Understanding Time-Lagged Causal Effects of Environmental Drivers on Vegetation Carbon Uptake

Global warming, primarily caused by rising levels of carbon dioxide (CO₂) concentration in the atmosphere, poses a severe threat to the Earth. Perhaps the most effective and important strategy is achieving carbon neutrality. While a reduction in carbon emissions is a go-to approach for achieving carbon neutrality, enhancing or conserving carbon uptake or sequestration through land vegetation—via photosynthesis, commonly known as gross primary production (GPP)—is equally crucial in reducing atmospheric CO₂ concentrations and mitigating global warming. Beyond direct and indirect human interventions, environmental drivers such as temperature, rainfall, vapor pressure deficit (VPD), solar radiation, and soil moisture (SM) significantly influence vegetation dynamics, influencing carbon sequestration. Therefore, identifying the interactions among various environmental drivers and GPP is vital for enhancing the accuracy of carbon budget estimations and refining climate-carbon feedback models to better project land-based carbon sinks under climate change. Despite this fact, there are very limited studies, especially in India, to investigate these relationships. Consequently, to ascertain the links between environmental drivers and GPP, we have adopted the data-driven causal technique rather than the conventional correlation approach. The results for average conditions in India indicate that root-zone SM has a strong connection with GPP, while rainfall presents a weaker connection with a larger lag than SM. It emphasizes the significance of irrigation in India's vegetation, as the country's land is predominantly occupied by shallow-rooted crops. Mean VPD demonstrates a moderate influence; in contrast, both mean temperature and net solar radiation show weaker effects, with almost equal lag time.  The overall findings reveal the varying influences of different environmental variables on the GPP, offering crucial insights for improved regional land-atmosphere modeling to better replicate the carbon balance, thereby reducing the uncertainty associated with current estimates.

Keywords: Causal inference, climate-vegetation interactions, gross primary production (GPP)