全球生態系碳停留時間與容量之評估於 紅樹林、鹽沼、海草床、泥炭地、巨藻林與生物幫浦

Abstract

氣候變遷迫切的威脅與全球二氧化碳排放量的持續增加，促使各界積極尋求有效的碳移除策略，海洋與浸水型(waterlogged)環境因可吸收約30%的人為排放，被視為潛在的碳中和途徑；碳若能在儲存在系統中達100年，便被視為已完成碳封存，然而此一常用準則可能大幅高估浸水型環境的碳儲存量；此外，現有研究在評估不同浸水型生態系之碳穩定性時，缺乏統一的分析框架，限制了跨系統比較的可行性。 本研究透過一致的箱體框架(box framework)以及整合現有文獻中可取得的碳庫存與通量資料，對浸水型系統中的碳穩定性與儲存容量進行系統性的全球評估，浸水型生態系包含紅樹林、鹽沼、海草床、泥炭地、巨藻林及生物幫浦(BCP)；同時，本研究亦建立一維擴散的海洋模型，以模擬海洋垂直碳傳輸並評估注入碳的長期洩漏行為。 模型模擬結果顯示，本研究可重現文獻所報導之全球尺度與區域尺度(大西洋、太平洋與印度洋)碳洩漏行為於±20%的範圍內重現注入碳回返大氣的比例誤差範圍內，且顯示在碳注入深度越深，碳洩漏速率越慢；進一步結果指出，注入的碳會隨時間逐漸返回大氣，顯示以100年作為碳封存指標可能高估實際有效碳儲存量約20–40%。在各類浸水型生態系中，超過90%的總碳輸入量並未滯留，而是快速通過系統；不同生態系之碳停留時間差異顯著，紅樹林、鹽沼、海草床與巨藻林的停留時間少於30年，而泥炭地與BCP則可達約200年；紅樹林、鹽沼、海草床與泥炭地雖具有可觀的碳儲存容量(20–130 kg C m-2)，但是其碳累積速率相對緩慢(10–300 gC m-2 y-1)；然而，過往研究指出，棲地流失可能迅速釋放60–80%的碳庫，進而使生態系轉變為碳源；過去文獻中預測亦顯示，海洋碳吸收受到碳酸鹽系統機制的限制，預期於本世紀末達到約 4–6 Pg C y⁻¹ 的上限，顯示海洋碳匯容量有限。 整體而言，浸水型生態系並非一種即時且高效的碳移除途徑。因此，在評估碳封存潛力時，納入完整的時間尺度至關重要，單純依賴100年的封存指標，可能會高估碳的穩定度。

The impending threats of climate change and the ever–growing global CO2 emission has prompted for major effective carbon removal strategies. Storage in oceans and other waterlogged environment have been suggested as a plausible carbon neutralization option as they can absorb ~30% of anthropogenic emissions. Although carbon is commonly considered as sequestered if it can remain stored/immobile for 100 years, such criterion may substantially overestimate total carbon storage in waterlogged environment. In addition, assessment of carbon stability across waterlogged ecosystems also suffer from the use of inconsistent frameworks. This study conducts a systematic global assessment on carbon stability and capacity in waterlogged systems by applying a consistent box framework to mangroves, saltmarshes, seagrass meadows, peatlands, macroalgal forests, and the biological carbon pump (BCP) integrating carbon stocks and fluxes availability in current literature. A one–dimensional diffusion–based ocean model is used to simulate vertical carbon transport and investigate long–term leakage of injected carbon. Model simulations reproduce global and basin–scale (Atlantic, Pacific, and Indian) estimates of the fraction of injected carbon returning to the atmosphere within ±20% and show slower leakage for deeper injections. These results further indicate that injected carbon gradually returns to the atmosphere, showing that the 100–year sequestration criterion likely overestimates effective carbon storage by 20–40%. Across waterlogged ecosystems, more than 90% of total carbon influx pass through the systems without being retained. Carbon retention times vary widely, from <30 years in mangroves, saltmarshes, seagrass meadows, and macroalgal forests to ~200 years in peatlands and the BCP. These ecosystems, referring to mangroves, saltmarshes, seagrass meadows, and peatlands, have substantial storage capacities (20–130 kg C m-2) but accumulate carbon slowly (10–300 gC m-2 y-1). However, these systems could become carbon sources as previous studies suggested habitat loss could release up to 60–80% of stored carbon therein. Similarly, literature simulations also projected that oceanic uptake carbon will peak at 4–6 Pg C y-1 by the end of the century dur to carbonate chemistry, indicating ocean as a finite carbon sink. Overall, waterlogged ecosystems do not represent an immediate or highly effective pathway for carbon removal, as reflected by their carbon retention times and capacities. Assessment of carbon sequestration with longer temporal scales is also warranted as the 100–year criterion can significantly overestimate carbon stability.