27 April 2015

Nepal's earthquake, groundwater and damage

The producer asked me to look at this documentary on Kathmandu's evolving water crisis a few months ago. I finally got 'round to it a few days ago, just before the earthquake hit, killing thousands and damaging many buildings. The documentary is interesting in its own right (government monopoly and corruption are driving the crisis), but the part that I immediately remembered was how local experts predicted (at 19:09) terrible damage in the event of an earthquake. Why? Groundwater overdrafting destabilizes subsurface soils.

Bottom Line: Groundwater abuse doesn't just deplete the water that you (or your neighbors) may want later -- it increases risk in ways that may be very costly.


  1. The groundwater-quake nexus in Nepal is scary ass shit for us living in depletion/seismic zones. Any thoughts on parallels for CA?

    1. I'm no seismologist, but we know that (1) reclaimed land "liquifies" in earthquakes (e.g., SF's Marina District), (2) water is important to subsurface stability (bedrock has no water; more porous subsurfaces do), and (3) land subsides when aquifers are overpumped, thereby (de)stabilizing the soil against movements.

      Although impacts may go either way, I *suspect* that it depleted groundwater cannot be good.

  2. Over-drafting of groundwater tends to result in compaction of sediments over time, which is why surface fractures develop and subsidence of the ground surface occurs. Such groundwater withdrawal and compaction would tend to increase density and stability -- not the opposite.
    However, the compaction of sediments, in this case a thick sequence of old lake beds, can result in greater amplification of seismic energy.
    In the case of the Nepal EQ, a number of factors are significant when explaining the resulting damage in the capital of Kathmandu and these factors are discussed in the summary of conditions and the EQ presented by the USGS:
    The hypocenter of this EQ was shallow (approx. 11 km) and fault rupture (and energy) propagated eastward toward Kathmandu (roughly 77 km from the epicenter) causing ground shaking and surface ruptures within the city. As seismic energy passed through the sedimentary sequence, the energy was amplified by the sediments and produced amplified ground shaking. Buildings and other structures are typically composed of unreinforced masonry that is highly susceptible to collapse from ground shaking. In this case, Modified Mercalli Intensities of VII to VIII were observed. Further structural failures have resulted from the on-going series of moderate (or larger) magnitude aftershocks.

  3. Groundwater depletion does not 'destabilize' soils but there are a few well-known links between groundwater pumping (either into or out of aquifers) and earthquakes

    In no order:

    1) unloading faults: groundwater depletion, depending on the geometry of the fault system, can the reduce the weight and therefore stabilizing force that keeps faults 'locked' together. Removing groundwater in such systems, like the Central Valley and the San Andreas fault, has been shown to increase micro-seismicity (small quakes, not large ones) due to 'unloading' of the weight on the fault.
    2) groundwater can 'lubricate' faults, increasing the number of smaller earthquakes. this has been proposed as a means for releasing stress on major faults to avoid 'the big one'
    3) now we recognize that injecting wastewater into deeper disposal wells (i.e. from fracking) has greatly increased the number of small earthquakes in a region (latest example is Oklahoma)
    4) Not groundwater, but, building large reservoirs depresses Earth's crust and is at least in the folklore, has triggered more earthquakes as the crust accommodates, in both directions -- so like 1) there may be a seasonal cycle as big reservoirs fill and empty. I'm not so sure on this one, and only recently came across it.

    The other things you mentioned:
    Liquefaction has little to do with groundwater depletion. It is common occurrence in some soil types and depends mostly on the natural structural integrity of the soil and perhaps its water content (soil moisture, not groundwater) during the quake. Areas that have been filled with sand and gravel (think the Marina District in the SF earthquake in 1989) have little soil structural integrity and loose what they have when the ground starts shaking violently. The soils then effectively flow like liquids. The moisture content in the soils at the time of the quake is a factor, but I'm not sure how great.

    Subsidence can be driven by groundwater depletion but there is no connection to earthquakes that I know of. This mainly happens in aquifers that contain lots of clay minerals. As the water is removed from the aquifer, the clay minerals, which are flat themselves, sort of stack up and flatten out. Think of a bunch of dirty dishes thrown into a sink randomly, versus a neatly stacked pile. The latter takes much less room. That's what's happening with subsidence.

    Nepal is of course an active area tectonically so the likelihood is low that there is some tie to water use and the recent quake. A bigger issue throughout the region is likely glacial outburst floods (when glaciers melt they form lakes at their snouts which are retained only by the glacial deposits in front of the receding glaciers) which can easily be triggered by earthquakes. Also the number of new dams proposed in such a tectonically active region should also be a concern (see the new article in the New Yorker).

    Regarding the claim in the video (at 9:09)....

    He is exagerrating and conflating a few things.

    He implies that if there's an earthquake (actually 'any natural disaster') then the place where there is depletion 'will collapse,' which is not true at all

    The only connection I might make is that where there is documented and signficant subsidence, then buildings, construction, etc on the surface may already be starting to sink or tilt, so an earthquake might make that worse.

    But it sounds to me like he is conflating subsidence and liquefaction which are two completely separate processes.

    In short, I would leave it out unless you knew that subsidence was actually starting to tip stuctures over or otherwise cause stuctural damage, tilting, displacement, etc that shaking by an earthquake would worsen.


Spam will be deleted. Comments on older posts must be approved.
If you're having problems posting, email your comment to me