For decades, the global carbon cycle relied on a predictable buffer: terrestrial ecosystems absorbed roughly a quarter of anthropogenic carbon emissions. When tracking this sink, scientists primarily monitored deforestation—the outright clear-cutting of forests for agriculture, roads, or pasture. If a forest pixel on a satellite map remained green, it was largely assumed to be intact and functioning as a carbon sink.
However, a major paradigm shift is occurring in earth system science. Recent research demonstrates that tracking land-cover change alone misses the majority of carbon dynamics. Forests are losing massive amounts of carbon within standing canopies through a process known as forest degradation. This subtle thinning of vegetation—driven by selective logging, edge effects, fragmentation, and climate-induced mortality—is quietly shifting entire continents from net carbon sinks into net carbon sources.
The Measurement Gap: From Hectares to Biomass
Why did this massive carbon loss remain hidden for so long? The answer lies in the limitations of traditional optical remote sensing. Optical satellites measure greenness at the surface; a degraded forest with a partially thinned canopy can look just as green from above as a pristine, old-growth forest.
To build a true accounting of carbon stocks, researchers had to shift from measuring forest area to measuring forest volume. This has been achieved by leveraging low-frequency "L-band" Vegetation Optical Depth (L-VOD) microwave satellite data. Because microwave signals can penetrate the canopy and are highly sensitive to the water content in woody biomass, they provide a direct proxy for aboveground carbon (AGC) density.
Using these tools, Lei Fan et al. (2019) revealed a startling reality: across the pantropical belt, carbon losses from forest degradation, selective logging, and climate-induced mortality are approximately twice as large as losses from clear-cut deforestation. Similarly, Yu Feng et al. demonstrated that subtle thinning of vegetation is the primary driver of biomass decline across the tropics, outstripping outright land clearing.
The problem is not unique to tropical zones. In North American boreal forests, Ling Yu et al. found that forest degradation contributes more to carbon loss than actual forest cover loss, emphasizing that forest quality and health, rather than just forest extent, dictate global carbon retention.
Traditional Optical Sensing: [Green Canopy] ──> Assumed Healthy Sink
Microwave L-VOD Sensing: [Thinned Wood] ──> Hidden Degradation Carbon Loss
The Mechanics of Degradation Feedbacks
To understand why degradation is such a potent driver of carbon loss, one must look at the compounding feedbacks it triggers. When a forest is degraded—whether by human activity like selective logging or natural disturbances like low-intensity fires—the structural integrity of the ecosystem is compromised.
Jiaying He et al. note that degradation accounts for 32% to 69% of the total carbon impact in tropical regions, representing a continuous "pulse" of emissions rather than a single clearing event. This structural damage has dual consequences:
- Biogeochemical Feedbacks: Dead wood decays, and surviving trees become more vulnerable to windthrow and environmental stress, accelerating long-term carbon release.
- Biophysical Feedbacks: Canopy thinning alters how the forest interacts with energy and water. Lei Zhu et al. discovered that the "biophysical" warming caused by changes in water evaporation and surface reflection in degraded tropical moist forests is nearly equal in magnitude to the "biogeochemical" warming caused by the carbon released from these damaged forests.
The Asymmetry of Forest Age
A common counterargument is that degraded or harvested forests will eventually regrow, recapturing the lost carbon. However, the temporal dynamics of forest growth create a severe carbon deficit.
Simon Besnard et al. investigated how global forest age transitions covary with the net carbon balance. They found that while young, regenerating forests exhibit rapid growth rates and can temporarily boost localized carbon uptake, they fundamentally cannot compensate for the massive, long-term carbon storage lost when old-growth forests are disrupted. The global decline in forest age—driven by escalating disturbances—creates a net loss of aboveground carbon that is particularly severe in critical basins like the Amazon, Congo, and Siberia.
A Continental Shift from Sink to Source
The cumulative effect of degradation and climate stress is driving a widespread transition of terrestrial ecosystems from carbon sinks to net sources. This shift is manifesting across distinct geographical zones:
The Amazon Basin
Long considered a vital global carbon sink, the Amazon is tipping. Dominic Fawcett et al. showed that accelerating deforestation and subsequent degradation losses now collectively outweigh the carbon gains from secondary forest regrowth, leading to a net loss of old-growth biomass. This trend is exacerbated by climate change; Yitong Yao et al. project that drought-induced tree mortality risks turning large swaths of the Amazon into permanent carbon sources by the mid-21st century. High-resolution disaggregation by Arthur Nicolaus Fendrich et al. confirms that these human-driven degradation pressures have intensified significantly over the past decade.
Northern and Boreal Ecosystems
In colder climates, warming temperatures were historically expected to lengthen growing seasons and boost carbon uptake. However, Tao Wang et al. demonstrated an emerging negative impact of warming on summer carbon uptake in northern ecosystems, signaling a decline in vegetation's positive response to temperature. In Siberia, Lei Fan et al. (2022) found that increased disturbances from wildfires and droughts have reduced the Siberian forest sink to a near-neutral carbon balance, decoupling superficial greening trends from actual biomass storage.
The African Continent
Africa's carbon balance has historically been highly dynamic. Shilong Piao et al. documented a shift from a net carbon source in the 1980s to a sink in the 1990s, driven by precipitation variability and $CO_2$ fertilization. However, recent comprehensive budgeting by Yolandi Ernst et al. indicates that this sink capacity is now diminishing, with the continent trending back toward a net carbon source due to escalating environmental pressures.
| Region | Historical Role | Current/Emerging Status | Primary Driver of Shift |
|---|---|---|---|
| Amazon | Major Carbon Sink | Net Carbon Source | Degradation & Drought Mortality |
| Siberia | Significant Carbon Sink | Near-Neutral Balance | Wildfires & Drought Disturbances |
| Africa | Dynamic Sink (1990s) | Diminishing Sink / Source | Climate Extremes & Land Use |
Implications for the Global Carbon Budget
When these regional shifts are integrated globally, the macro-picture reveals that the assumption that the land sink will indefinitely absorb a fixed portion of our emissions is breaking down.
In a major revision of the global carbon budget, Pierre Friedlingstein et al. demonstrated that the global land carbon sink is smaller and land-use emissions are larger than previously estimated, with climate change actively reducing sink efficiency. This vulnerability was starkly illustrated in 2024. Philippe Ciais et al. reported a severe, widespread weakening of the global land carbon sink driven by extreme heatwaves and droughts, which triggered record-high atmospheric $CO_2$ growth rates.
Understanding that forests are bleeding carbon from the inside out changes how we must approach climate mitigation. Conservation policies can no longer succeed merely by drawing lines around forests to prevent clear-cutting; they must actively manage and protect the structural health and functional integrity of standing forests to prevent them from becoming the very engines driving climate acceleration.
Key Foundational Work
Emerging negative impact of warming on summer carbon uptake in northern ecosystems (2018) — contributed 2 concept elements to this direction.
Variability and recent trends in the African terrestrial carbon balance (2009) — contributed 1 concept elements to this direction.