Misattributed biospheric carbon sink to the Three-North Program in the Taklamakan Desert may lead to an impractical policy: A commentary

[Geoengineering News]

https://www.pnas.org/doi/10.1073/pnas.2607916123

Authors: Tian Gao, Jiaojun Zhu, Xiao Zheng, Lining Song, Yirong Sun, Dexiong Teng, and Xinhua Zhou

12 June 2026

https://doi.org/10.1073/pnas.2607916123

Abstract 

Afforestation is a widely accepted practice to build terrestrial carbon sinks, but its applications are limited by climatic conditions (1, 2). Recently, Noor et al. (2026) demonstrate that ecological restoration (China’s Three-North Protective Forest Program, TNP) is transforming hyperarid environment into a carbon sink, providing an encouraging projection for carbon sink built up through afforestation, and extrapolating it to the entire Taklamakan Desert (3). However, given extremely limited effective water to plants in the region, potential ecohydrological processes derived from nonfield data could easily misinterpreted real field cases.

Hydrothermal conditions are the fundamental constraint, but these are not adequately addressed by Noor et al. (3). Based on precipitation, wet season (Jul–Sep) and dry season (Dec–Feb) are defined for analysis. However, the Taklamakan Desert features hot, dry summers, and cold, dryer winters, with annual precipitation less than 80 mm, while annual potential evaporation more than 2,500 mm; i.e., this hyperarid region experiences a sole “dry season” year-round without moist seasons. The defined dry season is exactly the mid of winter (nongrowing season), with an average temperature of ca. −5 °C, when plant activity ceases at all. Therefore, the comparison of vegetation activity between the dry and wet seasons has no physiological basis. Additionally, given the huge summer evaporation potential (4, 5), the precipitation is apparently insufficient to support plant growth. Furthermore, the authors attributed the summer decline in atmospheric CO2 to regional afforestation. However, such declines are a well-known seasonality driven by widespread vegetation growth outside this desert across the Northern Hemisphere. Due to the low afforestation covers, vegetation contribution is locally limited. Observed 3-ppm (3) decline falls within 4 to 6 ppm seasonal fluctuations of background atmospheric CO2 (6, 7), making it baseless to be inferred as the results from the Taklamakan afforestation.

Afforestation can build carbon sinks, but extrapolating its potential of 58.7 Mt CO2 y−1 into the entire Taklimakan Desert is unrealistic (Fig. 1). Regional evaporation leaves little rainfall available for summer plant uptake, and almost all of artificial vegetation relies on irrigation. Sustaining large-scale vegetation has to persistently pump groundwater. Hydrological evidence indicates a long-term decline of 8.6 mm y−1 in regional groundwater (8). Therefore, achieving carbon sinks would depend on large-scale water-diversion projects. Moreover, arid ecosystems exhibit low productivity and stability, and climate warming-induced drought may further cause vegetation decline and even shift the region to carbon sources. So, the carbon budget and afforestation cost-effectiveness should be reasonably evaluated.

Source: PNAS