The bloom is now being asked to file its own color report
the bloom is now being asked to file its own color report: a self supervised model fuses five satellites and solar fluorescence to map algal bloom severity without labels, a nasa tool stitches multi mission data to catch harmful blooms off florida and california, and a transformer turns a weather satellite into a ten minute ocean color camera.
the green tide now files its own chlorophyll ledger.
Context
SIT-FUSE is a self-supervised framework that fuses reflectance from VIIRS, MODIS, Sentinel-3 and PACE with TROPOMI solar-induced fluorescence. It produces harmful algal bloom severity and speciation maps without per-instrument labeled data, using hierarchical deep clustering. The arXiv paper validates against in-situ data from the Gulf of Mexico and Southern California for 2018 to 2025 and reports strong agreement with total phytoplankton, Karenia brevis, Alexandrium and Pseudo-nitzschia measurements.
NASA JPL's May 20, 2026 release describes the same tool, published in AGU Earth and Space Science. It fuses findings from five space missions or instruments, was trained on satellite data from 2018 and 2019, and was evaluated on later periods in the same areas, in western Florida and Southern California. The team says it is still working on more coastlines and on lakes.
A Chinese Academy of Sciences release dated July 8, 2026 describes a transformer that retrieves remote-sensing reflectance from the Himawari-8 geostationary satellite every 10 minutes. It combines a conventional atmospheric correction with a transformer trained on MODIS Aqua and AERONET-OC data. Against AERONET-OC it reports RMSE reductions of 34%, 26% and 12% at 470, 510 and 640 nm compared with the official hourly product.
These are two different jobs. SIT-FUSE maps bloom severity and species from several satellites and checks the result against field samples. The Himawari-8 transformer retrieves remote-sensing reflectance, and the CAS release does not report a chlorophyll or bloom map itself. The note's 'ten minute ocean color camera' fits the reflectance claim, and 'chlorophyll ledger' goes beyond what the release reports. The correlation above 0.98 in the CAS release is on test data; the 34, 26 and 12 percent figures are against in-situ stations. The JPL release calls the result initial and the tool a possible aid for choosing where to sample, not a replacement for water testing.
Related work
- NASA-developed AI Could Help Track Harmful Algae (NASA JPL) ↗NASA's release on the multi-mission bloom tool, with the five-mission count and the 2018 to 2019 training years.
- Fusing Multi- and Hyperspectral Satellite Data for Harmful Algal Bloom Monitoring with Self-Supervised and Hierarchical Deep Learning ↗The SIT-FUSE paper with the sensors, species and validation regions.
- New Algorithm Unlocks 10-Minute Ocean Color Observations from Himawari-8 (Chinese Academy of Sciences) ↗The source for the 10-minute cadence and the RMSE reductions.
Watch next
- Look for bloom-event skill, such as how early SIT-FUSE flags a bloom relative to field closures, and for results beyond the two coasts. For Himawari-8, look for a chlorophyll or bloom product and for performance in water types outside the training data.
Sources
- NASA-developed AI Could Help Track Harmful Algae (NASA JPL)jpl.nasa.gov
- Fusing Multi- and Hyperspectral Satellite Data for Harmful Algal Bloom Monitoring with Self-Supervised and Hierarchical Deep Learning (arXiv)arxiv.org
- New Algorithm Unlocks 10-Minute Ocean Color Observations from Himawari-8 (Chinese Academy of Sciences)english.cas.cn
Provenance
The note above is reproduced unedited from the original post, first published on Threads on 4 October 2026 at 03:33 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 →