衛星埋め込みフィールドは浅海サンゴ礁生息地について何を教えてくれるか？グレートバリアリーフにおけるAlphaEarthのデコーダビリティ調査

Unofficial AI-generated summary based on the public title and abstract. Not an official translation.

🔬研究者:Provides a rigorous framework for validating satellite embedding-based ecological indicators, with cautionary lessons on cross-validation and annotation schemes.

🏢実務担当者:Offers insights into the reliability of satellite-based coral reef monitoring, useful for environmental impact assessments and nature-related disclosure reporting.

🏛政策担当者:Highlights the potential and limitations of satellite embeddings for ecosystem monitoring, informing policy on climate adaptation and biodiversity conservation.

Annual satellite embedding fields compress a year of multi-sensor observation into a 64-dimensional vector per 10 m pixel, globally; whether they retain ecologically meaningful information about the submerged reef benthos is untested. We conducted a registered decodability survey of the AlphaEarth Foundations embedding over shallow (< 10 m) Great Barrier Reef habitats: every candidate output was registered with mechanism, test and pass gate before testing, evaluated under one fixed validation protocol (spatial and temporal holdouts, buffered isolation, baseline ablations, cluster-robust uncertainty), and reported pass or fail. Ground truth comprised 104,668 photo-quadrat images (37 sites, 7 years) at one reef, photo-transects at 67 further reefs in two eras, and independent point-label and map products. Benthic state is broadly decodable with linear read-outs: eight channels pass at one or more of six scales (10–100 m), led by hard coral (held-out rho .73 at 10 m, .80 at 50 m). A persistence control shows this cross-sectional skill largely reflects stable habitat structure, refreshed annually. Six channels replicate on never-trained reefs and across survey eras and methods, and decoded sand agrees with the independently produced Allen Coral Atlas at AUC .96 (sand is itself optically recoverable). Year-over-year change, measured at the single multi-year reef, is decodable only in the two documented bleaching-event years and only at 20–50 m: in the stronger (2024) event the coral-loss axis discriminates ≥ 10-point loss blocks at AUC .83 (cluster CI .69–.92; detection 13–85 % at a 10 % false-alarm budget). Quiet-year change lies below the noise floor of the registered linear protocol at every scale and temporal lag. A post-registration campaign against independent AIMS monitoring archives (184 manta-tow reefs; 306 fixed photo-transect sites, 289 after a coordinate screen) bounds deployment: site-level cover transfers at roughly a third of the agreement between the survey programmes themselves (reef windows, a quarter), but decoded trajectories fail with a directional bias (post-disturbance surfaces read as coral, so apparent gains rank below chance), and reef-mean aggregation resolves ≈ 7-point changes only at locally calibrated reefs. Ground-truth reliability places the decodability ceiling near rho .8: the limit is the signal itself, not the amount of data. Naive cross-validation inflated skill by up to +.28 rho, and annotation-scheme differences between ground-truth programmes exceeded ecological signal; both cautions generalise to any embedding-decoding study. The catalogue, its conditional channels and its documented negatives define what satellite embeddings can and cannot yet contribute to reef monitoring.

• EarthArXiv https://eartharxiv.org/repository/object/14589/download/25421/first seen 2026-08-26 04:13:49 · last seen 2026-09-08 04:15:12