Description of the Research Opportunity
Machine learning is a tool that can identify relationships in datasets without explicit programming. With the capability of analyzing vast amounts of data, machine learning has the potential to identify patterns and complex relationships that would be difficult, or even impossible, for humans and traditional methods to accomplish. Once these relationships have been developed, machine learning can also aid manually intensive tasks, improving the speed and accuracy of data analyses.
In seismology, recent advancements with machine learning and correlation-based techniques have improved our ability to produce more complete earthquake catalogs. While many additional earthquakes can be identified with these techniques, characterizing these new, often poorly recorded earthquakes has proven to be more challenging. With machine-learning enhanced catalogs of earthquakes, there is potential to gain greater insight into seismological processes if these events can be better characterized. The growing number of seismic instruments, variable quality of recordings, and new types of instrumentation (e.g., nodal arrays, distributed acoustic sensing) can be more fully leveraged to better characterize earthquakes and Earth’s subsurface structure.
We invite research proposals that can improve our understanding of earthquake hazards by developing new analyses and tools to take greater advantage of the vast amount of seismological data and interpreted products produced over the past decades. We seek a Mendenhall Postdoctoral Scholar who will use and/or develop innovative machine learning techniques to improve the characterization of earthquakes or subsurface structure with the aim of better understanding earthquake processes and their impacts. This could involve the development or integration of existing approaches (e.g., correlation-based detectors) with machine learning techniques. The candidate could explore a range of machine learning applications, including, but not limited to:
- Earthquake detection
- Phase arrival picking
- Seismic phase association
- Earthquake location
- Magnitude determination
- Source discrimination
- First-motion polarity classification
- Focal mechanism determination
- Source parameter estimation
- Stress inversion
- Seismogenic fault mapping
- Velocity model inversion
- Earthquake declustering
Projects could address natural, induced, or volcanic seismicity at local or regional scales. In addition to the decades of data recorded from regional seismic networks, additional datasets from local networks, nodal arrays, and distributed acoustic sensing would be available for use. The methodologies and scientific insights gained from the project would be expected to be widely applicable to other regions.
We encourage proposals that expand upon or integrate existing USGS research efforts, which could include improving comprehensive catalogs of seismicity, taking full advantage of locally deployed seismic network arrays, or advancing products like the National Seismic Hazard Model. Mentors from several USGS projects are available to advise this postdoctoral research across multiple USGS Science Centers.
Interested applicants are strongly encouraged to contact the Research Advisor(s) early in the application process to discuss project ideas.
Proposed Duty Station(s)
Moffett Field, California
Pasadena, California
Areas of PhD
Geophysics, seismology, geodesy, geomechanics, or related fields (candidates holding a Ph.D. in other disciplines, but with extensive knowledge and skills relevant to the Research Opportunity may be considered).
Qualifications
Applicants must meet one of the following qualifications: Research Geophysicist or Research Computer Scientist.
(This type of research is performed by those who have backgrounds for the occupations stated above. However, other titles may be applicable depending on the applicant's background, education, and research proposal. The final classification of the position will be made by the Human Resources specialist.)
