Regional changes in streamflow after a megathrust earthquake

Abstract

Moderate to large earthquakes can increase the amount of water feeding stream flows, mobilizing excess water from deep groundwater, shallow groundwater, or the vadose zone. Here we examine the regional pattern of streamflow response to the Maule M8.8 earthquake across Chile's diverse topographic and hydro-climatic gradients. We combine streamflow analyses with groundwater flow modelling and a random forest classifier, and find that, after the earthquake, at least 85 streams had a change in flow.Discharge mostly increased (n = 78) shortly after the earthquake, liberating an excess water volume of >1.1 km 3 , which is the largest ever reported following an earthquake.Several catchments had increased discharge of >50 mm, locally exceeding seasonal streamflow discharge under undisturbed conditions. Our modelling results favour enhanced vertical permeability induced by dynamic strain as the most probable process explaining the observed changes at the regional scale. Supporting this interpretation, our random forest classification identifies peak ground velocity and elevation extremes as most important for predicting streamflow response. Given the mean recurrence interval of ~25 yr for >M8.0 earthquakes along the Peru-Chile Trench, our observations highlight the role of earthquakes in the regional water cycle, especially in arid environments.

Stronger Peak Ground Motion, Beyond the Threshold to Initiate a Response, Does Not Lead to Larger Stream Discharge Responses to Earthquakes

“…Our findings thus favor a binary response, that is, that streamflow changes occur if ground motion is large enough, but the magnitude of response is independent of the ground motion beyond the threshold. A previous study was unable to identify hydrogeological, tectonic, or land cover controls (Mohr et al, 2017). Similarly, Cox et al (2015) report a threshold for PGA of 0.5% g for springs in New Zealand to respond to earthquakes.…”

“…PGVs provided by ShakeMap combine the maximum value observed on the two horizontal components of motion. We do not compare responses with static volumetric strains because a previous analysis of the response to the 2010 M8.8 Maule earthquake and studies elsewhere found no general correlation between the sign of the volumetric strain and the sign of the discharge change (e.g., Mohr et al, 2017). We do not compare responses with static volumetric strains because a previous analysis of the response to the 2010 M8.8 Maule earthquake and studies elsewhere found no general correlation between the sign of the volumetric strain and the sign of the discharge change (e.g., Mohr et al, 2017).…”

“…We follow previous studies (e.g., Manga et al, 2016;Mohr et al, 2017;) by using a one-dimensional groundwater flow model in which excess water is made available, either by changing vertical permeability or releasing water from storage . The model neglects quick flow since rainfall was not registered nor does recession analysis point to quick flow (in the case of incomplete rainfall time series).…”

“…The solution for Q as a function of time t since the earthquake is (Mohr et al, 2017) The aquifer extends from x = 0 at the catchment divide, where ∂h/∂x = 0, to x = L at the stream with h (L, t) = 0.…”

Water-table and discharge changes associated with the 2016–2017 seismic sequence in central Italy: hydrogeological data and a conceptual model for fractured carbonate aquifers

“…The mechanism causing both sustained negative and positive hydraulic changes is frequently related to static stress modifications evidenced by comparison between pre-seismic and post-seismic conditions (Jonsson et al 2003;Montgomery and Manga 2003;Parvin et al 2014;Mohr et al 2016). The difference between transient dynamic oscillations (Cooper et al 1965) and the offset-type Fig.…”