bathymetry contour lines

lamont · Post by **lamont** » Sun Feb 24, 2008 5:23 pm

experimenting with generating contour lines --

http://www.scriptkiddie.org/bath/contour.html

lines are at 33 foot intervals:

33 feet
66 feet
99 feet
132 feet
165 feet

Joshua Smith · Post by **Joshua Smith** » Sun Feb 24, 2008 5:26 pm

Pretty neat- maybe a bit easier to read than the colored bathymetry one, too.

CaptnJack · Post by **CaptnJack** » Sun Feb 24, 2008 7:02 pm

Very nice. Any chance you could color code the lines? (I'm guessing labels would be a bigger pain to scale up and down)

ppharman · Post by **ppharman** » Sun Feb 24, 2008 7:04 pm

Why not just get a nautical chart. There is large scale charts (small area) for all of Puget Sound. You can even get the original surveys for little or no cost from NOAA. Why guess/interpolates a small scale (large area) road map.

I may be missing something here, sorry if I am off base.

Pat Harman

CaptnJack · Post by **CaptnJack** » Sun Feb 24, 2008 7:08 pm

Lamont likes converting digital data in graphics. The contour lines and/or color version is actually much finer resolution that the NOAA charts. Its based on a 1m DTM that's available from the UW. NOAA charts aren't nearly that high a resolution and the interpolation is just winged in by a cartographer (which sometimes is the best way). But digital interpolation is interesting and sometimes shows stuff that NOAA does not.

BDub · Post by **BDub** » Sun Feb 24, 2008 7:17 pm

Good job, Lamont. Very helpful...

airsix · Post by **airsix** » Sun Feb 24, 2008 8:02 pm

Hi Lamont,
I think my brain processes the color gradient better than the contour lines. Great work you're doing, by the way. I spend quite a bit of my idle time at /bath/index.html

-Ben

lamont · Post by **lamont** » Sun Feb 24, 2008 9:24 pm

ppharman wrote:Why not just get a nautical chart. There is large scale charts (small area) for all of Puget Sound. You can even get the original surveys for little or no cost from NOAA. Why guess/interpolates a small scale (large area) road map.

I may be missing something here, sorry if I am off base.

Pat Harman

the DEMs that i'm drawing that information off of already incorporate the raw data from the NOAA surveys which went into the nautical charts. i've done quite a lot of research into the raw surveys themselves, but its difficulty to pull information out of them which can be used as overlays on top of google earth -- for example there's a higher resolution survey around the vertical barges investigating them for the AWOIS database and the production of nautical charts but the data isn't gridded, and its hard to find non-commercial software to resample it.

nautical charts are also geared towards 'soundings' and for boat traffic that just cares about not running aground. i'm more interested in things like the 100 foot contour running around puget sound.

lamont · Post by **lamont** » Sun Feb 24, 2008 9:26 pm

CaptnJack wrote:Lamont likes converting digital data in graphics. The contour lines and/or color version is actually much finer resolution that the NOAA charts. Its based on a 1m DTM that's available from the UW. NOAA charts aren't nearly that high a resolution and the interpolation is just winged in by a cartographer (which sometimes is the best way). But digital interpolation is interesting and sometimes shows stuff that NOAA does not.

actually most of it is a 10m DEM (30 ft x 30 ft pixels), the 3m and 4m stuff is around elliott bay. i'd *love* to get 1m resoultion bathymetric DEMs, but none of the data out there is quite that good...

smike · Post by **smike** » Mon Feb 25, 2008 10:46 am

Lamont,

How are you getting the information from the UW soundings into another program? I noticed the info is in 10m squares, and was wondering how I was going to convert that to global position.

Do you have the data in a google earth format? I would like it in sketchup, and Google earth is easily imported to Sketchup. I WAS planning on writing a ruby script for converting directly to Sketchup, and maybe I will still do that.
I've looked, and the best data that seems to be available is the UW data set.

BTW - good work. I don't have the time to work on this for a couple of weeks, but I am very interested in what you are doing. NEAT!

S'Mike

lamont · Post by **lamont** » Mon Feb 25, 2008 2:15 pm

smike wrote:Lamont,

How are you getting the information from the UW soundings into another program? I noticed the info is in 10m squares, and was wondering how I was going to convert that to global position.

Do you have the data in a google earth format? I would like it in sketchup, and Google earth is easily imported to Sketchup. I WAS planning on writing a ruby script for converting directly to Sketchup, and maybe I will still do that.
I've looked, and the best data that seems to be available is the UW data set.

BTW - good work. I don't have the time to work on this for a couple of weeks, but I am very interested in what you are doing. NEAT!

S'Mike

i'm doing kind of crazy stuff. i've got a little perl script which strips out the land and then converts the PSDEM co-ordinate system to a wgs84 (lat+long) georeferenced tiff (using primarily, gdalwarp). then the bulk of the work begins which is to use gdalwarp to cut out tiles of 256x256 pixels with the appropriate bounding boxes which i then covert to PNGs to create a tiling set which google maps can display directly.

i do this individually for psdem 2000, psdem 2005, and the nisqually, duwamish and puyallup delta DEMs which i then do another pass over to generate a single tiling set which is the result of stacking those individual tilings.

there's other ways to generate the maps, but pre-generating the tilings is the fastest way to display them (i was using UNC mapserver to display them on-the-fly, but nothing beats static content serving).

...

so here, i've got two perl scripts for stripping land (they take ASC in and produce ASC out):

http://www.scriptkiddie.org/bath/script ... ip_land.pl
http://www.scriptkiddie.org/bath/script ... ip_land.pl

then this Makefile has most of the rest of the pre-processing before the tiling:

http://www.scriptkiddie.org/bath/scripts/Makefile

i embed the decimeters-to-meters conversion on psdem2000 in my perl script that also strips out the land so the magic conversion to get a WGS84 geotiff is:

gdalwarp -s_srs "+proj=utm +zone=10 +datum=NAD27 +units=m" -t_srs WGS84 -of GTiff -srcnodata 0 -dstnodata 0 psdem_2000_noland.asc psdem_2000_noland_wgs84.tif

and for psdem2005:

gdalwarp -s_srs "+init=nad83:4601 +units=ft" -t_srs WGS84 -of GTiff -srcnodata 0 -dstnodata 0 psdem_2005_noland.asc psdem_2005_noland_wgs84.tif

the program which does the tiling is basically 200 lines of uncommented spaghetti perl:

http://www.scriptkiddie.org/bath/scripts/tiling.pl