Investigating aquifer properties in the Himalayas through stream flow response to the 2015 Gorkha earthquake

High-mountain water storage in the form of ice, snow, and groundwater is crucial for predicting water routing to rivers. While snow and ice volumes are diminishing considerably in mountains as a result of global warming, subsurface storage volumes and transfer mechanisms remain largely elusive.

Earthquakes can act as natural experiments by changing aquifer properties and causing a hydrological response in streams. These responses can provide indications of aquifer parameters that are impossible to acquire without intrusive measurements. A range of effects including changes in streamflow, spring discharge and water table were reported. These observations can be explained by mechanisms involving water release from new sources by changing hydrological conductivity, e.g. opening new cracks or un-clogging existing conduits. The way that hydrological systems respond to seismic event can provide valuable insights into the underlying aquifers properties.

On the 25^th of April 2015, which represents the end of the dry season, when rivers levels are low, a 7.8 Mw earthquake occurred in Gorkha, in central Nepal, rupturing a 140 km segment propagating from west to south-east. We observed that rivers draining the rupture area responded by a marked increase in rivers discharge. We analyzed 26 river gauging stations covering the wider rupture area. Stations within the rupture area recorded an instantaneous rise in river water level, lasting for 1-2 days after the earthquake. Stations outside the rupture area also exhibited delayed but noticeable responses, surprisingly only in the east. The 16 stations showing a marked stream flow response are spread over an area of 90,000 km², from middle to eastern Nepal.

To identify the factors influencing the patterns of response, we compared disturbed hydrographs with precipitation data, watershed characteristics, and changes in boundary conditions.  For watersheds located at the western end of the rupture zone, the co-seismic response is transient while at the eastern end the response is sustained until the onset of monsoon. The time delay recorded at the outlets of large watersheds corresponds to the time required for water to travel from the seismic affected areas to the outlet.

To estimate the additional groundwater release induced by the earthquake, we applied a low-pass filter to the hydrographs. Then, we analyzed the recession curve parameters before and after the earthquake to investigate the modification of the aquifer permeability by the event.

During the dry season, rivers are predominantly groundwater fed. In the absence of recharge from precipitation, the volume of groundwater stored in aquifers decreases, leading to a decline in water table levels. Since the event occurred at the end of this period, the excess of water likely source from deep groundwater. Our estimates indicate that the event released around 1.3-1.5 km³ of  additional groundwater. Furthermore, the sustained rise in water levels following (and induced by) the earthquake, suggests the presence of an important deep groundwater reservoir in the Himalayan mountain range.

Earthquakes provide valuable opportunities to investigate the dynamics of fractured bedrock aquifers in high mountains such as the Himalayas.