Estimating the impact of irrigation and groundwater pumping on regional hydroclimate using an Earth System Model

Irrigation is a significant anthropogenic forcing to the Earth system, altering water and heat budgets at the land surface and inducing changes in regional hydro-climate conditions across various spatiotemporal scales. These impacts of irrigation are expected to intensify in the future due to growing food demand and the pervasive effects of climate change. Therefore, it is imperative to better understand its nature, extent, and mechanisms through which irrigation affects the Earth system. However, despite its increasing importance, irrigation remains an emerging component in Earth system modeling community, necessitating further advancements in modeling approaches and a deeper understanding.

Our research aims to improve the quantitative understanding of how irrigation and groundwater use, as anthropogenic drivers, affect regional climate and environmental changes. To achieve this, we developed an enhanced Earth system modeling framework based on MIROC-ES2L (Hajima et al., 2020, GMD), integrated with hydrological human-activity modules (Yokohata et al., 2020, GMD). This framework enables simulations of coupled natural-human interactions, including hydrological dynamics associated with irrigation processes. Using this Earth system model, we carried out numerical experiments at T85 spatial resolution with an AMIP-style setup. Our large ensemble simulations allow statistical quantification of irrigation impacts, statistically distinguishing them from uncertainties arising due to natural variability.

Our investigation identified specific regions and seasons where irrigation exerts notable influences on regional hydro-climate. In particular, our results reveal substantial disparities—comparable to or exceeding inter-annual variability—between simulations with and without irrigation processes, especially in heavily irrigated regions such as Pakistan and India. Our model demonstrates that artificially wet soils due to irrigation alter the land surface hydrological balance, which consequently impacts the overlying atmosphere. However, significant uncertainties remain in the impact estimates for several variables in some regions, even those heavily irrigated, including the central United States and eastern China. This highlights the necessity of employing appropriate statistical approaches to evaluate irrigation impacts, accounting for inherent natural variability.

Additionally, our study estimates regional variations in the contributions of groundwater and surface water use to irrigation impacts. Our estimate indicates that approximately two-fifths of global irrigation water depend on groundwater resource, while this groundwater dependency ratio may still be underestimated. By emphasizing the importance of understanding regional and seasonal characteristics, our study underscores the importance of comprehending the complex interactions between irrigation-related human activities and the Earth's climate system. Nevertheless, we may still underestimate the full impacts of irrigation because irrigation water demand estimated by our coupled simulations is lower than that derived from preceding offline simulations or reported statistics. In this presentation, we will discuss this challenge as well.