Tengchong is a young volcanic area near the Euro-Asian collision boundary in the southeastern Tibetan margin. Holocene volcanoes are mostly concentrated in the center of the Tengchong basin, where they construct a string-like volcanic cluster along the north-to-south direction. In order to study the magma activity and its relation with the volcanic distribution, we deployed a temporary earthquake observation within the Tengchong basin. The survey line is approximately 40 km long between the Heshui town in the south and the Gudong town in the north, passing through the Laoguipo, Dayingshan, Dakongshan, Xiaokongshan volcanoes and the Crater Lake. Broadband seismic instruments were installed at nine stations with an average 4 km spacing. The observation operated for a period of 14 months from May 2015 to July 2016. According to the Chinese Earthquake Bulletins, we selected 61 teleseismic events with magnitudes greater than M s 5.0 and distance ranges between 30° and 90°. They include 1 M s≥8.0 event, 9 M s≥7.0 events, 49 M s≥ 6.0 events, 2 M s≥5.0 events. Most of the events were located in the western Pacific region, however, they were very rare in west of the study area. Finally, 612 teleseismic waveforms are obtained from the recoding data at nine stations. A time-domain deconvolution technique is employed to compute receiver functions, in which the maximum entropy as the rule to determine auto-correlation and cross-correlation functions is used to calculate receiver functions. The advantage of this is to extract valid signals that are overwhelmed by background noise. The length-scale of seismic structure is sensitive to the waveform frequency, thus two Gaussian filters (1.0 and 2.5) were applied to constrain the receiver functions. The former is used to extract large-scale velocity structures and the latter to resolve detailed structures. Due to the low signal/noise ratio for receiver functions with G >2.5, they are not used in our inversion. The neighborhood algorithm (NA) was used for the receiver function inversion. This approach is a derivative-free searching method for finding models of acceptable data fit in a multi-dimensional parameter space. The objective is to find an ensemble of models that preferentially sampleregions of parameter space that exhibit a good fit to data, rather than seeking a single optimal model. A synthetic receiver function is calculated from the true model using the Thomson-Haskell matrix method. We selected 1000 minimal misfit models to calculate the average S-wave velocity model for each station. Results from nine stations show that there is an interconnected low-velocity zone beneath the Dayingshan, Dakongshan and Heikongshan volcanoes. It is 6–15 km at depth and ranges ~16 km from north to south. A small, low-velocity zone is found beneath the Crater Lake, which is situated at 9–16 km depth with a size of ≤8 km. They are considered to indicators of magma chambers beneath the volcanoes. A heat flow channel which connects two magma chambers can be recognized in the deep crust. To the south, prominent low velocities are observed beneath the Laoguipo volcano, suggesting strong magma activities at depths of 10–25 km, partly affected by the intersection of the Dayingjiang fault zone and the Tengchong volcanic fault zone. It may belong to another magma chamber within the crust also connected to the nearby Maanshan volcano. The Moho depth of the volcanic area varies between 38 and 41 km and is partially elevated beneath the Dakongshan volcano group. For some stations, the Moho appears to be a transition zone with an open channel for the heat flow into the crust from the upper mantle.
Read full abstract
Oct 11, 2023
Obula Reddy Kummitha + 5
Open Access