Date of Award


Degree Name

Doctor of Philosophy




Aaron A. Velasco


We present in this study numbers of technique to understand the volcanic processes using seismology. The types of techniques we used are earthquake location (double difference travel-time, cross-correlation tremor technique, and time reversal), stresses (b-values, focal mechanism), and tomography imagery (ambient noise tomography). The areas of study we focus on are in two different tectonic settings. The East Africa Rift System at Kenya, studying four volcanoes: Menengai Caldera, Silali, Paka, and Korosi volcanoes. The other area of study is located in a subduction zone in Central America, Salvador at San Miguel volcano.

In this study we aim to calculate accurate hypocentres for earthquakes and the b-values that have occurred in the Menengai Caldera, Silali, Paka, Korosi. This is achieved through the use of seismic velocity profiles we obtain using earthquake arrival time data and a relative relocation procedure that utilizes the precise travel times we measure using waveform cross-correlation of earthquake seismograms.

At Menengai network, we identify small magnitudes and are caused either by natural heat loss, which causes thermal contraction and cracking or by interaction of magma with its surrounding medium. The catalogue is complete down to Mc = 2.1 this area, and the b-value range is 0.96 - 1.09, revealing earthquake span in a swarm zone. We interpret our results as magma passing through conduits of the magma chamber and/or fluid being transported as a function of magma movement or hydrothermal activity. We calculate the focal mechanism and there are different orientated nodal planes; we look over the T-axis which is the Dilatation axis minimum compressive stress and it is orientation is dominated by northward and NE-SW

At Paka and Silali volcanoe, we observed earthquake sources in this area are not triggered by fluids but rather by the regional change in stress, which in turn is induced by the dike system. The catalogue is complete down to magnitude completeness Mc = 2.3 this area, and the b-value is 0.84, revealing earthquake span in a swarm zone with a focal depth 2.1 â?? 10 km. Specially, we attempt to interpret the distribution of clustered earthquake hypocentres. Since there are no stations at Korosi volcano, earthquakes locations are pulled toward the networks to the north of Paka and Silali volcano resulting in bias up to several km. This is reflected in our results, by very low b-values, which suggests that no significant or no fluid transport at all occurred there. We calculated fault plane solutions to quantify the style and orientation of fault slip in response to magma reservoir or tectonic stresses.

We successfully applied Ambient Noise Tomography to obtain images of volcanic structures â?? being especially promising for imaging volcano reservoirs at unprecedented resolution. It reveal different geological structures at global, regional, and local scales using only a few hours to a few months of continuous seismic noise.

In this study, we images Menengai caldera to determine the geometry and location of the magmatic reservoirs. We find three anomalies. Two of them (with S-wave velocity of about 1.2-1.4 km/s) are located below the Menengai Caldera and the other outside the Menengai summit flank of the volcano. Both are shallow (<1 km depth) has oblate spheroidal shape. Third is a deeper and located between a depth of 4 and 7 km below the base. It has funnel-like shape. Three anomalies are strongly elliptical in an EW direction and separated by a 2-3 km thick zone with Vs of 1.8 â?? 2 km/s. As far as these anomalies are located under the hydrothermal activity of Menengai Caldera.

In Silali, Paka and Korosi volcanoes, we reveal two anomalies identify as magmatic reservoirs. The first anomaly (Region A) is deep, between 4 and 6 km depth below the active Silali volcano. The second anomaly (Region B) is deep, between 3 and 6 km away from Paka volcano. Naturally interpreted as a dike system, but we could not determine the correct size and location. We believe is identify as a dike system migrating to the magma chambers for Paka and Korosi volcanoes. The third anomaly (Region C) is deep, between 0 and 7.5 km beneath the Korosi volcano.

The concept of time reversal was previously successfully applied for acoustic waves in many fields like medical imaging, underwater acoustics and non-destructive testing. We propose the regional earthquakes (e.g., Lg phase) triggers volcanic activity in San Miguel Volcano. That wave energy propagates through the earth upper crust and perturbed within the volcanic structure triggering LPs and tremors. Itâ??s extremely difficult to locate long period and tremor sources with classic hyponcentre determination. We present the first application of a time reverse location method in a volcanic setting, for a long-period (LP) and monochromatic tremor on San Miguel Volcano. Time reversal methods involve using the reversal of the seismic signal to produce a wave simulator where the signal propagates through a 3D numerical model. That wave propagation travels back in time until begins at its original source location. The source location is the computational 3D grid cell of the amplitude signal. Here, we investigate the feasibility of this method for several LP and tremor sources and present a gallery of time reversal source images.

We pre-process the signals by filtering the monochromatic tremor filter of 2 Hz and long period filer of band-pass 0.2 â?? 0.7 Hz. Then, we damped the amplitude by removing the instrumentation response, since the amplitude is important factor to locate the source. The monochromatic tremors determine a good precise source location with 4 to 6 km focal depth that is close near the vent of the volcano. The long period earthquakes do not present a good accuracy location, since we used a small geometry network. The location reveals inside the San Miguel network and we proposed that the long period are triggered by regional events below the flank of the volcano. We located 11 long periods and two monochromatic tremors the lapse time of the event been located. The events sequences follow the dike system of the volcano.

We compared the time reversal with relocation earthquakes and cross-correlation tremor location technique using long period and tremor. To see any relationship that could favor interpretation of the volcanic processes of San Miguel. We relocated approximately 600 events using the double-difference inversion and reveal that the earthquakes are spread on the conduit of the volcano with a depth of 0.5- 2km. For the cross-correlation tremor location, we located the long period and tremor SW and SE nears the flank of the volcano. This is important observation demonstrating the location volcanic processes interpreted as intrusive magma passing through the dyke system or long periods been trigger below the volcano by regional stresses.

Our approaches, help monitor the volcanic process such as the movement of magma, map the geometry of the magma reservoir, and observed the changes in volume that could lead to volcanic eruption.




Received from ProQuest

File Size

141 pages

File Format


Rights Holder

Ezer Patlan