Wout M. van Dijk , Alex Densmore, Christopher Aiden-Lee Jackson , Jonathan Mackay, Suneel Joshi, Rajiv Sinha, Shashank Shekhar, Sanjeev Gupta

Unsustainable exploitation of groundwater in northwestern India has led to extreme but spatially variable depletion of the alluvial aquifer system in the region. Mitigation and management of groundwater resources require an understanding of the drivers behind the pattern and magnitude of groundwater depletion, but a regional perspective on these drivers has been lacking. The objectives of this study are to (1) understand the extent to which the observed pattern of groundwater level change can be explained by the drivers of precipitation, potential evapotranspiration, abstraction, and canal irrigation, and (2) understand how the impacts of these drivers may vary depending on the underlying geological heterogeneity of the system. We used a transfer function-noise (TFN) time series approach to quantify the effect of the various driver components in the period 1974-2010, based on predefined impulse response functions (theta). The dynamic response to abstraction, summarized by the zeroth moment of the response M0, is spatially variable but is generally large across the proximal and middle parts of the study area, particularly where abstraction is high but alluvial aquifer bodies are less abundant. In contrast, the precipitation response is rapid and fairly uniform across the study area. At larger distances from the Himalayan front, observed groundwater level rise can be explained predominantly by canal irrigation. We conclude that the geological heterogeneity of the aquifer system, which is imposed by the geomorphic setting, affects the response of the aquifer system to the imposed drivers. This heterogeneity thus provides a useful framework that can guide mitigation efforts; for example, efforts to decrease abstraction rates should be focused on areas with thinner and less abundant aquifer bodies.