Global greening refers to the widespread increase in vegetation cover, leaf area, and productivity observed via satellites (using metrics like NDVI, LAI, or GPP) since the early 1980s. It is one of the clearest large- scale responses of the biosphere to rising atmospheric CO₂, climate change, and human land management.

Satellite records (AVHRR, MODIS, VIIRS, and harmonized products) show consistent global greening of vegetated land since ~1982. Roughly 25- 50% or more of vegetated areas have shown significant increases in leaf area, equivalent in some earlier analyses to adding leaf area on the scale of two continental United States.

The trend has persisted and, in recent years, strengthened. Global mean vegetation greenness set records in 2024 and again in 2025.

In 2024, ~67.7% of vegetated land surfaces showed greening (positive anomalies), raising the 2000- 2024 trend.

In 2025, ~68.2% of vegetated surfaces greened, further steepening the trend (to ~8.5 × 10⁻⁴ yr⁻¹ for the 2000–2025 period in one analysis). Herbaceous ecosystems (grasslands ~72%, croplands ~78%) dominated the recent signal.

Greening is not uniform: some regions experience browning due to drought, heat, deforestation, or other stressors (e.g., parts of the Amazon, southern Africa, or eastern Siberia in certain years). Overall, greening far outweighs browning at the global scale.

CO₂ fertilization is widely identified as the dominant global driver. Higher CO₂ boosts photosynthesis and improves water- use efficiency (plants partially close stomata, losing less water). Attribution studies using models and statistical approaches commonly attribute ~70% (or higher, e.g., ~83% in one 1982- 2021 analysis) of the greening to rising CO₂.

Other important factors, varying by region:

Climate — Warming extends growing seasons (especially autumn phenology in extratropical forests) and benefits high- latitude/Arctic areas; precipitation increases drive recent hotspots in grasslands and croplands.

Land use and management — Agricultural intensification, irrigation, fertilization, and afforestation are major contributors, especially in China and India (together accounting for a large share of new vegetation since 2000). Human activities can outweigh pure CO₂ effects in managed drylands and croplands.

Nitrogen deposition and other factors play smaller roles globally.

In drylands specifically (linking to the prior discussion of aridity and CO₂), greening is widespread despite rising temperatures and aridity indices in many areas. CO₂- driven water savings allow more vegetation growth for a given amount of moisture; human management (e.g., cropland expansion) also contributes substantially. Some studies find greening in a majority of drylands, with net productivity gains.

Regional patterns and recent highlights

• China and India — Major greening hotspots due to tree planting/afforestation and intensive farming.
• Northern mid- to- high latitudes and Arctic — Strong greening from warming and longer growing seasons; some polar regions (e.g., Antarctic Peninsula, Svalbard) also show expansion of vegetation.
• Grasslands, savannas, and croplands — Prominent in recent record years (Eurasia, tropics, southern Hemisphere).
• Tropical forests — More mixed signals; some recovery or greening, but also browning linked to droughts/El Niño and deforestation.
• Drylands — Overall greening exceeds desertification in many assessments.

Phenology contributes meaningfully: lengthened autumn growing seasons drive a substantial portion of greening in extratropical forests, while maximum greenness increases dominate in non- forests (aided by CO₂ and land use).

Greening enhances the terrestrial carbon sink (helping offset some fossil- fuel emissions), increases evaporative cooling in places, and can improve water- use efficiency. However:

• It does not fully counteract climate impacts (e.g., dryland expansion risks, extreme heat/drought stress, or biodiversity losses).
• Increased vegetation can sometimes raise water demand or alter albedo.
• Resilience may decline in some greening humid regions even as greenness rises.
• Models often under- or over-estimate certain regional drivers (especially human land management in drylands).
• Trends can accelerate or slow with interannual climate variability (e.g., El Niño effects).

In summary, Earth has been getting greener for decades, driven primarily by CO₂ fertilization globally, with strong regional contributions from climate and human activities. Recent years (2024- 2025) set new records, particularly in grasslands and croplands, reinforcing the multi- decadal signal. This continues even as other climate stressors intensify, illustrating the complex, dual role of rising CO₂. Ongoing satellite monitoring and improved models remain essential for tracking future changes.

A new study finds that rising atmospheric CO₂ is expected to slow (but not stop) the expansion of Earth’s drylands this century, mainly because plants partially close their stomata and lose less water via transpiration.

The research, published in Environmental Research Letters (Cai et al., 2026), updates global dryland mapping and future projections using climate models from CMIP6 and historical meteorological data (1960- 2023).

Drylands are defined by the aridity index (AI = precipitation / potential evapotranspiration, or PET), with AI < 0.65 classifying an area as a dryland.

Key findings:

Current extent (baseline 1994– 2023, excluding Antarctica): Drylands cover about 38.58% of Earth’s land surface. Subtypes break down roughly as semi-arid (~14.72%), arid (~11.24%), dry sub- humid (~6.35%), and hyper- arid (~6.27%).

Future projections (2071- 2100): Under different greenhouse-gas emissions scenarios, drylands are projected to expand modestly to 39.31- 40.28% of land area. Hyper- arid zones show the largest relative increase. This equates to a potential gain of up to ~2.37 million km² (roughly the size of Algeria).

CO₂’s moderating role: Many earlier projections did not fully account for how higher CO₂ reduces plant stomatal conductance. This lowers water loss through transpiration, which modifies PET and yields lower aridity estimates than conventional approaches. The authors note that ignoring this effect can inflate aridity projections and shift the location of hotspots. With the CO₂- adjusted PET framework, expansion is noticeably smaller than some prior estimates (e.g., one older study suggested up to ~56% coverage by 2100).

Regions expected to see the most pronounced shifts include Australia (the strongest changes), Brazil, and parts of Africa. Expansion into formerly more humid areas could intensify water scarcity, ecosystem stress, and desertification risks, with implications for sustainable development.

This physiological response of plants to elevated CO₂ is a well- documented process that improves water-use efficiency and has been observed in experiments and models; the new work incorporates it more explicitly into global aridity- index projections.

Note that aridity- index metrics focus on atmospheric demand relative to precipitation and do not fully capture all ecohydrological or vegetation responses (e.g., actual soil moisture, plant growth, or greening trends observed in some drylands).

Other studies using different indices have sometimes projected little or no net global dryland expansion once plant physiological effects are considered more comprehensively.

CO₂-altered evapotranspiration moderates future global dryland expansion under climate change.

Rising atmospheric CO₂ reduces stomatal conductance, which lowers potential evapotranspiration (PET). When this physiological effect is included in the aridity index (AI = precipitation / PET), projected expansion of global drylands is moderated (smaller and slower) compared with conventional calculations that ignore the CO₂ effect on PET. The spatial pattern of drying and wetting hotspots also shifts.

Key results:

Current baseline (1994- 2023): Drylands (AI < 0.65) cover 38.58% of global land area (excluding Antarctica). Semi- arid zones form the largest share; hyper- arid zones are concentrated in places such as Algeria, Libya, and Saudi Arabia.

Historical trend (1960- 2023): Dryland extent increased overall at a rate of 3.91 × 10⁴ km² yr⁻¹, though the rate of expansion has slowed in recent decades.

Future projections (2071- 2100) with CO₂-modified PET (CMIP6 multi-model ensemble):

• SSP1-2.6: ~39.31%
• SSP2-4.5: ~39.68%
• SSP5-8.5: ~40.28%
  Absolute expansions relative to the baseline: 1.09, 1.58, and 2.37 million km², respectively. Hyper-arid zones contribute most to the increase.

Methods highlights:

AI classification follows UNEP: hyper- arid (AI < 0.05), arid (0.05-0.2), semi- arid (0.2- 0.5), dry sub- humid (0.5- 0.65).

PET calculated with a CO₂- modified Penman- Monteith equation that accounts for reduced stomatal conductance at higher CO₂. Parallel runs without the CO₂ term allow direct comparison.

Historical data: CRU precipitation and ERA5 meteorology (1960- 2023). Future: 19 CMIP6 models under three SSPs, bias-corrected using the 2015- 2023 overlap period.

Hotspots defined by significant AI trends (Theil- Sen and modified Mann- Kendall, p < 0.01).

Regional patterns:

• Persistent or emerging drying hotspots: Australia, Brazil, northern Africa.
• Possible wetting trends: parts of Northwestern China.
• Transition example: parts of Central Australia may shift from wetting to drying hotspots.
• Scenario differences are strongest along dryland margins (Brazil, North Africa- West Asia, Australia, Sahe- East Africa, northern China).

Conventional aridity projections that neglect CO₂ effects on PET tend to overestimate future dryland expansion and can mislocate hotspots.

The physiological response (stomatal closure → lower transpiration demand) provides a partial brake on aridification.

The authors note that full risk assessments should still incorporate land- use change, actual water resources, and ecosystem responses beyond the climate- based AI.

The paper is open access, so the full text, figures, and supplementary material are freely available at the link.

Published: Environmental Research Letters, Volume 21, Number 17 (2026)

DOI: 10.1088/1748-9326/ae9b29 (open access)

Authors: Yifei Cai, Feng Wang^*, Yude Pan, Xubin Pan and Qi Lu

Abstract

Drylands are highly vulnerable to climate change, yet their recent spatial redistribution and future evolution remain insufficiently understood. Conventional projections often neglect CO₂ effects on potential evapotranspiration (PET), which may yield higher aridity estimates and alter aridity-change hotspot patterns. To address this gap, this study maps the current global distribution of drylands, quantifies historical and future changes in dryland extent, and identifies drying and wetting hotspots under different emission scenarios using a CO₂-modified PET framework. Using 1994–2023 as the baseline period, global drylands currently occupy approximately 38.58% of the global terrestrial area. From 1960 to 2023, dryland extent showed a fluctuating but overall increasing trend, with an expansion rate of 3.91 × 10⁴ km² yr⁻¹. Future projections indicate continued but slower expansion through the 21st century compared with the historical trend. By the end of the 21st century, drylands are projected to cover approximately 40.28% of the global terrestrial area under SSP585. Accounting for rising CO₂ effects moderates projected dryland expansion and alters the spatial pattern of drying and wetting hotspots compared with approaches that neglect CO₂ effects on PET. Regionally, Australia, Brazil, and northern Africa are projected to remain or emerge as persistent drying hotspots, whereas parts of Northwestern China may maintain wetting trends. Parts of Central Australia may undergo a wetting-to-drying hotspot transition, indicating the emergence of new drying hotspots. These findings provide a climate-based indication of regions where dryland risks may be spatially reconfigured. Future dryland risk assessments should further incorporate land-use and water-resource conditions to better evaluate adaptation and management implications.