China's Great Green Wall Turns Taklamakan Desert Margins Into a Carbon Sink

Satellites detect carbon capture at the desert's edge

Vegetation planted around China's Taklamakan Desert is absorbing more carbon dioxide than it releases, according to research published in the Proceedings of the National Academy of Sciences. Scientists have identified a functioning carbon sink along the margins of a desert where shifting sands cover more than 95% of the landscape, recording atmospheric carbon dioxide concentrations 1 to 2 parts per million below those in surrounding areas.

The findings connect decades of planting under China's Three-North Shelterbelt Program with a measurable change in the atmosphere. Launched in 1978 and commonly called the Great Green Wall, the program was designed to slow desert expansion and protect land. Its planted vegetation is now also drawing carbon dioxide from the air through photosynthesis, the process plants use to turn sunlight, water and carbon dioxide into material they need to grow.

The study describes the first documented case of human intervention turning an extreme desert environment into a stable carbon sink. The change is concentrated around the Taklamakan's periphery, where planted vegetation has access to water flowing from surrounding mountains. Hardy shrubs are central to that transformation.

The Taklamakan covers approximately 130,000 square miles, or 337,000 square kilometers, in the Tarim Basin of northwestern China. It is slightly larger than Montana and roughly the size of Germany. High mountain ranges surround it, blocking moist air from entering the interior for much of the year. Those conditions helped give the desert its nickname, the Sea of Death.

Satellite images now show a green belt around its margins. China completed the desert's vegetative encirclement in 2024, a milestone in a planting effort that has continued for more than four decades. The findings offer evidence that even very dry landscapes can contribute to carbon storage, while also showing how strongly that contribution depends on water.

A planting program stretching from 1978 to 2050

Chinese officials began the Three-North Shelterbelt Program in response to accelerating desertification, the process through which land loses vegetation and becomes increasingly desertlike. During the 1950s, rapid urbanization and the expansion of farmland stripped away natural vegetation, leaving land exposed to worsening sandstorms and threatening agricultural productivity.

The government committed to planting billions of trees around the Taklamakan and Gobi deserts by 2050. The aim was to hold shifting dunes in place and stop the deserts from spreading. The schedule covers 72 years, with implementation milestones in 1998, 2011 and 2021.

The program's main dates trace a long effort rather than a single construction project:

• 1978: China launched the Three-North Shelterbelt Program.
• 1998, 2011 and 2021: The program reached key implementation milestones.
• 2024: Vegetation completed an encircling belt around the Taklamakan.
• 2050: The program's planned timeline reaches its endpoint.

More than 66 billion trees have taken root since the program began. Across China, forest cover has risen from 10% of the country's area in 1949 to more than 25% today. The Taklamakan's completed belt is part of that wider change in vegetation.

Protecting farmland and connecting oases

A 3,046-kilometer green corridor built through the desert is among the project's engineering achievements. It threads vegetation across the landscape and connects oases that were previously isolated. Around the desert's edges, planted belts also serve as barriers against moving sand.

Farmland protection was a central reason for the undertaking. Unrestrained desert growth threatened productive land, while dust storms affected cities across northern China. Vegetation helps by stabilizing dunes and reducing wind erosion, allowing roots and soil to hold material that would otherwise be carried away.

Beijing's objectives also included regional stability in western China, where tensions between minority ethnic groups and Han Chinese leadership have historically created friction. Securing agricultural land and limiting desert expansion formed part of that wider political purpose. Addressing China's carbon footprint became another aim.

Political continuity helped keep funding and labor available over more than four decades. The Chinese initiative persisted without interruption, unlike similar international efforts, including those launched in the Sahara Desert by the United Nations. That continuity gave vegetation time to establish itself and soils time to change.

How researchers measured the change from space

Scientists from the University of California, Riverside worked with researchers from the University of Houston, Tsinghua University and Caltech to investigate the desert's carbon dynamics. Their analysis used satellite observations to connect changes in vegetation with changes in atmospheric carbon dioxide.

Data from NASA's Orbiting Carbon Observatory and the MODIS satellite supported an examination of carbon dioxide concentrations and vegetation greenness over the Taklamakan's rim across 25 years. The team focused on two indicators: lower concentrations of carbon dioxide above planted areas and higher levels of solar-induced fluorescence.

Solar-induced fluorescence is a faint light that plants emit during photosynthesis. Satellites can detect it, giving researchers an indirect measure of how actively vegetation is using sunlight. When that signal rises alongside a fall in atmospheric carbon dioxide, scientists can examine how plant activity is contributing to carbon uptake.

The observations identified a carbon dioxide cold spot over treated areas, with concentrations 1 to 2 parts per million below surrounding regions. Parts per million, usually shortened to ppm, describes the share of a gas in the air. A concentration of 416 ppm means there are about 416 carbon dioxide molecules for every million molecules of air.

King-Fai Li, an atmospheric physicist at the University of California, Riverside and a collaborator on the study, explained why even modest vegetation matters in this setting:

This is not like a rainforest in the Amazon or Congo. Some afforested regions are only shrublands like Southern California's chaparral. But the fact that they are drawing down CO2 at all, and doing it consistently, is something positive we can measure and verify from space.

Seasonal rain strengthens the carbon drawdown

Rainfall gives the planted vegetation a seasonal boost. During the wet season, from July through September, precipitation averages about 0.6 inches, or 16 millimeters, per month. That is roughly 2.5 times the monthly precipitation during dry periods.

The extra moisture supports more vegetation and greater photosynthetic activity. Atmospheric carbon dioxide concentrations fall from about 416 ppm during dry periods to 413 ppm during the wet season, a seasonal decline of 3 ppm.

The measurements describe both a difference between planted areas and their surroundings and a change between seasons. The 1 to 2 ppm cold spot identifies lower carbon dioxide over the treated landscape. The decline from 416 to 413 ppm shows how carbon drawdown becomes stronger when seasonal moisture supports plant growth.

The Taklamakan remains an exceptionally dry environment despite this seasonal increase. Its planted margins capture carbon under conditions far removed from those of a tropical forest, making the role of water especially visible in the satellite observations.

Why shrubs matter as much as trees

Hardy shrubs are the main drivers of the carbon capture identified by the research, even though the program is often described as a tree planting effort. These woody plants resemble the chaparral of Southern California and can continue photosynthesizing in dry, hot conditions where many tree species would struggle.

Vegetation acts as a biological pump. Plants take carbon dioxide from the atmosphere and use its carbon to build roots and stems. Some of that carbon then enters soil organic matter, creating storage beneath the plants as well as within them.

Earlier research suggested that desert sand could also trap carbon dioxide through physical processes linked to temperature changes between day and night. Heating and cooling cause air in the sand to expand and contract, potentially storing roughly one million tons of carbon annually.

Scientists warned that this form of storage could become unstable as the climate warms. Higher temperatures could expand the trapped air and release stored carbon. The new findings identify biological storage through photosynthesis as a more stable and substantial sink.

The planted shrubs also change the physical landscape. By anchoring sand and reducing wind erosion, they create conditions in which further ecological succession can occur. Ecological succession means changes in the plants and other living organisms occupying an area as conditions develop over time.

Soil provides another layer of storage

Research along the Taklamakan's southern edge examined how planting affects carbon inside soil. Scientists used the delta13C natural abundance method to trace carbon moving through different soil fractions. The method follows naturally occurring differences in forms of carbon to help identify its movement.

The analysis found that afforestation, the establishment of vegetation on land previously without forest cover, increased both the organic carbon content of the soil as a whole and the formation of soil aggregates. The largest gains appeared after seven years of plant establishment.

Soil aggregates are clusters of particles bound together. Microaggregates are smaller clusters that can act as both sources and sinks for carbon moving through desert ecosystems. Their formation helps explain how a planted landscape can retain carbon beyond the roots and stems visible above ground.

Soil organic carbon is the largest carbon pool in terrestrial ecosystems, and approximately 90% of topsoil carbon is stored within soil aggregates. The findings from the desert's southern edge connect plant establishment with changes in these structures, showing why the effects of planting extend beneath the surface.

Seven years is a meaningful interval for a project intended to last decades. Carbon storage develops as vegetation becomes established and soil structure changes, alongside the more immediate benefits of holding sand in place.

Water sets the limit on further greening

Mountain runoff supplies the moisture and nutrients that allow the current shrublands to survive. Water flowing down from the surrounding highlands sustains the planted belts around the Taklamakan, tying their carbon uptake to conditions beyond the desert itself.

Expansion farther into the interior would require reliable water supplies. Water is already increasingly scarce across the planet, and the desert's harsh interior places a firm constraint on how far planting can spread.

Climate change could also alter the water feeding the existing vegetation. Changes in snowfall or glacier melt may weaken the runoff that supports the planted margins, potentially weakening or destroying the carbon sink. The green belt depends on a continuing connection between mountain water and desert plants.

The project therefore combines two environmental tasks: maintaining vegetation in an extreme landscape and sustaining the water supply that keeps it alive. The 2024 completion of the encircling belt marked a planting milestone, while the vegetation's continued carbon uptake depends on its ability to survive those conditions.

A measurable contribution with a modest global scale

Even complete vegetation cover across the Taklamakan would capture only about 60 million tons of carbon dioxide annually, according to the estimate discussed by the researchers. That scenario covers the entire desert, an area roughly the size of Germany, rather than just its planted margins.

The estimated 60 million tons would equal about 10% of Canada's annual emissions. Against global annual emissions of 40 billion tons, it would amount to approximately 0.15%. The comparison shows the scale of the contribution that even an enormous desert planting effort could make.

King-Fai Li stressed that desert greening alone cannot solve the climate crisis. Its value also lies in understanding where carbon dioxide can be absorbed, how much can be captured and which environmental conditions make that possible.

The research brings together several benefits in the same landscape: protecting farmland, reducing erosion, stabilizing dunes and storing carbon. Those benefits are linked through the survival of vegetation, rather than through a technology that removes carbon dioxide independently of local growing conditions.

What the Taklamakan experience offers other dry regions

The Taklamakan project provides a rare example of desert greening sustained over decades where other efforts have failed. Consistent funding and labor allocation gave plants time to grow and soils time to accumulate organic carbon, creating changes that projects of shorter duration may struggle to achieve.

Yuk L. Yung, a professor of planetary science at Caltech and senior research scientist at NASA's Jet Propulsion Laboratory who helped write the study, described the finding:

We found, for the first time, that human-led intervention can effectively enhance carbon sequestration in even the most extreme arid landscapes, demonstrating the potential to transform a desert into a carbon sink and halt desertification.

Deserts long regarded as ecological dead zones can participate in carbon management when vegetation has the conditions it needs. The Taklamakan demonstrates that possibility through measurable plant activity, lower atmospheric carbon dioxide and changes in soil storage.

Wildfires are increasingly turning forests from carbon sinks into carbon sources across Europe, Latin America and North America. Against that backdrop, the Taklamakan offers an example of human intervention helping a landscape move in the opposite direction.

The green belt is a practical demonstration of a natural approach to carbon storage that depends on time, continuity and ecological understanding rather than advanced carbon removal technology. Its lessons remain tied to the limits identified by the research: scarce water, dependence on mountain runoff and a contribution modest in relation to global emissions.

The Essentials

• China launched the Three-North Shelterbelt Program in 1978, with a planned endpoint in 2050. The Taklamakan's vegetative encirclement was completed in 2024.
• More than 66 billion trees have taken root since the program began, while China's forest cover has risen from 10% in 1949 to more than 25% today.
• Satellite observations found carbon dioxide concentrations 1 to 2 ppm below surrounding areas over treated land. Seasonal concentrations fell from 416 ppm during dry periods to 413 ppm during the wet season.
• Hardy shrubs drive much of the carbon capture. Research along the southern desert edge found increased soil carbon and aggregate formation, with the largest gains after seven years.
• Mountain runoff supplies the water and nutrients sustaining the planted belts. Changes in snowfall and glacier melt could threaten that supply.
• Vegetation covering the entire desert would capture an estimated 60 million tons of carbon dioxide annually, about 0.15% of global annual emissions of 40 billion tons.