Estimation of Global Irrigation Water Use by the Integration of Multiple Satellite Observations

Quantification of the global irrigation water use (IWU) is crucial to understanding the anthropogenic disturbance of the natural hydrological cycle and optimal agricultural water management. However, it is challenging to obtain time series data with the conventional survey-based approach, while the current satellite-based IWU estimations are subject to data gaps and the model structure. In this paper, we propose a comprehensive framework to couple the different processes associated with irrigation and integrate multiple satellite observations to estimate the global IWU. The ensemble IWU estimate demonstrates an improved performance when compared to the IWU obtained from individual satellite observations. The results show reasonable correlation with the survey-based irrigation water withdrawal in states of the US (bias = -0.42 km(3)), provinces of China (bias = -3.10 km(3)), and country statistics from the Food and Agriculture Organization (bias = -10.84 km(3)). Large amounts of IWU are apparent in India, China, the US, Europe, and Pakistan, making up >70% of the global IWU. A general underestimation of IWU is found both in this work and previous studies, due to the coarse resolution and asynchronism of the various satellite products, the changes in irrigated areas, and the deficiency in detecting irrigation events under the case of saturated soil moisture. Nevertheless, we demonstrate advantages in integrating multiple satellite observations to reduce the uncertainty in estimating global IWU. However, additional efforts are needed to produce high-quality and finer spatiotemporal resolution satellite-based products, to further improve the accuracy of the global IWU estimation.

Automated summary

This study proposes a comprehensive framework to estimate global irrigation water use (IWU) by integrating multiple satellite observations. The results show that the integrated IWU estimate outperforms individual satellite observations, but there is a general underestimation of IWU globally. Further efforts are needed to improve the accuracy of global IWU estimation by producing high-quality and finer spatiotemporal resolution satellite-based products.