Bug #925

GSHHS - NE Greenland

Added by John almost 4 years ago. Updated over 3 years ago.

Status:In ProgressStart date:2016-07-08
Priority:NormalDue date:
Assignee:Paul% Done:


Target version:-
Affected version:5.x-svn Platform:Mac OS X


Is the coastline issue in NE Greenland still in GSHHS? Danish GEUS coastline still differs considerably.

Thanks - John

Coastline_Comparison.pdf (414 KB) John, 2016-07-29 08:59

diff.pdf (374 KB) Andreas, 2016-07-30 01:45


#1 Updated by Paul almost 4 years ago

  • Status changed from New to Feedback

We fixed something in Greenland but that was some years ago. if the danes would share updated data then we will consider adding it. However, we are not on the prowl for improved data for GSHHG 2.

#2 Updated by John almost 4 years ago

Thanks, Paul. I meant to post this in the forum, and clicked new issue by mistake. Yikes!

I thought NE Greenland was fixed a few years ago, too, but I just found that it was still there. For a while, NSIDC had the Danish coastline, but I'll be darned if I can find it again.


#3 Updated by John almost 4 years ago

A little follow-up on this. Finally found the Danish coastline data set on NSIDC. Take a look in this directory:


There are 1:100k and 1:500k level coastline data sets available, there.

Attaching a map showing the discrepancy. GMT & gshhs coastline is pretty seriously wrong in NE Greenland.

#4 Updated by Paul almost 4 years ago

  • Status changed from Feedback to In Progress
  • Assignee set to Paul

Thanks John, your timing is impeccable. The GMT team will be having a developer summit at Scripps the week of Aug 15-19 and we will release GSHHG 3.5.6; should be able to address Greenland in that.

#5 Updated by Paul almost 4 years ago

Hm, having some trouble. Registered and downloaded the files but
ogr2ogr -f "GMT" mog500_geus_coastline_v1.1.gmt mog500_geus_coastline_v1.1.shp
ERROR 1: Features without geometry not supported by GMT writer.
ERROR 1: Unable to write feature 1 from layer mog500_geus_coastline_v1.1.
ERROR 1: Terminating translation prematurely after failed
translation of layer mog500_geus_coastline_v1.1 (use -skipfailures to skip errors)

How did you extract the data from this file?

#6 Updated by John almost 4 years ago

I had the same issue, but I used the suggestion: use -skipfailures to skip errors.

ogr2ogr -skipfailures -f "GMT" mog500_geus_coastline_v1.1.gmt mog500_geus_coastline_v1.1.shp

That will get around the mystery "feature 1" errors.

Resulting file will be in XY Polar Stereographic space (in meters). So I used a mapproject command:

mapproject infile -Js-45/90/70/1:1 -R-180/180/30/90 -Fe -C -I > outfile to get lon,lat pairs.

When I plotted it, it looked fine and complete. Have no idea why the shape file won't convert... John

#7 Updated by Paul almost 4 years ago

Got it, thanks.

#8 Updated by Andreas almost 4 years ago

There is some inconsistency in the coastline of Svalbard and Jan Mayen, Norway (see figure) which you might also want to address. A high resolution coastline set is available from the Norwegian Polar Institute. I've used the 1:100 000 set in my figure (highest resolution).

GSHHG is plotted with highest resolution (-Df).

Vector data, Svalbard
1:100 000 - https://data.npolar.no/dataset/645336c7-adfe-4d5a-978d-9426fe788ee3
1:250 000 - https://data.npolar.no/dataset/a23acc28-288b-49ba-ac6d-025d1fdee246
1:1 000 000 - https://data.npolar.no/dataset/63730e2e-b7a6-4d14-b341-c661ccdc5254

Vector data, Jan Mayen
1:100 000 - https://data.npolar.no/dataset/f16ab884-a66e-453e-be30-70f480da6691
1:1 000 000 - https://data.npolar.no/dataset/d4b774f2-77b3-4ac2-9f16-520905aecfe8

#9 Updated by John over 3 years ago

Now that the GMT developer summit is well past, I wondered whether any updates or decisions were made regarding this issue? Thanks! John

#10 Updated by Joaquim over 3 years ago

I'm afraid the new aren't going to be good. Paul will elaborate more but he passed some days working in this issue and gave up due to data madness things.

#11 Updated by John over 3 years ago

Copy that, Joaquim. NE Greenland is a part of the world that is under intense study. I would think that having corrected coastlines would be a high priority and that "data madness" issues can ultimately be overcome. Thanks for the reply. This, indeed, isn't good news...

#12 Updated by Paul over 3 years ago

Have now obtained the new good data. Now will be looking for time to do the work.

Also available in: Atom PDF