コロンビア・カリブ海Cispatá湾マングローブ生態系におけるブルーカーボン露出とENSO関連分類不確実性の空間評価
Spatial Assessment of Blue Carbon Exposure and ENSO-Associated Classification Uncertainty in the Mangrove Ecosystems of Cispatá Bay, Colombian Caribbean (原題)
Juan David Medina-Olivera, Ricardo Diaz, Luis Carlos Sandoval Herazo
🤖 gxceed AI 要約
日本語
Sentinel-1/2画像のマルチテンポラルRandom Forest分類（2016–2025年）により、コロンビア・カリブ海Cispatá湾のマングローブ炭素貯留を評価。総生態系炭素密度511 Mg C ha−1を用い、生データ推定は700万〜890万tC、高信頼コア8,894 haで454万tCを貯留。コアのほぼ全量が人為的圧力から250m以内に位置し、空間的曝露を定量化した。
English
Multitemporal Random Forest classification of Sentinel-1/2 imagery (2016–2025) assessed blue carbon stocks in Cispatá Bay, Colombian Caribbean. Using a Tier 1 density of 511 Mg C ha−1, raw total-stock estimates ranged 7.0–8.9 MtC, with a high-confidence mangrove core of 8,894 ha storing 4.54 MtC. Nearly the entire core lies within 250 m of anthropogenic pressure, quantifying spatial exposure; ENSO differences are exploratory only.
Unofficial AI-generated summary based on the public title and abstract. Not an official translation.
📝 gxceed 編集解説 — Why this matters
日本のGX文脈において
日本ではブルーカーボンがJ-クレジット制度や生物多様性クレジットの議論で注目されるが、本論文は衛星分類と不確実性評価の手法面で参考になる。SSBJ・有報の自然資本開示を検討する企業にとって、リモートセンシングによる炭素貯留モニタリングの限界と検証手法は示唆的。
In the global GX context
Blue carbon sits at the intersection of biodiversity and climate disclosure under TNFD and CSRD, yet remains under-specified in ISSB-aligned reporting. This study's design-based validation and explicit treatment of classification uncertainty offer a methodological template for nature-related metrics, though its Colombian case limits direct transferability.
👥 読者別の含意
🔬研究者:リモートセンシングによるブルーカーボン推定の不確実性評価とdesign-based検証の手法が参考になる。
🏢実務担当者:自然資本・生物多様性開示を検討する企業にとって、衛星ベース炭素モニタリングの限界と検証要件を理解する材料。
🏛政策担当者:ブルーカーボンクレジットや沿岸生態系保全政策のモニタリング基盤設計に示唆を与える。
📄 Abstract（原文）
This study assessed blue carbon stocks in the Cispatá Bay Regional Integrated Management District (DRMI), Colombian Caribbean, using multitemporal Random Forest classification of Sentinel-1/2 imagery (2016–2025) spanning contrasting ENSO conditions. Mangrove total ecosystem carbon density was estimated using the transferred Tier 1 value of 511 Mg C ha−1; the locally reported aboveground-carbon value was used only to describe component partitioning and does not change that total density. Raw mapped total-stock estimates ranged from 7.0 to 8.9 million tC, with mangroves accounting for more than 92% of the total. A high-confidence classification-persistent mangrove core of 8894 ha stored 4.54 million tC. Using the unfiltered 2024 agriculture/pasture plus aquaculture layer, the entire core occurred within 250 m of mapped anthropogenic pressure (mean distance 24.1 m; maximum 186.8 m). In a conservative sensitivity scenario that removed pressure patches < 0.5 ha, 99.1% remained within 250 m and 100% within 500 m (mean 53.9 m; maximum 382.1 m). Independent design-based validation was available only for 2020 and 2024, with area-adjusted overall accuracies of 80.2% and 72.4%, respectively; the difference in conventional overall accuracy (54.0% versus 50.4%) was not statistically significant. Consequently, ENSO-related differences are interpreted as exploratory associations rather than causal effects. The results quantify spatial exposure to surrounding land use and provide a monitoring baseline that complements field-based programs in Cispatá Bay.
🔗 Provenance — このレコードを発見したソース
- openalex https://doi.org/10.3390/earth7050165first seen 2026-10-07 05:04:19
🔔 こうした論文の新着を逃したくない方は キーワードアラート に登録（無料・3キーワードまで）。
gxceed は公開メタデータに基づく研究支援データセットです。要約・翻訳・解説は AI 支援で生成されています。 最終的な解釈・検証は利用者が原典資料に基づいて行うことを前提とします。