RGI6 corrections (all regions)

[samherreid]

The link below contains the shapefiles I manually drew over satellite imagery to correct the entire RGI6. The resulting RGI update I derived for the article is mostly more recent than 2000, but I designed the methods to work backwards and forwards in time to easily accommodate automated updates. For each RGI region, I drew shapefiles that encompassed: 1. missing glacier area (debris-covered or bare ice): 2. false positive glacier area (bedrock mapped as glacier): 3. terminus change; and 4. cast shadows where bare glacier ice in shade would be falsely mapped as debris cover.

Included in the link are the files for 1. and 2. for each RGI region (3. and 4. are specific to the satellite imagery library I used and are specific to the timing of satellite image acquisition).

1. is somewhat time sensitive, but ice missed in RGI6 mapped in, say, 2016, was almost surly present in 2000. The shapes in 1. (subscript "addArea") are intended to be merged with RGI6 glacier geometry. If a similar addition/correction was already applied in RGI7 the merge would have no effect.

2. is in the form of lasso areas (subscript "error_debLasso"). Assuming that debris cover maps will be generated with RGI7, area mapped as debris that falls within the lasso areas should be excluded both from the debris map as well as the glacier geometry.

paper: https://www.nature.com/articles/s41561-020-0615-0 shapefiles: https://www.dropbox.com/s/890wwksng8g17jl/RGI_errorLocations_SH.7z?dl=0

[fmaussion]

Dear Sam,

thank you so much for this contribution! It is a herculean effort you have made, and is undoubtedly a dataset of huge value to the community.

Unfortunately, I don't exactly know what to do with this tremendous amount of information. RGI7 aims to be closest as possible to 2000. RGI7 will be a strict subset of GLIMS, the reference inventory for outlines. We cannot apply corrections to the GLIMS outlines after pulling them out of the database, for reproducibility and attribution reasons. Here I explained our workflow and here is a description of the current state of RGI. Note that many regions have improved greatly since RGI6, and I would be curious to know your opinion about whether some of the issues you found in RGI6 are corrected now.

For your outlines to be used in RGI, the first step would be to incorporate them into GLIMS first (there are some guidelines, I think that the most constraining will be that they need to know exactly which images have been used for the outline definition, and who the contributors are - in your case it would be you + whoever did the RGI6 outlines in the first place). The outlines also need to be consistent and comprehensive, i.e. they have to cover all ice bodies in a given region. Finally, your outlines will probably be flagged as of year 2016 (approx), which will make them candidates for the forthcoming RGI version targeting outlines at year 2015.

Do you have another idea about how to incorporate your data in RGI7?

[samherreid]

Hi Fabien,

Thanks for your reply, this is what I figured, but I wanted to make these shapefiles at least available and not just living on my harddrive when they have potential value to the community. Image tags for each shape in "addArea" is no problem. The debris lasso shapes aren't exactly fixed in time to a particular image, although I used imagery to draw them. Short of glacier surges with a terminus advance that took place between 2000 and the 2016 (approx) imagery I used, and assuming debris maps will be generated with imagery around 2000, I think these corrections can easily be incorporated into a 2000 timestamp RGI7. But the strict subset of GLIMS is the limiting factor. I work at a coffee shop now, so I don't have too much spare time to work on glacier inventory data management. Still, I'm curious also to see if the corrections agree with RGI7 updates, I'll run a few intersections and let you know what I find out :)

Thank you and the steering committee for all of your work coordinating this effort!

[bruceraup]

Hi Sam,

I echo Fabien's thanks for this work. If you find yourself with time to submit these data to GLIMS, let me know and we can talk through the best way to do that. Probably region-by-region would be easiest, and grouping outlines into a single submission that are from the same time is also best. If you already have image IDs in the files, then that's most of the work already done. Let me know, if and when.

Cheers, Bruce

On Tue, Feb 8, 2022 at 12:59 PM Sam Herreid @.***> wrote:

Hi Fabien,

Thanks for your reply, this is what I figured, but I wanted to make these shapefiles at least available and not just living on my harddrive when they have potential value to the community. Image tags for each shape in "addArea" is no problem. The debris lasso shapes aren't exactly fixed in time to a particular image, although I used imagery to draw them. Short of glacier surges with a terminus advance that took place between 2000 and the 2016 (approx) imagery I used, and assuming debris maps will be generated with imagery around 2000, I think these corrections can easily be incorporated into a 2000 timestamp RGI7. But the strict subset of GLIMS is the limiting factor. I work at a coffee shop now, so I don't have too much spare time to work on glacier inventory data management. Still, I'm curious also to see if the corrections agree with RGI7 updates, I'll run a few intersections and let you know what I find out :)

Thank you and the steering committee for all of your work coordinating this effort!

— Reply to this email directly, view it on GitHub https://github.com/GLIMS-RGI/rgi_issue_tracker/issues/31, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAL3IPSCTDP2KAQMTGOJ2KDU2FY33ANCNFSM5NYWG5MA . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

You are receiving this because you are subscribed to this thread.Message ID: @.***>