This study used a dense dataset of 2- and 3-D seismic reflection data. Interpretations of the entire dataset across the Tugulu, Manas and Huoerguosi anticlines were used to constrain the geometry of the Southern Junggar Thrust (SJT). These interpretations, in turn, acted as the constraints for the 3-D fault model shown in Figure 3 of the main text.
Figure S1. Seismic reflection data base map of the southern Junggar basin. Black lines indicate 2-D seismic reflection profiles. Yellow boxes are 3-D seismic reflection surveys. White lines are 2-D seismic profiles shown in other figures. Seismic reflection data were migrated and depth converted using a velocity model derived from stacking velocities and sonic logs. All available seismic reflection data were interpreted and used to develop the 3-D fault model in Figure 3. Landsat image is shown in bands 7-4-1 (R-G-B).
Figure S2. Uninterpreted seismic reflection profile across the Tugulu and Qigu anticlines. See Figure 2 for interpretation. See Figure S1 for line location.
Figure S3. Uninterpreted seismic reflection profile across the Huoerguosi anticline and Tian Shan range front. See Figure S1 for line location.
Figure S4. Interpretation of Figure S3. Stratigraphic interpretations are constrained by surface outcrop, well penetrations, seismic reflection character, and regional interpretations. See Figure 2 for description of interpretation. Double-arrow arches across the lower ramp of the SJT depict the upper and lower bounds of the fault ramp uncertainty based on interpretations of all seismic reflection data. See Figure S1 for line location.
Figure S5. Uninterpreted seismic reflection profile across the Manas anticline and Tian Shan range front. See Figure S1 for line location.
Figure S6. Interpretation of Figure S5. Stratigraphic interpretations are constrained by surface outcrop, well penetrations, seismic reflection character, and regional interpretations. See Figure 2 for description of interpretation. Double-arrow arches across the lower ramp of the SJT depict the upper and lower bounds of the fault ramp uncertainty based on interpretations of all seismic reflection data. See Figure S1 for line location.
This study used seismicity data downloaded from the China Seismograph Network (CSN). This data was downloaded from http://www.csndmc.ac.cn/csndmc/catalog_direct_link.htm. Events of any magnitude within 41°N - 46°N and 84°E - 90°E were originally downloaded (1377 in total). These events range in moment magnitude from M 2.0 – M 6.2. Eight events in the catalog lacked depth data and were omitted. Because the CSN catalog does not contain location nor depth uncertainties, we omitted the 10 events with Z > 40 km below Earth's surface (Figure S7). After removing these deepest events, we followed the workflow of Nazareth & Haukkson (2004) to model the base of the seismicity in our study area (Figure S8). The parameters we used were: 0.8° x 0.8° bins, a minimum of 8 events had to be present within a bin, and we interpolated down to 99.0% of the Mo in each bin. These parameters were modified from the original Nazareth & Haukkson workflow due to the sparse dataset that we used (1359 events). The Z > 40 km threshold mentioned above is 7 km deeper than the mean depth of our modeled base of seismicity surface (Z = 33 km) and all of these deepest events that were omitted had M < 5.2. Therefore, we conclude that excluding so few events did not have significant effects on this modeled surface.
Figure S7. Depth-distribution plot of the raw CSN seismicity catalog used to model the base of the seismogenic crust in the study area. Events deeper than 40 km were omitted from the dataset used in our model.
Figure S8. A) Perspective view up and to the NE displaying our base of seismicity surface with the hypocenter events used in our model. B) Perspective view down and to the NE displaying our base of seismicity surface with the hypocenter events used in our model. Note that the displayed seismicity data in A and B are only the events within the X-Y domain of the displayed base of seismicity surface. Hypocenter colors correspond to their moment magnitude.
This study used empirical relations to yield moment magnitude estimates for each SJT fault area scenario. We used the Wells & Coppersmith (1994) equation for reverse fault area:
M = 4.33 + 0.90*log(RA)
where M is moment magnitude and RA is rupture area in square kilometers. In addition, we used the Leonard (2010) equation for scarp forming intraplate ruptures (SCR):
log(Mo) = 6.38 + 1.50*log(A)
where Mo is seismic moment in N*m and A is rupture area in square meters. We then used the definition of moment magnitude (Hanks & Kanamori, 1979):
M = -10.7 + (2/3)*log(Mo)
to relate the log(Mo) value from Leonard (2010) calculations to moment magnitude. Note that the Mo value in Leonard (2010) has units of N*m and the Hanks & Kanamori (1979) definition of moment magnitude reported here requires Mo to be in units of dyn*cm.
Table S1. Fault area and corresponding moment magnitudes for the SJT 3-D fault model. Base of Seismicity 2 and 3 refer to our uncertainty analysis where we raised and lowered, respectively, the original base of seismicity surface by 5 km (see primary manuscript for details).
Hanks, T. H. & H. Kanamori (1979), A moment magnitude scale, J. Geophys. Res., 84, 2348-2350.
Leonard, M. (2010), Earthquake fault scaling: self-consistent relating of rupture length, width, average displacement, and moment release, Bull. Seismol. Soc. Am., 100, 1971-1988, doi:10.1785/0120090189.
Nazareth, J.J., and E. Hauksson (2004), The seismogenic thickness of southern California crust, Bull. Seismol. Soc. Am., 94, 940-960.
Wells, D.L., and K.J. Coppersmith (1994), New empirical relationships among magnitude, rupture length, rupture width, rupture area, and surface displacement, Bull. Seismol. Soc. Am., 94, 974-1002.
[ Back ]
