A Simple and Reliable Measurement of Upper Crustal Vp/Vs Ratios
Abstract
The P- and S-wave velocity ratio, Vp/Vs, is an important parameter indicating the properties of medium. It has been analyzed for elucidating the presence of melts beneath volcanoes, investigating the rock composition of medium, and studying tectonic structures. Also, the Vp/Vs ratios are useful for better locationing of seismic events in regions where S wave velocity models are not available. A conventional way to measure the Vp/Vs ratios is to use Vp and Vs models that are determined separately. However, Vp and Vs models have their own limits in resolution and accuracy. Thus, manipulation of the independently-determined parameters causes difficulty in sustaining the reliability in the resultant Vp/Vs models. We use a simple Wadati method that is based on the P- and S-wave travel times (Tp, Ts). The Wadati diagram presents the time differences between Tp and Ts as a function of Tp. The Wadati diagram was originally introduced to determine the origin time and Vp/Vs ratio using a linear regression for a single event. We plot all travel time data on a diagram, and determine stable and representative Vp/Vs ratios. The Vp/Vs ratios are directly determined from the travel time information, and are rarely affected by the source depths that are usually critical in seismic tomography. We determine the upper crustal Vp/Vs ratios in the southern Korean Peninsula from local seismic events with distances less than 50 km. The earthquakes occurring around the Koran Peninsula are intraplate events with shallow focal depths less than 25 km. Dense seismic networks are operated in the Korean Peninsula, allowing us to collect high-quality local seismic records. We calculate the Vp/Vs ratios for regions that are discretized by 0.7°-by-0.7° bins with 5 data or more. We test the reliability of the calculated Vp/Vs ratios using two stability tests. We conduct bootstrap resampling method to estimate quantitative stability of data sets. Also, we examine the variations of Vp/Vs ratios for possible contamination of errors in origin times and arrival times. We find that the regional variations of Vp/Vs ratios are generally consistent with major tectonic features. It suggests that the observed Vp/Vs ratios are reflect well the compositional lithology and physical properties of medium.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2011
- Bibcode:
- 2011AGUFM.T33B2400J
- Keywords:
-
- 7205 SEISMOLOGY / Continental crust;
- 7270 SEISMOLOGY / Tomography