Estimating spatiotemporal pumping amounts using multiple signal decomposition methods

Abstract

Effective water resource management requires accurate estimation of the volume of water withdrawn through anthropogenic groundwater pumping. However, insufficient pumping information makes accurate estimation difficult. To address this issue, this study proposes a novel method for estimating pumping volumes that does not rely on pumping information. The proposed method uses high-resolution groundwater level (GWL) data from densely monitored wells and integrates Hilbert–Huang transform and empirical orthogonal function analysis to extract pump-associated signals, which can then be used to estimate total pumping. Our method was validated by analyzing GWL data from monitoring stations on the Choshui River alluvial fan in Taiwan. The results indicate the estimated pumping amount of 1.71×109 – 2.05×109 m3/year consistent with previous estimates obtained using survey and physics-based methods ranging from 9×108 m3/year to 2.2×109 m3/year. In the areas with abundant groundwater head observations, the proposed method can efficiently estimate pumping rates with a high space–time resolution, and may serve as a useful reference for groundwater management.