Marine Science

this page is a work in progress

Tools for marine scientists

Bathymetry processing

• Basic DEM creation: v.in.ascii + v.surf.rst
• r.in.xyz for processing multibeam sonar swaths
• r.surf.nnbathy add-on script for creating bathymetic DEMs from input x,y,z data

• Multibeam sonar swath transect planning script

Bathymetric data

• See the ETOPO2 article by M.H. Bowman in the GRASS Newsletter, 1:8-11, August 2004.

• Smith and Sandwell 1-minute global elevation, v9.1, August 21, 2007

http://topex.ucsd.edu/marine_topo/mar_topo.html (712mb)

global_topo_1min/README_V9.1.txt file:

Version 9.1 has a very different FORMAT than V8.2
The main differences are that the grid spacing in 
longitude is now 1 minute rather than 2 minutes.
In addition, the latitude range is increased to 
+/- 80.738.  Like the old versions, the elevation(+)
and depth(-) are stored as 2-byte integers to the nearest meter.
An odd depth of say -2001m signifies that this pixel was constrained
by a real depth sounding while an even depth of say -2000m is
a predicted depth.

Here are the parameters for the old and new versions:
param    V8.2     V9.2
___________________________
nlon     10800    21600
nlat     12672    17280
rlt0   -72.006  -80.738
rltf    72.006   80.738
___________________________

The binary format of the integers is bigendian so the bytes need to be 
swapped if you are running on an Intel processor.
Here is a typical command for swapping bytes:
dd if=topo_9.1.img of=topo_9.1.img.swab bs=21600 conv=swab.

Import

Process with GMT's img2grd to convert from spherical Mercator projection to geographic coordinates, then import into GRASS

 img2grd topo_9.1b.img -T1 -S1 -V -R0/360/-80.738/80.738 -m1 -D -Gtopo_all.grd
 # (out of memory, needs 1.4gb)
 # try just for NZ   (W/E/S/N bounds)
 REGION=160/180/-50/-30
 img2grd topo_9.1b.img -T1 -S1 -V -R"$REGION" -m1 -D -Gtopo_NZ.grd
 grd2xyz topo_NZ.grd -S > topo_NZ.xyz

 # get adjusted region bounds and resolution from img2grd output
 # ** check that rows and columns match **
 g.region n=-29.9945810754 s=-50.0056468984 w=160E e=180 \
    ewres=0:01 nsres=0.0126094 -p

 r.in.xyz in=topo_NZ_180.xyz out=topo_NZ_1min fs=tab
 r.colors output=topo_NZ_1min color=etopo2

 # create a r.in.xyz "n" map to test input point coverage
 r.in.xyz in=topo_NZ.xyz out=topo_NZ_1min_n fs=tab method=n
 # check rast map stats, min=max=1 and there should be no null cells
 r.univar topo_NZ_1min_n
 # cleanup
 g.remove topo_NZ_1min_n

or, import GMT .grd file directly (introduces FP +0.005 elev error??)

 # convert COARDS-compliant netCDF grdfile to old GMT native .grd
 grdreformat topo_NZ.grd topo_NZ_old.grd=bf
 # import
 r.in.bin -hf in=topo_NZ_old.grd out=topo_NZ_old

To load it into GRASS lat/lon location (spherical):

Location setup:

http://thread.gmane.org/gmane.comp.gis.gmt.user/918

http://article.gmane.org/gmane.comp.gis.proj-4.devel/192/

Is it even possible to load directly into GRASS? How to set up Merc/Sphere proj?

 # the following does not work correctly, just a trial
 # offset n,s,e,w by 1/2 a grid cell?
 r.in.bin input=topo_9.1b.img output=topo_9.1b \
      title="1' worldwide relief (1.852 km-sq)" \
      -b -s bytes=2 rows=17280 cols=21600 \
      n=80.738 s=-80.738 w=0 e=360

  r.colors output=topo_9.1b color=etopo2

Sidescan sonar processing

• i.gdalwarp script for georectifying and mosaicking scanned paper rolls into a GeoTIFF

• v.swathwidth script for planning swath bathymetry surveys

Wave exposure

• Using GRASS to prepare and process data for the SWAN Wave Model
  • preparing input DEM
  • r.in.mat and r.out.mat

Circulation models

• Preparing input grids
  • r.in.mat and r.out.mat

Tutorials

Remote Sensing

• Importing MODIS Aqua SST and Chlorophyll-a data

Mapping and Cartography

• Using GMT with GRASS
• Using vector electronic navigation charts with GRASS
• Using rasterized navigation charts with GRASS