Michael Manga, Professor

Open (1) Monitoring California's volcanoes through their hydrology

Open. Apprentices needed for the fall semester. Please do NOT contact faculty before September 11th (the start of the 4th week of classes)! Enter your application on the web beginning August 16th. The deadline to apply is Tuesday, August 29th at 8 AM.

Unrest at volcanoes (and their eruption) affects fluids in the crust and the temperature, discharge and composition of water at springs around volcanoes. At present, these springs are not well monitored. The first step in this project is to identify where we should monitor springs and what data exists. The second step is to make measurements at some of these springs. At the same time existing data can be used to answer key questions for interpreting this data: are there signals from climate change? can we measure how much heat the volcano is producing? do we see changes from earthquakes or unrest at the volcanoes?

The project will begin by looking for existing data in library and web resources and geological maps. A field trip is possible.

Day-to-day supervisor for this project: Michael Manga

Qualifications: Some familiarity with Earth science. Experience with handling data would be useful.

Weekly Hours: to be negotiated

Related website: http://seismo.berkeley.edu/~manga/

Open (2) Imaging magma under California volcanoes with seismic attenuation

Open. Apprentices needed for the fall semester. Please do NOT contact faculty before September 11th (the start of the 4th week of classes)! Enter your application on the web beginning August 16th. The deadline to apply is Tuesday, August 29th at 8 AM.

Understanding the dynamics beneath active volcanoes requires identifying the presence and position of magma bodies. Geophysical imaging, specially seismic tomography, is becoming a widely used technique for this purpose. However, there are still uncertainties related to the interpretation of tomography images and the joint interpretation of different physical parameters seems to be a better way to recognize and interpret volcanic structures.

Recent publications demonstrate that seismic attenuation (Q) is an useful parameter due to its strong relationship to temperature and the structural complexity of the medium.

The student involved in the URAP project will learn about seismic imaging at active volcanoes and will interpret obtained Q-images together with other geophysical images (magnetotellurics, seismic velocity).

Download seismic data from IRIS (data repository); analyze data with codes that are already written; eventually, create image of the volcano

Day-to-day supervisor for this project: Janire Prudencio, Post-Doc

Qualifications: matlab; some familiarity with Earth science; some understanding of wave physics will be useful

Weekly Hours: to be negotiated

Related website: seismo.berkeley.edu/~manga

Open (3) Finding submarine volcanic eruptions using machine learning and satellites

Open. Apprentices needed for the fall semester. Please do NOT contact faculty before September 11th (the start of the 4th week of classes)! Enter your application on the web beginning August 16th. The deadline to apply is Tuesday, August 29th at 8 AM.

Although the majority of volcanic eruptions are expected to occur on the ocean floor, few such eruptions have been identified. By identifying pumice (porous volcanic rock) rafts in satellite images, we can identify potential locations where submarine eruptions have occurred.
The project is to use machine learning techniques to identify pumice rafts from satellite images (see second link for example eruption).

The student will learn applications of machine learning, manipulation of large satellite imagery data sets, and volcanology.

We are seeking a student to use machine learning techniques to search for signatures of submarine volcanic eruptions.

Collect satellites images; construct machine learning classifier for documented eruptions; apply algorithm to larger data set

Day-to-day supervisor for this project: Robert Citron, Ph.D. candidate

Qualifications: some scripting knowledge; python or matlab; experience with machine learning helpful but not essential

Weekly Hours: to be negotiated

Related website: seismo.berkeley.edu/~manga
Related website: https://www.nature.com/articles/ncomms4660