Estimating Global “Blue Carbon” Emissions from Conversion and Degradation of Vegetated Coastal Ecosystems

Every country in the world tracks its carbon emissions. Every smokestack, every tailpipe, and every hectare of cleared forest has an entry somewhere in an official accounting sheet. And yet there is an entire category of emissions — potentially billions of tons of carbon dioxide every year — that has never appeared on any of those sheets. Not because it's small, but because until recently, no one had ever measured it. The missing entry is the carbon buried under salt marshes, mangrove forests, and seagrass beds along every coastline on Earth. Pendleton and colleagues set out to fix that in their landmark paper estimating global blue carbon emissions from the conversion and degradation of vegetated coastal ecosystems. Blue carbon is simply the carbon stored by these three habitat types — tidal marshes, mangroves, and seagrasses. Together, they cover roughly 49 million hectares worldwide. What makes them unusual isn't just that they absorb carbon from the atmosphere. It's where that carbon goes. It sinks into the sediment beneath them, into waterlogged, low-oxygen mud where microbial decomposition slows almost to a halt. These sediments can be several meters deep, and the carbon locked inside them has been accumulating for centuries. That distinction matters enormously. Before this paper, attention to blue carbon had focused on annual sequestration rates — the flux of carbon these ecosystems pull down each year. But that framing misses most of the story. Coastal habitats sit on top of a vast archive of previously stored carbon, and when you destroy the habitat, you don’t just stop future uptake. You crack open the archive. Disturbed sediments are exposed to oxygen, microbial decomposition accelerates, and carbon that has been locked away for generations can escape to the atmosphere in years. Pendleton and colleagues cite empirical evidence from cleared Panamanian mangroves where sediment carbon dropped by 50 percent within just eight years of land clearing. So the question Pendleton and colleagues asked was: how much carbon is actually being released this way, globally, right now? To answer it, they built the first quantitative global estimate, combining three inputs for each of the three ecosystem types: total habitat area, the annual rate at which that area is being lost, and the near-surface carbon stock, meaning the carbon in vegetation plus the top meter of sediment. The numbers going into the model reflect how much uncertainty already exists in the underlying data. Tidal marsh area was estimated at a central value of 5.1 million hectares but ranged from 2.2 to 40 million — a factor of nearly twenty. Mangrove area was better constrained, centered at 14.5 million hectares. Seagrass coverage was estimated at 30 million hectares but with a wide range of 17.7 to 60 million. Annual loss rates across the three ecosystems ran from under one percent to over three percent per year. Per-hectare carbon stocks, once converted to potential carbon dioxide emissions using the standard factor of 3.67, came out to roughly 949 metric tons of carbon dioxide per hectare for tidal marshes, 1,492 for mangroves, and 522 for seagrasses. To combine those ranges into a single estimate, the team used a Monte Carlo framework. Think of it this way: instead of plugging in one value for each input and getting one answer, you run 50,000 simulations, each time drawing a different plausible value from the known range of each input. Some runs use conservative estimates of area and loss rate, while others use high ones. The full cloud of 50,000 outputs gives you a distribution of possible answers, and from that distribution, you can read off a range that captures 90 percent of the outcomes. That's what Pendleton and colleagues report — a 90 percent confidence interval rather than a false-precision single number. One additional key assumption: they modeled a range of 25 to 100 percent for how much of the near-surface carbon pool is actually released to the atmosphere upon conversion, acknowledging that some carbon might be buried or redistributed rather than emitted. The result: between 0.15 and 1.02 petagrams — that's billion metric tons — of carbon dioxide released per year, with a central estimate of 0.45 petagrams. To anchor that in something familiar, global deforestation emits around 5.5 petagrams of carbon dioxide per year. Blue carbon losses from coastal conversion are equivalent to 3 to 19 percent of that. At the central estimate, you're talking about roughly the same annual carbon dioxide emissions as the United Kingdom. At the high end, you approach Japan's annual fossil fuel emissions. The wide range isn't a sign of a weak estimate; it's an honest reflection of genuinely sparse data. Pendleton and colleagues ran a sensitivity analysis to find which inputs drove the uncertainty most. The answer was clear: not the carbon stock values, not even the assumed fraction of carbon released upon conversion, but the global extent of these habitats and how fast they're being lost. Uncertainty in tidal marsh spatial extent alone accounted for 30 percent of total uncertainty. Uncertainty in mangrove conversion rates contributed 18 percent. Seagrass conversion rates contributed 14 percent. By contrast, uncertainty in the proportion of carbon lost upon conversion contributed just 2 percent for tidal marshes. Better maps and better loss-rate data would tighten this estimate dramatically. Of the three ecosystems, mangroves contribute the largest share of total emissions — roughly half — because they combine large per-hectare carbon stocks with substantial conversion rates. Seagrasses contribute the second largest share, compensated by their greater global area despite lower per-hectare stocks. And these are conservative estimates in a specific technical sense: the analysis counts only the top meter of sediment plus aboveground vegetation. Deep sediment carbon, which can extend several more meters down, is not included. Neither is lost future sequestration. Methane dynamics are acknowledged as a gap — changes in methane emissions from converted coastal wetlands could alter the carbon dioxide-equivalent total by up to around 10 to 15 percent, but the authors could not fully quantify that from available data. To translate emissions into economic terms, Pendleton and colleagues applied a social cost of carbon of 41 U.S. dollars per metric ton of carbon dioxide — the marginal value of climate damages attributable to each additional ton emitted, expressed in 2007 dollars. The math is direct: emissions in tons multiplied by 41 dollars per ton. That yields annual economic damages of 6.1 to 41.9 billion dollars, with a central estimate of 18.5 billion dollars per year. Every year, from ecosystems that appear on no carbon accounting sheet anywhere. That's the structural gap this paper is pointing at. These emissions are real, they're ongoing, and they're completely invisible to carbon markets and national emissions inventories. Frameworks like Reducing Emissions from Deforestation and Forest Degradation, often abbreviated as REDD+, which reward countries for reducing deforestation emissions, don't cover coastal wetlands. Neither do most carbon market protocols. The paper notes that guidance for forest reference emission levels could, in principle, be extended to mangrove belowground carbon, and that nascent mechanisms for wetland drainage and rewetting under the Kyoto Protocol offer potential pathways. But right now, those pathways aren't being used. The research agenda that falls out of this work is specific. Pendleton and colleagues call for better global mapping of tidal marshes and seagrasses, more robust area conversion rates for mangroves and seagrasses, and empirical studies tracking what actually happens to sediment carbon in the years following different types of coastal disturbance. Those are tractable problems. And crucially, the authors argue that policy responses do not need to wait for perfect data. The lower bound of this estimate — 0.15 petagrams of carbon dioxide per year — is already globally significant. The floor of the range alone justifies action. What would inclusion look like? If coastal blue carbon were recognized in carbon accounting frameworks, it would change the economics of coastal land use. Conservation and sustainable management would generate carbon credits. The opportunity costs that currently drive conversion to aquaculture, development, and agriculture would be offset by payments for keeping sediment carbon in place. The value of a mangrove forest wouldn't just be the shrimp farm it could become. It would include the centuries of stored carbon it currently holds in trust. The world's carbon accounts have been missing a line item. Pendleton and colleagues didn't just find it — they measured it, estimated its cost, and pointed clearly at the policy machinery needed to address it. The sediments have been holding this carbon for generations. The question now is whether the accounting systems we use to govern climate will finally catch up. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.