This electronic supplement contains an overview map of the region of interest in Oklahoma along with seismicity that occurred in the three months before and after the Pawnee earthquake. This map also shows the location of the two regions we further study in the main article, surrounding the 3 September 2016 Pawnee and 7 November 2016 Cushing earthquake epicenters and their relation to the monitoring wells, Payne 07 and Payne 08 (Fig. S1). Additionally, we include a well completion diagram for each of the aforementioned monitoring wells (Figs. S2 and S3). Figure S4 shows the sensitivity of the root mean square (rms) error (between observed and predicted fluid levels) to the seven input parameters used to compute the predicted fluid levels. Three hypotheses that may explain the spatial and temporal conditions of the Arbuckle formation in the vicinity of the Payne 07 and Payne 08 monitoring wells are examined in the main article. Hypothesis 3 is preferred; however, hypotheses 1 and 2 cannot be ruled out as possible conditions. Therefore, Tables S1 and S2 show the best-fitting source and poroelastic parameters for the hypotheses 1 and 2 that are discussed in the main article.
Table S1. Best-fitting source parameters for hypotheses 1 and 2.
Table S2. Best-fitting set of poroelastic parameters determined for hypotheses 1 and 2 in the main article (H1 and H2, respectively). Note that B is derived from Ku, Kfl, and ϕ. There are two sets of poroelastic parameters, because this hypothesis assumes the poroelastic parameters are spatially variable and are different between each well.
Figure S1. Location of monitoring wells installed by the end of 2016 in northern and central Oklahoma. Seismicity during the three months before (open circles) and after (green circles) the Pawnee earthquake, scaled by magnitude. Focal mechanisms for the three largest events provided by the National Earthquake Information Center (NEIC, see Data and Resources) catalog. Orange boxes denote the regions mapped in Figure 1 of the main article (larger box in Fig. 1a, and smaller box used in Fig. 1b in the main article).
Figure S2. Well completion diagram showing approximate depths and thicknesses of geologic zones of interest. The Arbuckle Group is estimated to be 940 ft (286.5 m) thick at this location with about a 940-ft-thick openhole injection interval in the well. The pressure gradient measured in this well is 0.483 psi/ft (10.92 kPa/m).
Figure S3. Well completion diagram showing approximate depths and thicknesses of geologic zones of interest. The Arbuckle Group is estimated to be 973 ft (296.6 m) thick at this location with about a 760-ft-thick openhole injection interval in the well. The pressure gradient in this well is measured to be 0.479 psi/ft (10.84 kPa/m).
Figure S4. Sensitivity analysis conducted with a one-at-a-time test, wherein the predicted fluid levels are computed for the best-fitting set of model parameters (listed in Tables 3 and 4 of the main article), varying one parameter at a time over the range examined (listed in Tables 1 and 2 of the main article). We perform this analysis for 30 values of each uniformly distributed variable. Values of each parameter increase with increasing parameter index. The sensitivity to each variable is given by the colored lines in each figure according to the legend in (b). The rms error computed for the predicted fluid levels in Payne 07 and Payne 08 separately for each earthquake, because the best-fitting parameters result from hypothesis 3 which suggests the poroelastic parameters are uniform in space and heterogeneous in time. An asterisk is placed at the best-fitting value of each parameter. We find that the predicted fluid levels due to the Pawnee event are most sensitive to Bulk modulus Ku and the amount of slip, and to a lesser extent, the porosity. Similarly, the predicted fluid levels due to the Cushing event are sensitive to Ku and slip but are also sensitive to the strike of the fault plane as well.
The U.S. Geological Survey (USGS) National Earthquake Information Center (NEIC) moment tensors were found on http://earthquake.usgs.gov/earthquakes/eventpage/us10006jxs#moment-tensor and http://earthquake.usgs.gov/earthquakes/eventpage/us100075y8#moment-tensor (last accessed December 2016). NEIC catalog used in Figure S1 is obtained from https://earthquake.usgs.gov/earthquakes/map/#%7B%22feed%22%3A%221489434437142%22%2C%22sort%22%3A%22newest%22%2C%22mapposition%22%3A%5B%5B35.008%2C-98.229%5D%2C%5B36.99%2C-95.328%5D%5D%2C%22viewModes%22%3A%7B%22help%22%3Afalse%2C%22list%22%3Atrue%2C%22map%22%3Atrue%2C%22settings%22%3Afalse%7D%2C%22autoUpdate%22%3Afalse%2C%22search%22%3A%7B%22id%22%3A%221489434437142%22%2C%22name%22%3A%22Search%20Results%22%2C%22isSearch%22%3Atrue%2C%22params%22%3A%7B%22starttime%22%3A%222016-06-03%2000%3A00%3A00%22%2C%22endtime%22%3A%222016-12-03%2023%3A59%3A59%22%2C%22maxlatitude%22%3A36.99%2C%22minlatitude%22%3A35.008%2C%22maxlongitude%22%3A-95.328%2C%22minlongitude%22%3A-98.229%2C%22minmagnitude%22%3A1%2C%22orderby%22%3A%22time%22%7D%7D%7D (last accessed December 2016).
[ Back ]
