The bird is now being asked to keep its own travel diary
the bird is now being asked to keep its own travel diary: 108 autonomous sensors along a european migration route record and classify calls around the clock, google’s perch v2 pulled six sentinel seabird species from 49,383 acoustic events in patagonia, and a radar classifier lifts species resolution nearly 8 fold in israel.
the flock now files its own migration log.
Context
TABMON (preprint January 2026, later in Methods in Ecology and Evolution) is a standardised network of 108 autonomous sensors across Norway, the Netherlands, France and Spain along a bird migration route. Audio is recorded continuously, uploaded in near real time and passed through an automated analysis pipeline built to support expert validation. The authors say it shows both the opportunities and the current limits of near real time acoustic monitoring.
For Perch 2.0 in Patagonia, the closest source found is a GitHub issue (April 2026) from a conservation project at Meullin-Puye sanctuary in Chilean Patagonia. It says Perch 2.0 run on 200 ambient recordings from three sensors produced 27,460 detections with a high false-positive rate for non-Patagonian species, and asks for help fine-tuning for Southern Hemisphere species. It lists steamer-ducks, petrels and albatrosses as species of interest.
The radar paper (bioRxiv, 2026, Yuval Werber, University of Haifa) divides targets seen by the Birdscan MR1 vertical radar into morphology and movement based morphotypes. Using data from the Hula Valley station in Israel it reports a nearly 8-fold increase in bird classification resolution, and says most morphotypes correspond to 1 to 10 local species.
The note's 'six sentinel seabird species from 49,383 acoustic events' could not be matched. The Patagonia source found reports 27,460 detections from 200 recordings, lists more than six species of interest, and says the model needed fine tuning because of false positives, so it supports a different picture from 'pulled out' species cleanly. A related article on Hackernoon could not be read, so the 49,383 figure is unsupported here, not refuted. The issue is a project's own request to the Perch team and not a peer reviewed result. The radar claim is about taxonomic resolution (number of classes), not accuracy of species identification, and it rests on one site and one author. TABMON is a design and deployment paper, so it shows the network exists, not how accurate its classifications are.
Related work
- TABMON: Design and deployment of a transnational passive acoustic monitoring network for European birds (EcoEvoRxiv) ↗Source for the 108 sensor network and its near real time pipeline.
- TABMON: Design and deployment of a transnational passive acoustic monitoring network for European birds (Methods in Ecology and Evolution) ↗Published version of the TABMON paper.
- Perch 2.0: The Bittern Lesson for Bioacoustics (arXiv) ↗The model paper for Perch 2.0.
- Missing Southern Hemisphere seabird species in ebird2021 taxonomy (GitHub issue, google-research/perch) ↗Closest source found for Perch 2.0 on Patagonia recordings. Its numbers differ from the note.
- The morphotype approach to classification of aerial animals in radar data (bioRxiv) ↗Source for the morphotype method and the 8-fold resolution gain in Israel.
Watch next
- Find the source for 49,383 events and six sentinel species. Check the radar morphotype approach against species confirmed on the ground.
Sources
- TABMON: Design and deployment of a transnational passive acoustic monitoring network for European birds (EcoEvoRxiv)ecoevorxiv.org
- Missing Southern Hemisphere seabird species in ebird2021 taxonomy (GitHub issue, google-research/perch)github.com
- The morphotype approach to classification of aerial animals in radar data (bioRxiv)biorxiv.org
Provenance
The note above is reproduced unedited from the original post, first published on Threads on 3 October 2026 at 23:47 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 →