In forty-six years, China has planted 66 billion trees in its northern regions. The Gobi, which was gaining 10,000 km² per year in the 1980s, was retreating by more than 2,000 km² per year in 2022. Sandstorms have declined by more than 20% since 2000. And on November 28, 2024, the green belt of 3,046 km surrounding the Taklamakan Desert was completed. In January 2026, a study published in the Proceedings of the National Academy of Sciences established, for the first time, that plantations from the Three-North Shelterbelt Program have transformed part of the Taklamakan Desert into a measurable carbon sink visible from space. This milestone deserves careful examination, without either magnifying or diminishing it.

The Essentials

  • China claims 66 billion trees planted since 1978 as part of the Three-North Shelterbelt Program, across 13 provinces and autonomous regions.
  • The green belt of 3,046 km surrounding the Taklamakan, the most arid desert in Central Asia, was completed on November 28, 2024.
  • The Gobi, which was gaining 10,000 km² per year in the 1980s, was retreating by more than 2,000 km² per year in 2022. Sandstorms have declined by more than 20% since 2000.
  • Tens of millions of rural residents have directly benefited from the program, through the stabilization of agricultural lands and the development of a forestry and fruit-growing economy.
  • A PNAS study (January 2026) proves that these plantations constitute a carbon sink now detectable by satellite.
  • Real limitations exist: vulnerable monocultures, groundwater under pressure, and the objective of forest coverage by 2050 remains uncertain in the most arid zones.
  • The challenge for the coming years is to maintain the gains without exhausting the water resources that make them possible.

The Largest Forestry Project in Human History

The Three-North Shelterbelt Program was launched in 1978, the same year China began its economic reforms. The ambition matched the scale of the problem: sandstorms swept across northern agricultural lands each spring, reached Beijing, and pushed millions of hectares toward sterility. The program planned to plant trees across a 4,480-kilometer band traversing northeastern, northern, and northwestern China by 2050.

Forty-six years later, the quantitative results are unparalleled in the annals of reforestation. According to official Chinese data, 66 billion trees have been planted across 13 provinces and autonomous regions. More than 30 million hectares reforested. China today accounts alone for approximately one-quarter of the net global gain in vegetation measurable by satellite since 2000, according to analyses published in Nature Sustainability. Tens of millions of people have been lifted out of poverty thanks to the development of forestry and fruit-growing industries generated by the program.

The green belt of 3,046 km surrounding the Taklamakan Desert, completed on November 28, 2024 according to Chinese official media, is the most recent milestone of this program. The Taklamakan is the world’s second-largest sand desert, entirely hyperarid, nestled in the Xinjiang basin between the Tian Shan mountain range to the north and the Kunlun to the south. The idea of stabilizing its borders with a belt of trees and shrubs resistant to these conditions represents, on paper, geographic obstinacy. The first results prove that obstinacy has paid off.

Deserts Becoming Carbon Sinks

The scientific contribution of the January 2026 PNAS study is precise. Researchers analyzed carbon fluxes above reforested areas of the Three-North Shelterbelt Program by combining remote sensing data and biogeochemical models. Their conclusion: the plantations have generated a net carbon sink, measurable from space, in areas that previously constituted net emission sources through wind erosion and organic matter decomposition.

This result matters for a reason that extends beyond China. Large arid zones represent approximately 40% of the Earth’s land surface. They are traditionally excluded from climate mitigation models, considered too dry, too poor in organic carbon, too unstable to contribute to the global balance. The PNAS study shows that this exclusion deserves reconsideration. Hyperarid deserts, under conditions of intensive reforestation and irrigation, can shift toward the carbon sink side. The order of magnitude remains modest at the planetary scale, but the principle is established.

This climate dimension adds to already-documented benefits: reduction in the frequency and intensity of sandstorms — more than 20% fewer since 2000 — stabilization of agricultural lands, improvement in air quality in northern cities. The Gobi, which was engulfing 10,000 km² of land per year in the 1980s, was retreating by more than 2,000 km² per year in 2022. These are entire landscapes, entire rural communities that did not have to migrate. Agricultural yields that did not collapse. The annual value of ecosystem services generated by the program is estimated at 2,340 billion yuan by Chinese authorities — a figure whose methodology is debatable but revealing of the scale of transformations at stake.

It is useful to place this type of program alongside other environmental progress dynamics. Hannah Ritchie, in her analysis of global environmental data, recalls that long-term trends are often less catastrophic than dominant discourse suggests, but that this improvement is never automatic: it results from targeted public investments and political decisions sustained over time. The Three-North Shelterbelt is a canonical example. For further analysis of this framework, the book review of Hannah Ritchie’s Not the End of the World offers complementary insights.

The Limitations That Official Narrative Minimizes

The program is not exempt from serious criticism, and ignoring them would amount to confusing results with narrative.

The first limitation is ecological. A large part of the plantations relies on monocultures, notably poplars and spruces selected for their rapid growth and drought resistance. These species have their utility, but their vulnerability to disease and climate variations is documented. Entire zones have withered in Inner Mongolia and Shaanxi, as shown in monitoring studies published in Forest Ecology and Management. A screen of poplars is not a forest: its biodiversity is limited, its long-term resilience uncertain.

The second limitation is hydrological. In hyperarid regions like Xinjiang, vegetation cannot exist without water. Plantations bordering the Taklamakan depend heavily on irrigation, itself fed by fossil aquifers and the melting of Tian Shan glaciers. Yet the Tian Shan glaciers have lost between 20 and 30% of their volume since the 1960s, according to Chinese Academy of Sciences surveys. The green belt around the Taklamakan is, in part, built on a water resource whose future availability is uncertain. If glaciers continue to retreat and aquifers continue to deplete, portions of the belt could wither before reaching ecological maturity.

The third limitation is statistical. The regional forest coverage objective set by the Chinese government appears within reach when reading current figures. But these aggregate highly heterogeneous situations. Some northeastern provinces have exceeded 40% coverage; others, notably in Xinjiang and Inner Mongolia, remain well below 5%. The national average masks considerable regional disparities, and the easiest gains have already been made. Future gains will be on more hostile terrain, with higher marginal costs in water and maintenance.

What Scientists and Engineers Have Built

The scale of the program requires that its concrete mechanisms be described. This is not a communication operation: the 66 billion trees planted represent a mobilization of human and technical resources without precedent in the history of reforestation.

Forest engineers from the State Forestry and Grassland Administration have developed, over the decades, implantation techniques adapted to the extreme conditions of Central Asian deserts. Staggered planting, which reduces evapotranspiration compared to straight rows. Woven straw barriers driven into the sand, which stabilize dunes before shrub implantation. The use of Haloxylon ammodendron (saxaul) and other xerophytes capable of surviving on less than 100 mm of annual precipitation. These techniques have been tested in experimental stations, refined through trial and error, then deployed at large scale under difficult logistical conditions.

The financial cost of the program over its total duration is not consolidated in a single accessible source. Figures advanced by different Chinese institutional sources vary and are not independently verifiable. What is documented is that the program has mobilized hundreds of thousands of rural workers, teams from the People’s Liberation Army in certain provinces, and a network of state nurseries covering the entire target area.

Chinese and foreign researchers played a determining role in monitoring and adjusting the program. Teams from the Desert Research Institute of the Chinese Academy of Sciences, based in Lanzhou, have published hundreds of studies on vegetation dynamics, erosion, and water balance in reforested zones. This accumulation of knowledge is what allowed for correction of errors in the first decades, notably excessive reliance on species unsuited to sandy substrates, and improvements in plant survival rates.

The International Comparison Often Missing

The Three-North Shelterbelt is often presented as a Chinese exception, made possible by a regime capable of imposing long-term objectives independently of electoral cycles. This reading is partially correct but reductive.

Other countries have conducted ambitious reforestation programs under different institutional conditions. Ethiopia planted 350 million trees in a single day in 2019, as part of its Green Legacy program. Pakistan launched a ten-billion-tree program under Imran Khan’s government. The African Great Green Wall, a pan-African initiative coordinated by the African Union, aims to restore 100 million hectares of degraded land across the Sahel by 2030. These programs have contrasting results, but they show that large-scale reforestation ambition is not the monopoly of authoritarian regimes.

What distinguishes the Chinese program is not so much its political model as its duration. Forty-six years of continuity in an ecological infrastructure program represents an institutional rarity, regardless of governance system. Most ambitious reforestation programs fade after a change of government or budgetary reorientation. China’s capacity to maintain course over several decades, through revisions and trajectory corrections, remains its main asset.

On the long-term carbon balance, a parallel is worth mentioning: industrial and environmental policies that produce measurable results often share the same characteristics, whether in energy, industry, or land management. The wolf returning to 34 European countries thanks to coordinated protection policies over several decades offers another example of what institutional perseverance can accomplish in environmental matters, as analyzed in this article on the wolf’s return to Europe.

What the Sand Still Hides

The official objective of regional forest coverage by 2050 is both near and distant. Near in terms of raw figures: the gains achieved since 1978 are spectacular on paper. Distant in terms of geophysical reality: remaining gains must be made in the most hostile zones, where water is scarcest.

The real challenge for the coming decades is not reaching an additional percentage, but maintaining what has already been gained in a context of climate change reducing precipitation and accelerating evaporation in arid regions. Regional climate modeling studies, notably those produced by the CORDEX-EA program, suggest that the arid zones of northwestern China could experience intensifying drought by 2050 under high-emission scenarios. In this context, keeping tens of millions of hectares of plantations alive will require increasing water resources as the climate warms.

Chinese foresters are aware of this. Since the 2010s, new plantations have favored low-water-consumption species, and several provinces have imposed moratoriums on planting in zones where groundwater shows signs of overexploitation. This is a significant trajectory adjustment, signaling that the program learns from its mistakes. But this adjustment does not resolve the central question: how well can vegetation installed over the past forty years maintain itself without artificial irrigation if hydrological conditions deteriorate?

The honest answer is that ecologists do not yet know with certainty. The oldest plantations, notably in Shaanxi and Ningxia, are beginning to show signs of partial autonomy, with high survival rates without artificial input. Those implanted in the most arid Xinjiang zones remain dependent. The gap between these two situations is the real indicator to watch in the years ahead.

The green belt of 3,046 km around the Taklamakan is an accomplished fact, measurable and documented by independent satellites. What PNAS researchers demonstrated in January 2026 goes beyond symbolism: zones classified as net carbon sources have become sinks. This shift is real. It does not erase future hydrological tensions, but it establishes that large-scale reforestation in arid zones is possible, and that it produces measurable climate effects. The question that will remain open for the coming decades is not whether China has managed to push back a desert. Data confirms it has partially succeeded. The question is whether the 66 billion trees planted will survive the climate change that these same deserts modestly help mitigate.

Sources

  1. Zheng et al. (2025), analysis of the Three-North Shelterbelt Program, Science China Life Sciences: https://link.springer.com/article/10.1007/s11427-024-2705-4
  2. PNAS study, January 2026 — Taklamakan carbon sink: https://www.pnas.org/doi/10.1073/pnas.2523388123
  3. Satellite analyses of global vegetation, Nature Sustainability — Chen et al. 2019 (25% global vegetation): https://www.nature.com/articles/s41893-019-0220-7
  4. State Forestry and Grassland Administration, annual reports of the Three-North Shelterbelt Program
  5. Desert Research Institute, Chinese Academy of Sciences, Lanzhou (publications available on the Academy portal)
  6. CORDEX-EA program, regional climate modeling for East Asia
  7. Completion of the Taklamakan green belt — SCMP, November 2024: https://www.scmp.com/news/china/science/article/3288549/great-wall-taklamakan-china-surrounds-its-largest-desert-giant-green-belt
  8. Chinese Academy of Sciences — 40-Year TNSP Assessment (2018): https://english.iae.cas.cn/News2017/201812/t20181228_202987.html
  9. Wikipedia — Tian Shan glaciers (27% loss since 1961): https://en.wikipedia.org/wiki/Tian_Shan
  10. UN DESA — Three-North Shelterbelt Program (14.95% objective): https://sdgs.un.org/partnerships/three-north-shelterbelt-program
  11. PNAS — Critical commentary Gao et al. on the Noor 2026 study: https://www.pnas.org/doi/10.1073/pnas.2607916123