The topsoil is now being asked to keep its own carbon ledger
the topsoil is now being asked to keep its own carbon ledger: a four dimensional deep model projects global soil organic carbon to 2100 with a 200 centimetre depth profile at an r squared of 0.90, a pretrained model trained on 106,167 global samples transfers to regions with scarce data, and an open machine learning product now maps soil organic carbon at 30 metre resolution worldwide.
the ground now files its own carbon balance.
Context
The pretrained model matches GSoilCPM work reported as Regional-scale soil carbon predictions can be enhanced by transferring global-scale soil-environment relationships. Its abstract says it uses a large amount of environmental covariate data and 106,167 soil samples across the globe to test whether global relationships improve regional predictions. The code is on GitHub as leizhang-geo/GSoilCPM.
The open product is OpenLandMap-soildb, a data description paper in Earth System Science Data (6 February 2026) by Tomislav Hengl and colleagues: global soil information at 30 m spatial resolution for 2000 to 2022 and later, based on spatiotemporal machine learning and harmonized legacy soil samples. Its tutorial and data are on GitHub under openlandmap/soildb.
The four dimensional deep model that projects soil organic carbon to 2100 with a 200 cm depth profile at R squared 0.90 was not located. Searches returned related spatiotemporal soil carbon work and a 0 to 200 cm random forest study of temperate grasslands, which are different models, so the 2100 projection and the 0.90 figure are unsupported here, not refuted. The note says the open product maps soil organic carbon at 30 m. The ESSD paper describes a 30 m soil information product, and whether soil organic carbon is among its layers at that resolution was not read in the excerpt, so that match is taken from the product's scope, not from a layer list.
Watch next
- The 4D deep model paper and its validation. The list of soil properties in OpenLandMap-soildb.
Sources
- Regional-scale soil carbon predictions can be enhanced by transferring global-scale soil-environment relationshipsexa.ai
- leizhang-geo/GSoilCPM (GitHub)github.com
- OpenLandMap-soildb: global soil information at 30 m spatial resolution for 2000-2022+ (Earth System Science Data)essd.copernicus.org
- openlandmap/soildb (GitHub)github.com
Provenance
The note above is reproduced unedited from the original post, first published on Threads on 4 October 2026 at 12:39 IST. Sources are the papers and datasets the note draws on.
View the original post ↗Embed this note
More notes
The air is now being asked to keep its own ledger
the air is now being asked to keep its own ledger: ecmwf’s aifs compo becomes the first ai model to forecast atmospheric composition globally every three hours, cleanair simulates 365 days of pm2.5 over china in ten seconds, and a unified framework maps six pollutants at one kilometer across the whole country. the air now files its own composition report.
read the note →The current is now being asked to draw its own map
the current is now being asked to draw its own map: china’s langya 2.0 predicts six ocean phenomena including internal waves and mesoscale eddies, a deep net called wenhai resolves eddies globally with air sea flux formulas built in, and scripps infers surface currents from the way temperature patterns deform in satellite images. the ocean now files its own circulation report.
read the note →The soil is now being asked to report its own carbon
the soil is now being asked to report its own carbon: a nix color sensor paired with generative data augmentation predicts soil organic carbon without a lab, random forest drives 74 percent of soil health mapping studies, and sentinel 2 tracks five year carbon change across france and italy from 922 samples. the dirt now files its own carbon account.
read the note →