Research ArticleGEOPHYSICS

Shear-wave anisotropy reveals pore fluid pressure–induced seismicity in the U.S. midcontinent

See allHide authors and affiliations

Science Advances  13 Dec 2017:
Vol. 3, no. 12, e1700443
DOI: 10.1126/sciadv.1700443
  • Fig. 1 Map of the study area in south-central Kansas and northern Oklahoma.

    Colored triangles are seismometer station locations, and colored circles are earthquake epicenters. Color identifies the time period of the earthquakes and the corresponding recording stations. Red: 2010–2012, EarthScope Transportable Array (TA); green: 2013–2015, Nanometrics Research Network (NX) and the U.S. Geological Survey (USGS) Networks (GS); blue: 2015–2016 Wellington, Kansas CO2 Sequestration Monitoring network (ZA). Downhole pressure measured in the KGS 1-28 well. Most events used in the study occurred in or near western Sumner County, Kansas. More distant events in northern Oklahoma were incorporated during early time periods, when there was very little seismicity in Kansas. The timing of earthquake occurrence suggests a progression of seismicity from south to north over the 7-year period.

  • Fig. 2 Temporal variation in φ and δ.

    (A1) Polar histogram of φ from 2010–2012 TA stations (red). Common φ values are near the maximum horizontal stress of 70° to 85° along with flipped values at ~330°. Zero degree values are most often null solutions. (A2) Polar histogram of φ from NX and 2013–2015 GS stations (green) shows the most common solutions in line with maximum horizontal stress and flipped solutions approximately 90° off of maximum horizontal stress. (A3) Polar histogram of φ from 2015–2016 ZA stations (blue) shows the most common solution to be flipped approximately 90° off of the maximum horizontal stress, a direct indicator of critical pore fluid pressure. Polar histograms depict all individual station-earthquake pairs. Black arrows indicate the orientation of maximum horizontal stress at 75°. (B) Average δt/km of earthquakes from 2010 to 2016 showing an increase in variance. Each circle depicts the average δt/km of all stations for an individual earthquake. Yellow stars correspond to average monthly pressure observations in the KGS 1-28 well at the Wellington Oil field. The initial pressure measurement in August 2011 was obtained when the well was drilled. Inset B1 is an expanded view of monthly average downhole pressures from April to November 2016.

  • Fig. 3 Spatiotemporal comparison of early time period data.

    Separation of early time period earthquakes (2010–2013) between southern group A below the red line at 36.5° latitude and northern group B above the red line at 36.5° latitude. All southern group A earthquakes occurred in 2010–2012 and correspond to the red TA events shown in Figs. 1 and 2. (A) The polar histogram shows the φ solutions of the southern group A earthquakes. (B) The polar histogram shows the φ solutions of the northern group B earthquakes, which include one event from 2010 (red) and the earliest green earthquakes that occurred in 2013. All flipped φ orientations from the red 2010–2012 data shown in Fig. 2 come from the southern group A earthquakes. Corresponding δt values are identified in Fig. 2B to the left of the vertical dashed line at the end of year 2013. The northernmost red event is the earliest earthquake in the data set.

  • Fig. 4 Raw seismograms depicting S-wave arrival.

    S-wave arrival from WK07 of the Wellington, Kansas CO2 Sequestration Monitoring Network (ZA) with no processing shown. The earthquake occurred 2 March 2016, 15 km from the recording station at a depth of 5.6 km and at M 2.0. Data are shown in north-south (N-S) and east-west (E-W) channels along with the vertical component. Right column figures are expanded views of the left column waveforms between the solid lines depicting S-wave arrival picks. Dashed lines mark the time separation in S-wave arrivals.

Supplementary Materials

  • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/3/12/e1700443/DC1

    fig. S1. Plot of the Wellington CO2 Sequestration Monitoring network (ZA) earthquake catalog consisting of 1676 events ranging in M from 0.4 to 4.3 and depth from 1 to 11 km.

    fig. S2. Depth and magnitude distributions of the 150 earthquakes used in this study.

    fig. S3. Plot of the minimization of the second eigenvalue (λ2) in ϕ and δt space from waveforms shown in fig. S4.

    fig. S4. Plot of raw channel data from station WK15 of an M 2.7 earthquake that occurred in July 2015.

    fig. S5. Hodogram plots of 0.1-s increments corresponding to the 2-s time window identified in fig. S4.

    fig. S6. Hodogram plot of S-wave splitting that aligns with the maximum horizontal stress at approximately 75° (marked with red dashed lines).

    Additional Acknowledgments

  • Supplementary Materials

    This PDF file includes:

    • fig. S1. Plot of the Wellington CO2 Sequestration Monitoring network (ZA) earthquake catalog consisting of 1676 events ranging in M from 0.4 to 4.3 and depth from 1 to 11 km.
    • fig. S2. Depth and magnitude distributions of the 150 earthquakes used in this study.
    • fig. S3. Plot of the minimization of the second eigenvalue (λ2) in ϕ and δt space from waveforms shown in fig. S4.
    • fig. S4. Plot of raw channel data from station WK15 of an M 2.7 earthquake that occurred in July 2015.
    • fig. S5. Hodogram plots of 0.1-s increments corresponding to the 2-s time window identified in fig. S4.
    • fig. S6. Hodogram plot of S-wave splitting that aligns with the maximum horizontal stress at approximately 75° (marked with red dashed lines).
    • Additional Acknowledgments

    Download PDF

    Files in this Data Supplement:

Stay Connected to Science Advances

Navigate This Article