Research ArticleEARTH SCIENCES

Monitoring transient changes within overpressured regions of subduction zones using ambient seismic noise

See allHide authors and affiliations

Science Advances  08 Jan 2016:
Vol. 2, no. 1, e1501289
DOI: 10.1126/sciadv.1501289
  • Fig. 1 Map of northern Costa Rica showing seismic velocity reductions following the 2012 Mw 7.6 Nicoya Peninsula, Costa Rica earthquake.

    The white dashed line marks the position where the origin of the subducting oceanic lithosphere changes from warmer Cocos Nazca Spreading center (CNS) to colder East Pacific Rise (EPR) (28). (A) The focal mechanism (~21-km centroid depth) and area of concentrated slip for the mainshock [dashed black contour (29)] are indicated. Black inverted triangles correspond to the broadband seismic stations used in this study. Triangles show the location of strong ground motion sensors from the Universidad de Costa Rica with their color indicating peak ground velocity (PGV). Colored lines indicate coseismic velocity reductions obtained in the period band 1 to 3 s along the interstation path. (B) Colored lines indicate seismic velocity reductions obtained in the period band 3 to 10 s. Inverted triangles indicate the Vp/Vs values obtained for the forearc crust from receiver function analysis (30). Note that there were no relative coseismic velocity increases in the frequency ranges in our analysis.

  • Fig. 2 Cross-correlation of ambient seismic noise.

    Daily GFs filtered between 1 and 10 s for the station pair INDI-SAJU. The black trace represents the reference GF (RGF). The dotted line indicates the time of the Nicoya Peninsula earthquake. Note the high symmetry between the causal and acausal signals.

  • Fig. 3 Seismic velocity reductions, GFs, and time delays associated with the 2012 Nicoya Peninsula earthquake for the station pair INDI-SAJU in the following period bands: 1 to 10, 1 to 3, and 3 to 10 s.

    (A) The preseismic, coseismic, and postseismic velocity variations between April 2012 and October 2013. The dashed line indicates the time of the mainshock. Note the near-identical velocity reduction in the 1- to 10-s and 1- to 3-s period bands and the lack of a velocity variation in the 3- to 10-s band. (B) RGF, daily GFs, and time delays resulting from the cross-correlation of their coda filtered between 1 and 10 s: 5 days before and 15 days after the earthquake. The longer time period used after the earthquake was required to avoid mixing of postseismic signals or aftershocks.

  • Fig. 4 Seismic velocity reductions, GFs, and time delays associated with the 2012 Nicoya Peninsula earthquake for the station pair INDI-LAFE in the following period bands: 1 to 10, 1 to 3, 3 to 10, and 5 to 10 s.

    (A) The preseismic, coseismic, and postseismic velocity variations between February 2012 and June 2013. The dashed line indicates the time of the mainshock. Note the similarity in the velocity reductions in the 1- to 10-s, 3- to 10-s, and 5- to 10-s period bands, confirming that the physical processes responsible are occurring at depths of ~5 to 15 km. The delay in the velocity reduction apparent in the 5- to 10-s period band reflects the longer, 50-day stacks required to stabilize the analysis. (B) RGF, daily GFs, and time delays resulting from the cross-correlation of their coda filtered between 3 and 10 s: 5 days before and 15 days after the earthquake.

Supplementary Materials

Stay Connected to Science Advances

Navigate This Article