ESA Active Layer Thickness

[nocollier]

As mentioned on our last dev call, here is the data reference. @dlawrenncar @ckoven

https://catalogue.ceda.ac.uk/uuid/67a3f8c8dc914ef99f7f08eb0d997e23

[nocollier]

Evaluating permafrost physics in the Coupled Model Intercomparison Project 6 (CMIP6) models and their sensitivity to climate change https://tc.copernicus.org/articles/14/3155/2020/tc-14-3155-2020.pdf

[nocollier]

Analysis of Permafrost Thermal Dynamics and Response to Climate Change in the CMIP5 Earth System Models https://journals.ametsoc.org/view/journals/clim/26/6/jcli-d-12-00228.1.xml

[nocollier]

• Implemented effective snow depth, seasonal and thermal offsets, mean annual ground temperature, mean annual ground surface temperature, mean annual air temperature, active layer thickness, and thawed fraction used in the papers above.
• I used the permafrost extent to mask out active layer thickness which has produced easier to read maps
• I also have included a score for the fraction of the CALM sites where permafrost was observed in the models as well as a modeled depth.
• Much of the Burke paper is about relationships which I am trying to include. This requires encoding some snow depth data. The monthly data is in ascii format and so conversion is taking some time.
• Still need to look rather (or additionally) at comparing a log transform of ALT

[dlawrenncar]

Nice work. I'd like to talk through these results at some point. I want to make sure that everything is being compared correctly. And, I'm looking at the data and some of it doesn't make sense. The low ALT values in the ESACCI dataset in the southern domains for example, must mean something other than low ALT. Also, and to this end, can you run the ESACCI dataset as another model against the CALM network to get a sense of how much agreement there is or isn't between these two datasets?

On Wed, Sep 6, 2023 at 9:00 AM nocollier @.***> wrote:

• Implemented https://github.com/rubisco-sfa/ILAMB/blob/nate/alt/src/ILAMB/ConfALT.py effective snow depth, seasonal and thermal offsets, mean annual ground temperature, mean annual ground surface temperature, mean annual air temperature, active layer thickness, and thawed fraction used in the papers above.
• I used the permafrost extent to mask out active layer thickness which has produced easier to read maps https://www.ilamb.org/dev/HighLat/
• I also have included a score for the fraction of the CALM sites where permafrost was observed in the models as well as a modeled depth.
• Much of the Burke paper is about relationships which I am trying to include. This requires encoding some snow depth data https://nsidc.org/data/nsidc-0447/versions/1. The monthly data is in ascii format and so conversion is taking some time.
• Still need to look rather (or additionally) at comparing a log transform of ALT

— Reply to this email directly, view it on GitHub https://github.com/rubisco-sfa/ILAMB-Data/issues/48#issuecomment-1708542434, or unsubscribe https://github.com/notifications/unsubscribe-auth/AFABYVDZCUU5LHWPUBRIJSDXZCF2XANCNFSM6AAAAAA22GM6ZA . You are receiving this because you were mentioned.Message ID: @.***>

[nocollier]

I have tried to be careful about implementing what the papers say they are doing but it is easy to make a mistake. Perhaps some time when we can review the code would be a good idea. I will add ESACCI as a model in the next run, but for now consider these comparison plots of ESA to CALM. Note though that this is using my coarsened ESA dataset.