Data for coastline refinements

Hi there,

the German wiki says that the coastlines generated automatically from PGS data are not very precise and need manual refinements. It recommends to use satellite images or gps tracking for that purpose.

I corrected the coastline for El Hierro (Spain) according to yahoo satellite images where it is visible, but large portions are cloud covered. Tracking the coastline is not much of an option in that area either.

However I observed the the 0m altitude level of the data generated from SRTM data (as displayed with cyclemap rendering) seems to follow the coastline much more precisely. Would it be legitime to use this data to refine the coastline? If so, it would probably a) get the raw data + b) convert it into gpx to get it shown in josm, right?

I think this is a good idea. You should check if it’s permitted to do so. Don’t know
if the STRM Data is public domain or another licence which allows them to be used
for this purpose.
Chris

Yes, both the SRTM data available from http://dds.cr.usgs.gov/srtm/ and the highresolution coastline data distributed with Generic Mapping Tools (http://gmt.soest.hawaii.edu/) seem to be public domain according to their documentation.

However I got stuck again trying to get the data into a usable form.

So far, I have after a lot of unsuccessful tries to handle the SRTM data (which seems to be overkill for my purposed anyway as it contains a lot more information than just the coastline which I am trying to extract), I have recently managed to get the GMT coastline data into human readable format.

My problem now: It is not a list of x, y (and possible z) coordinates that I need to generate somehow, rather some array structure that yields the information.

I can probably recalculate absolute latitude and longitude values from the relative coordinates given there once I fully understand the array structure, however, some guidance from anyone more experience would be highly appreciated.

I’ll add some sample data in a second post. This is political borders in low precision, not the coastline. This is because smaller file size only.

(deleted example data as this turned out to be a dead end)

Ok, I think I figured it out (mostly at least). Let’s do some math and see what I end up with.

After a lot of messing around with GSHHS coastline data trying to extract coordinates from it, I discovered that a) there is a tool for this purpose but b) the precision of the data was less than I had hoped.

So, it returned to my initial idea to investigate the SRTM data. Looking for a tool earlier in the process this time I found srtm2osm and managed to get the 5 m contour line which seems sufficient for redrawing the coastline at least if the coast is so cliffy (a lot better than the coastline in the Northern part of the island looked before anyway).

In case someone wants to use SRTM data for similar purposes in similar situation (cliffy coasts, no useful satellite images available), here is a brief description of what I did:

  1. download srtm2osm.exe for windows (there are instructions for linux as well)
  2. get the border coordinates of the area, e.g. from the coordinates tab in JOSM’s download window
  3. run strm2osm on the command line with option -bounds1 and the border coordinates and option -step 5
  4. edit the resulting osm file to get rid of 1. all ways with elevation > 5 + the corresponding nodes (a bit of fiddly thing, but my first idea to deleted these items within JOSM turned out not to be advisable)
  5. open the osm file in JOSM and export it as .gpx
  6. use this “pseudo-gpx-track” to redraw the coastline where necessary