Contribution seq 22639

Payload

{
  "excerpt": "The researchers mapped six natural and seminatural open wetland types across large parts of Europe at 10-m resolution, with overall accuracy of 96.6 ± 0.3%, covering an estimated 413,504 ± 17,782 square kilometres, or 6.8 ± 0.3% of the respective land surface. They used an XGBoost machine-learning algorithm trained on 101,000 CORINE Land Cover-labelled sample points and applied it to a multisensor satellite data cube spanning 2017–2019. The predictor set combined optical and radar observations from more than 470,000 Sentinel-1 and Sentinel-2 scenes with auxiliary layers on topography, climate, biomes and soil. An independent validation sample and design-based area estimates were used. European wetlands were fine-grained and highly fragmented: 27–33% of wetland area occurred in map-defined patches smaller than 25 hectares and 7–11% in patches smaller than 1 hectare. Small-patch prevalence was greatest for inland marshes, with 48.8% below 25 hectares and 19.8% below 1 hectare, while intertidal flats had the lowest shares. Coarser inventories commonly miss wetlands below 25 hectares and below 1 hectare.",
  "locator": {
    "section": "European-scale mapping of open wetlands"
  },
  "source_id": "cbcb46a1-0d18-8a24-a0b6-c2b9d0e155e4",
  "excerpt_hash": "sha256:eb436604e95f642f80233f44f11e28f6ed3c856c6e318094308a294a3372e38b",
  "locator_type": "TRANSCRIPTION"
}

Rows created

• SourceLocation e26f196d-9ff8-81fe-b8cd-fefa5cfed93e

Status history

Audits

None.

Audit sampling (audit-policy-v0.1): probability 1.0000, sampled yes; inputs {"n":"0.00","mean":"0.5000","domain":"general","task_type":"MANUAL","principal_tier":"PSEUDONYMOUS","downstream_count":0,"outcome_is_unusual":false}