sciscape Publications
← All publications
physics.geo-ph · Geophysics

Strong earthquake shaking did not make baseflow recession faster at United States stream gauges (PGV of 10 cm/s or more, 1960-2026); a first-year low-flow increase is not confirmed

Question: After strong earthquake shaking, does baseflow recession become faster, and do low flows change, in the following months and years? Faster recession would show that shaking increased the drainage efficiency (permeability) of catchments. Approach: We used 367 ComCat earthquakes (M 5 or more, conterminous United States, 1960 to 2026), merged into 155 episodes, with USGS ShakeMap peak ground velocity (PGV) at USGS stream gauges and NWIS daily discharge for 5 years before and after each episode. Treated gauges had PGV of 10 cm/s or more. Control gauges had PGV below 1 cm/s within 200 km. We compared each event-year with the 5 baseline years (difference in differences). The plan was written before discharge data was downloaded. Findings: In the first year, the annual recession residual changed by -0.08 ln units (95 percent episode-bootstrap interval -0.15 to +0.06; 22 episodes, 105 treated gauges; Holm p 0.28). A within-episode year-permutation test gave p 0.11, and equal episode weights gave 0.00. Thus recession did not become faster. Recession on days 2 to 90 was slower (-0.22, -0.34 to -0.04; permutation p 0.11, or 0.005 with equal weights). The 7-day low flow rose by +0.33 ln (0.00 to +0.52; Holm p 0.10; permutation p 0.046), which failed the planned test. It was +0.20 without the 2001 Nisqually episode and +0.17 without the interference rule. Years 2 to 4 showed no clear change. False events before the earthquakes showed that the bootstrap intervals are too narrow. The result agrees with Manga (2001): shaking adds water without a lasting change in drainage efficiency, but the added-water signs are not confirmed. Limitations: about 100 treated gauges in 16 to 32 episodes; one episode supplies 39 percent of treated gauges; no rain data; model-based PGV for old events; many regulated gauges; reference-gauge subset of 4 to 7 episodes.

earthquake hydrology · streamflow · baseflow recession · permeability · ShakeMap · peak ground velocity · low flow · difference in differences · USGS NWIS · recovery

Admitted 1 citing publication

References

8
  1. usesUS stream gauges rose more often after earthquakes than in placebo windows at energy density of 1 J/m3 or more and at 10 cm/s or more, in dry periods, 1950 to 2024 sci:sha256:960bb4a1dfbff9d17d6edc024941d90d49297037888d9b55a7c7fff7489689c8

    Found the 30-day post-earthquake discharge excess that this study follows up for recession and recovery; we reuse its GAGES-II reference-gauge list.

  2. United States stream gauges show no detectable median discharge change in rain-free windows after 1980-1990 earthquakes; evidence above 1 J/m3 is too sparse to test sci:sha256:11bb393c8865d0a13337a06dd8e0227508ae0a6fbf781f5bb9b7dbfdc3443751

    Rain-filtered 1980-1990 test of short-term discharge change; related prior campaign work.

  3. doi:10.1029/2000GL012481

    Manga (2001): baseflow recession unchanged after large postseismic streamflow increases; the hypothesis tested here.

  4. doi:10.1029/2018GL078621

    Mohr, Manga and Wald (2018): PGV threshold of about 10 cm/s for streamflow responses.

  5. doi:10.1007/978-3-030-64308-9

    Wang and Manga (2021), Water and Earthquakes: review of hydrologic responses.

  6. https://earthquake.usgs.gov/fdsnws/event/1/

    ComCat events and ShakeMap products.

  7. https://waterservices.usgs.gov/nwis/dv/

    NWIS daily discharge.

  8. https://waterservices.usgs.gov/nwis/site/

    NWIS gauge inventory.

Cited by

1 publication

Public catalog entries on Sciscape.