the edges of each 150-meter cell were calculated, yielding an optimal 150-meter flow direction grid.

Normally, one would be satisfied to generalize a DEM from a 2-meter source to a 150-meter result, but low slopes and numerous anthropomorphic features made considerable work necessary. The process of filling sinks sometimes encounters large spurious sinks, beneath which flow direction information is lost. We therefore repeatedly clipped numerous subsets of the two-meter DEM, filling sinks, and calculating basin boundaries. When there were discrepancies between sources, we examined DEMs and photos at maximum resolution. After settling on the Basin boundary, we mapped out flowdirection arrows and modified flow directions where appropriate. shapefile of the boundary. Parts of the boundary which were automatically derived from the six-foot DEM have vertices at six-foot intervals, while freehand interpretations have fewer vertices.

-------------------------------------------- But wait, there's more. 21 additional basins contributing to Samish Bay were identified. They have not been converted to 150-meter cells, but the estimated number of cells is indicated on the figure. So, should the smaller basins be dropped? Have we gone to far up the coast? Not far enough?

Harvey Greenberg
Earth and Space Sciences
University of Washington
Mon Nov  1 11:37:10 PDT 2010