The river is now being asked to file its own water test
the river is now being asked to file its own water test: four deep models race to forecast dissolved oxygen in the river lee up to 30 days out, a bayesian lstm with mode decomposition wins daily oxygen prediction across eight yangtze sites, and a graph model respects river network topology to warn whole basins.
the stream now files its own quality ledger.
Context
The River Lee study compares Temporal Fusion Transformer, Informer, LSTM and GRU for dissolved-oxygen forecasts at 7- and 30-day horizons, using temperature, turbidity, pH, conductivity, chlorophyll and river flow. Its abstract reports TFT leading RMSE and SMAPE at seven days and leading the 30-day comparison; the ranking depends on metric and horizon.
The Credit River watershed study trains graph and transformer models on a network of hydrometric and water-quality stations in Ontario. GNN-SAGE benefits from information at surrounding stations and reports R-squared of 0.97 and RMSE of 0.34 mg/L against its study benchmarks.
The River Lee result is a horizon-specific model comparison; the Credit River result concerns spatial information shared between monitoring stations. Their errors and scores are not a common basin-wide benchmark or proof of a deployed warning system. A publication-index record matches the eight-site Yangtze VMD-Bayes-LSTM claim, but its primary article and evaluation split were not recovered in this research: unsupported by a primary reading, not refuted. Bayesian hyperparameter optimization should not be read as Bayesian predictive uncertainty.
Related work
- New Graph-Based and Transformer Deep Learning Models for River Dissolved Oxygen Forecasting ↗A concrete watershed test of whether neighbouring-station information improves oxygen forecasts.
Watch next
- Recover the primary eight-site Yangtze article and check decomposition leakage, held-out stations and five-day lead performance. For warnings, test low-oxygen event recall, calibration and prospective alerts rather than average regression fit alone.
Sources
Provenance
The note above is reproduced unedited from the original post, first published on Threads on 4 October 2026 at 04:02 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 →