Construction of Digital Bathymetry for the Gulf of Maine
Generating the Grids
This image below shows the combined data coverage from the 7 data sources:
The 15 and 30 second bathymetry grids were constructed using the freely-available Generic Mapping Tools (GMT) software (Wessel and Smith, 1991). The gridding technique consists of (1) pre-filtering the data with a block-median procedure to avoid aliasing small wavelengths, then solving a partial differential equation that combines Laplacian and bi-harmonic terms. See the paper by Smith and Wessel (1990) for details. Here are the GMT scripts we used:
#!/bin/csh
# Generate 15 second grid for the Gulf of Maine using GMT
zcat *.xyz.gz | \
blockmedian -R/-71.5/-63/39.5/46 -I0.25m -V > gom15.xyz
surface gom15.xyz -R/-71.5/-63/39.5/46/ -Ggom15.grd -C0.1 -A0.73 -I0.25m -T0.35 -V
#!/bin/csh
# Generate 30 second grid for the Gulf of Maine using GMT
zcat *.xyz.gz | \
blockmedian -R/-71.5/-63/39.5/46 -I0.5m -V > gom30.xyz
surface gom30.xyz -R/-71.5/-63/39.5/46/ -Ggom30.grd -C0.1 -A0.73 -I0.50m -T0.35 -V
Note about NAD83 and WGS84 horizontal datums; The horizontal differences between NAD83 and WGS84 were determined to be insignificant at the resolution of the 15 and 30 arc-second grids. With cell sizes of ~500 and 1000 meters any displacement of data points because of differences in horizontal datum would have no effect on the outcome of the grid cell value after pre-filtering through the GMT block median procedure.
Note about grid cell sizes: we produced the 15 and 30 arc-second grids because of varying source data resolution in the near-shore and offshore areas. The 15 arc-second grid has much higher source data resolution(NOS data) in the nearshore areas and could possibly be recompiled at a higher resolution in certain areas. The 30 arc second grid was produced for a more uniform fit of the data since the DBDB-V, and the USGS DEM contained data spaced at this resolution.
For comments and questions, contact Rich Signell
[an error occurred while processing this directive]