The problem
About 56 million years ago, Earth warmed rapidly as carbon entered the atmosphere and ocean. How much came from each store, and which releases started the warming rather than followed it? Volcanoes could supply carbon directly or heat carbon-rich sediments; warming itself could also release carbon held in other stores.[1] [3] [4]
Earth warmed rapidly about 56 million years ago while carbon entered the atmosphere and ocean. What were the amounts, source mixture and sequence of those releases, including the initial input and later feedbacks? Carbon-isotope mass balance weights each source by its amount and isotope composition. Joint pH and carbon-isotope inversion can constrain an input more tightly: a 2025 study favors thermogenic release for the preceding excursion. Fang’s terrestrial-loss calculations instead respond to imposed climates, addressing a possible feedback rather than recovering an observed release history.[1] [3] [4]
Scope
The amount, source partition and release sequence of carbon during the Paleocene–Eocene Thermal Maximum, including initiating inputs versus warming feedbacks.
What remains open
The rocks show a major carbon-cycle disturbance, but several source mixtures can leave similar isotope signatures. Researchers have not uniquely determined how much came from each store or which releases started the event rather than amplified it.[1] [2]
Source mixtures remain nonunique, and a bulk isotope balance does not explicitly reproduce all ocean carbonate chemistry. Vegetation analogs and climate forcing introduce further uncertainty. A plausible feedback magnitude cannot by itself determine the trigger, the release rate or the division between mantle and thermogenic carbon.[1] [2] [3] [4]
Why it matters
The event can test how an initial carbon input and carbon stored on land and in the ocean interact during warming. Carbon released in response to warming could amplify the original disturbance, whereas an earlier release plays a different role. Separating those roles helps explain the warming rather than treating carbon recorded in the same sediments as one cause.[1] [4]
The PETM offers a test of how geological carbon release and climate-sensitive carbon stores act together. Carbon loss after warming may amplify a prior trigger; carbon required before warming has a different role. Chronology, ocean acidity and isotopes can test whether an account explains the forcing and feedbacks consistently. Matching a total carbon amount alone cannot establish that causal sequence or the release rate.[1] [2] [3] [4]
Research directions
1 direction to explore
Separate the trigger from carbon feedbacks
Use jointly dated carbon, temperature and ocean-chemistry changes to test when each proposed source contributed.
Definitions
3 terms
isotope signatures
Patterns in the relative amounts of heavier and lighter forms of the same element. Different carbon sources can have different patterns, while mixtures can leave similar overall changes.[2] [3] [4]
thermogenic
Produced by heating organic matter in sediments. In this context, magma intruding into carbon-rich rock can release sedimentary carbon; this differs from adding carbon directly from the mantle.[4]
feedback
A response that changes the process that caused it. Carbon lost from warming land could add to the original climate disturbance; that conditional response is distinct from the initial carbon input.[1]
Sources
Partial results
Ocean-chemistry reconstruction supports a large, predominantly volcanic-associated release. Scope: Boron and carbon isotopes combined with Earth-system modeling.
Its source interpretation includes distinctions between mantle carbon and carbon released by heating organic sediments.
Source date: 2017-08-31The study infers a large carbon input compatible with North Atlantic volcanism.Read the source ↗A multiproxy reconstruction allows methane, organic or mixed PETM sources. Scope: Pacific microfossil records and Bayesian SST/CO2 reconstruction.
The allowed source mixture differs from a unique volcanic-only explanation.
Source date: 2024-08-26Joint temperature and CO2 estimates constrain the inferred input.Read the source ↗Vegetation modeling quantifies scenario-dependent terrestrial carbon losses. Scope: Imposed high-CO2 PETM climate experiments.
These are modeled feedback responses, not a measured initiating release or a unique historical partition.
Source date: 2026-05-29Stronger forcing produces much larger simulated land-carbon losses.Read the source ↗Joint pH and carbon-isotope modeling favors thermogenic carbon in a preceding excursion. Scope: The pre-onset excursion in a high-resolution eastern Tethys record.
The chronology has alternative age models, and this source inference does not partition all carbon during the PETM.
Source date: 2025-06-30Data assimilation explores several onset-duration scenarios.Read the source ↗- Terrestrial carbon stock vulnerability during the Paleocene–Eocene Thermal Maximum
Notes on Terrestrial carbon stock vulnerability during the Paleocene–Eocene Thermal Maximum
Abstract and selected results, source-partition discussion and methods read. Vegetation experiments use modern functional analogs and imposed climate/CO2 scenarios.
- Long- and short-term coupling of sea surface temperature and atmospheric CO2 during the late Paleocene and early Eocene
Notes on Long- and short-term coupling of sea surface temperature and atmospheric CO2 during the late Paleocene and early Eocene
Primary abstract and indexed author-PDF significance/abstract read; full main paper and supplements not read. Separates PETM source inference from longer-term volcanic forcing.
- Very large release of mostly volcanic carbon during the Palaeocene–Eocene Thermal Maximum
Notes on Very large release of mostly volcanic carbon during the Palaeocene–Eocene Thermal Maximum
Primary abstract and selected indexed discussion read; complete publisher article unavailable. Volcanically triggered sedimentary carbon must be distinguished from mantle CO2.
- Millennial-timescale thermogenic CO2 release preceding the Paleocene-Eocene Thermal Maximum
Notes on Millennial-timescale thermogenic CO2 release preceding the Paleocene-Eocene Thermal Maximum
Abstract, introduction, selected chronology results and double-inversion methods read on October 2. The model attribution concerns the preceding excursion, with alternative age models; it is not a partition of every PETM input.