search

GLIMS-RGI / rgi_issue_tracker

Report issues about the RGI dataset
1 stars 0 forks source link

RGI 7.0: Glaciers in High Mountain Asia are improperly merged together #46

Open lucas-zeller opened 7 months ago

lucas-zeller commented 7 months ago

RGI version RGI v7.0

Describe the issue There are ~34 instances in regions 13, 14, and 15 of multiple glaciers being grouped together within a single RGI outline and given a single RGIId. These glaciers have multiple termini that drain into distinct basins.

Expected outcome The glacier outlines should be separated along their drainage divides and have unique RGI identifiers.

Screenshots image The blue outline in this example is a single glacier outline (RGI2000-v7.0-G-15-00060) from RGI v7.0. The orange polygons are 3 distinct glacier outlines in RGI v6 (RGI60-15.07374, RGI60-15.07439, RGI60-15.07375).

Additional context I found 34 instances of this occurring in regions 13, 14, and 15 by using the RGI6 to RGI7 link files to identify glaciers in RGI7 that were linked to multiple large (>1 km2) RGI6 glaciers, and then manually checking each of these. I checked the GAMDAM inventory and the same issues were found. The centerline products for these glaciers should also be recalculated after the geometries are properly split.

The full list of glaciers that I believe have this issue is: RGI2000-v7.0-G-15-00060 RGI2000-v7.0-G-15-04301 RGI2000-v7.0-G-15-11345 RGI2000-v7.0-G-15-11994 RGI2000-v7.0-G-15-14680 RGI2000-v7.0-G-15-14618

RGI2000-v7.0-G-14-32508 RGI2000-v7.0-G-14-33763 RGI2000-v7.0-G-14-17421 RGI2000-v7.0-G-14-12493 RGI2000-v7.0-G-14-12539 RGI2000-v7.0-G-14-24248 RGI2000-v7.0-G-14-08846 RGI2000-v7.0-G-14-05915

RGI2000-v7.0-G-13-39217 RGI2000-v7.0-G-13-24414 RGI2000-v7.0-G-13-24951 RGI2000-v7.0-G-13-42189 RGI2000-v7.0-G-13-43762 RGI2000-v7.0-G-13-48253 RGI2000-v7.0-G-13-47971 RGI2000-v7.0-G-13-47326 RGI2000-v7.0-G-13-65368 RGI2000-v7.0-G-13-64385 RGI2000-v7.0-G-13-63986 RGI2000-v7.0-G-13-70035 RGI2000-v7.0-G-13-69331 RGI2000-v7.0-G-13-73067 RGI2000-v7.0-G-13-74883 RGI2000-v7.0-G-13-51401 RGI2000-v7.0-G-13-54337

fmaussion commented 7 months ago

@lucas-zeller thanks for the report. This looks like a valid issue and is indeed very probably inherited from GAMDAM.

At this stage we won't be able to correct this for the RGI7 series and will be addressed in the RGI8 series...