China’s Great Green Wall (officially the Three-North Shelterbelt Program or Three-North Protective Forest Program) is a massive, multi-decade ecological engineering project launched in 1978 and planned to run until 2050.

It aims to combat desertification, reduce sandstorms and soil erosion, protect farmland, and increase forest cover across northern China (Northwest, North, and Northeast regions), covering about 4–4.5 million km²—nearly half the country’s land area.

It is not a single continuous wall of trees, but a network of shelterbelts, forest patches, shrublands, and grasslands forming protective barriers around deserts (including the Taklamakan and Gobi), along roads, canals, fields, and settlements.

The program combines engineering, biological, and technological approaches, adapted to local conditions (rainfall, soil, wind, water availability). Work is done in phases and focuses on “edge-locking” deserts rather than greening the entire interior.

In short, the Great Green Wall works by systematically stabilizing sand first, then establishing hardy vegetation with targeted water support and long-term management, creating living barriers that slow desert advance and gradually improve local ecosystems. It relies on massive state coordination, adaptive local techniques, and sustained effort over decades.

_____________________________________________________________________________________

China’s decades-long tree- and shrub-planting campaign around the Taklamakan Desert (part of the broader Three-North Shelterbelt Program, or “Great Green Wall”) has helped turn the desert’s margins into a measurable carbon sink, according to satellite-based research published in early 2026.

The Taklamakan (also spelled Taklimakan), in northwestern China’s Xinjiang region, is one of Earth’s largest and driest deserts—roughly 337,000 km² (about the size of Germany or slightly larger than Montana), often called a “biological void” or “Sea of Death” due to its hyper-arid conditions, shifting dunes, and historically minimal vegetation.

China began large-scale afforestation efforts in 1978 primarily to halt desert expansion, reduce sandstorms, stabilize dunes, protect farmland/infrastructure, and increase northern forest cover (with goals tied to broader national and UN forest targets).

In late 2024, authorities announced completion of a continuous ~3,046 km (about 1,892-mile) green belt of trees and shrubs encircling the desert, alongside plantings along the Tarim Desert Highway.

Tens of billions of trees have been planted across northern China as part of the wider program (figures around 66 billion are commonly cited for the north overall).

A 2026 study in the Proceedings of the National Academy of Sciences (PNAS), led by researchers including King-Fai Li (UC Riverside) and involving collaborators linked to NASA data and other institutions, analyzed multi-year satellite observations (NASA’s Orbiting Carbon Observatory for CO₂ and MODIS for vegetation/greenness/fluorescence) plus related models.

Key findings include:

• Strong seasonal dynamics: During the wet season (July–September), increased precipitation supports higher vegetation cover and photosynthesis along the planted margins, drawing down atmospheric CO₂ by roughly 3 ppm relative to dry-season baselines.
• Long-term trends: Rising vegetation cover and photosynthetic activity (solar-induced fluorescence) over ~25 years, with a strengthening net ecosystem exchange (NEE) indicating net CO₂ uptake, concentrated at the desert edges and aligning temporally with the planting program.
• The planted shelterbelt region shows average annual uptake on the order of ~1.74 t CO₂ per hectare. Earlier related estimates for broader periods (e.g., ~2004–2017) indicated net absorption (uptake exceeding release). Hypothetical full-desert coverage at similar rates could sequester ~50–60 million metric tons of CO₂ per year—meaningful locally but only a small fraction of global emissions.

Additional reported co-benefits include reduced wind erosion and sandstorm intensity/frequency in some areas, plus protection of nearby agriculture. The project used drought- and salt-tolerant species suited to the harsh conditions, often with irrigation support from runoff or systems.

Scientists describe this as evidence that large-scale ecological restoration can enhance carbon sequestration even in extreme arid environments—the first clear quantified example of human intervention turning hyper-arid desert margins into a net biospheric carbon sink in this way.

It was not primarily designed as a climate project but has produced this secondary effect. Limitations remain: success is mainly at the rims (not the vast interior dunes), survival rates and water use have faced past challenges in such projects, the overall climate impact is modest on a global scale, and long-term sustainability depends on continued management, water availability, and monitoring.

Separate studies have examined related shelterbelts (e.g., along the desert highway) and also noted increased carbon storage/NPP.

In short, the core of the headline is accurate based on the recent peer-reviewed satellite analysis: the planting effort has created measurable carbon uptake where little biological activity existed before.

_____________________________________________________________________________________

Human-induced biospheric carbon sink: Impact from the Taklamakan Afforestation Project

“Human-induced biospheric carbon sink: Impact from the Taklamakan Afforestation Project” is a research article published in the Proceedings of the National Academy of Sciences (PNAS) on January 20, 2026 (Vol. 123, No. 4, e2523388123; DOI: 10.1073/pnas.2523388123).

Lead authors include Salma Noor and Xun Jiang (University of Houston), with co-authors such as King-Fai Li (University of California, Riverside), Le Yu (Tsinghua University), Yuk L. Yung (Caltech / NASA Jet Propulsion Laboratory), and others. Funding included support from NASA’s Earth Sciences Division.

The Taklamakan Desert (≈337,000 km² in northwestern China), long viewed as a “biological void” due to extreme aridity (average annual rainfall <50 mm, >95% shifting sands), is being transformed into a carbon sink through large-scale ecological restoration under China’s Three-North Shelterbelt Program (started 1978; also called the Great Green Wall).

Using satellite data (MODIS NDVI for vegetation cover, TROPOMI solar-induced fluorescence/SIF for photosynthetic activity, OCO-2 for atmospheric CO₂) plus ground-based land-cover records and the MiCASA biospheric model (for net ecosystem exchange/NEE, net primary production, and respiration), the authors report:

• Seasonal dynamics: Wet-season (July–September) precipitation rises to ~16.3 mm/month, boosting vegetation and photosynthesis. This draws down atmospheric CO₂ by ≈3 ppm relative to dry-season levels (dry-season mean ≈416.3 ppm).
• Long-term trends (aligned with the planting program and concentrated along desert margins):
  • Vegetation cover (NDVI): +6.8 × 10⁻⁴ per year.
  • Photosynthetic activity (SIF): +6.1 × 10⁻³ W/m²/sr/µm per year.
  • Strengthening net CO₂ uptake (NEE trend): −5.2 × 10⁻¹² kg/m²/s per year.
• Annual mean NEE over the green shelterbelt (rim) region: −1.5 × 10⁻⁹ kg C m⁻² s⁻¹, equivalent to ≈−1.74 t CO₂ ha⁻¹ y⁻¹.
• Hypotheticals: Full-desert coverage at the same rate could remove ≈58.7 million metric tons CO₂ per year; scaling across all of China could remove ≈1,670 million metric tons CO₂ per year (roughly 14% of China’s annual emissions, though this is an extreme extrapolation).

The paper positions this as direct evidence that human intervention can enhance carbon sequestration in hyper-arid landscapes, halt desertification, and serve as a model for nature-based climate solutions in drylands (which are often overlooked in global carbon-cycle studies). It notes prior findings of modest abiotic carbon sinks in the desert sands themselves but emphasizes the biospheric (vegetation-driven) contribution from afforestation.

It highlights the Taklamakan’s emerging role in the global carbon cycle, shows that extreme deserts are not beyond recovery, provides benchmarks for Earth-system models, and supports carbon-management strategies focused on drylands.

In June 2026, PNAS published critical letters and an authors’ reply:

• Commentators (Nan Xu; Tian Gao, Jiaojun Zhu et al.) questioned the statistical significance of the NEE trend (signal-to-noise ≈1), potential contribution of abiotic processes (not fully resolved by the MiCASA model) to the seasonal CO₂ drawdown, the practicality of extrapolating rim rates to the entire desert or China (water scarcity, heavy reliance on irrigation, high costs), and whether the observed changes could be over-attributed to the Three-North Program versus natural variability or other factors. They noted that full-desert greening is unrealistic and could mislead policy.
• The authors replied that trends are supported by multiple independent datasets (NDVI, SIF, ground green-area increases of ≈162 km²/y), that local trends along the rim have higher significance, that seasonal NEE differences confirm biospheric (photosynthetic) influence, and that the work demonstrates success specifically at the margins as a proof-of-concept rather than a call for unrestricted expansion.

The study relies on remote-sensing and modeling rather than extensive new ground flux-tower measurements across the vast desert, which is a common limitation for such large-scale analyses.

Overall, it provides observational evidence of greening and enhanced carbon uptake linked to the multi-decade planting effort, while the critiques underscore uncertainties in attribution, magnitude, and scalability.

Published: Proceedings of the National Academy of Sciences (PNAS)

DOI: 10.1073/pnas.252338812

Authors: Salma Noor, Xun Jiang, Xinyue Wang and Yuk L. Yung

Significance

This study highlights the Taklamakan Desert’s emerging and previously underappreciated role in the global carbon cycle. By revealing how human-led afforestation can transform hyperarid landscapes into functioning carbon sinks, it demonstrates that even the most extreme deserts are not beyond ecological recovery and can be managed for carbon storage, thus helping to mitigate climate change. The observed greening trend suggests a significant climate impact, as increased CO₂ uptake in arid zones could influence regional and even global climate dynamics. This work provides critical constraints and benchmarks for refining Earth system models and supports the development of more effective carbon management strategies, particularly in dryland regions that have long been overlooked in global reforestation and carbon removal efforts.