Research fellowship – Machine learning methods for improved geothermal energy assessments
The U.S. Geological Survey (USGS) has opened a research opportunity on "Machine learning methods for development of improved geothermal energy (conventional hydrothermal) assessments", closing date is September 25, 2020.
The U.S. Geological Service (USGS) has published an open call under the Mendenhall Research Fellowship Program – S50. Machine learning methods for development of improved geothermal energy (conventional hydrothermal) assessments – Closing Date: September 25, 2020 – How to apply?
This Research Opportunity will be filled depending on the availability of funds. All application materials must be submitted through USAJobs by 11:59 pm, US Eastern Standard Time, on the closing date.
Since the last national assessments of hydrothermal-energy resources (circa 2009), both the USGS and USDOE have invested a large amount of resources into a better understanding of hydrothermal electricity power production favorability (e.g., Play Fairway Analyses; PFA). These efforts included a systematic compilation of existing data and the collection of targeted new data.
Methods of synthesizing PFA data into favorability maps were largely expert-systems driven, so mathematical relations used are likely biased predictors. The datasets collected and the resulting favorability maps represent the collective best-judgment of teams from the geothermal industry and academic and applied researchers, so unbiased analyses of these data should improve future hydrothermal assessments.
In addition to incorporating new data (collected since the last national assessment) into hydrothermal assessment methodology, the post-doctoral researcher will be expected to interact with a wide range of USGS subject matter experts, including Energy Assessment and Minerals Assessment personnel. There are many similarities and correlations between geothermal and mineral resource features. For instance, important minerals are enriched and emplaced by hydrothermal fluids, geophysical and geological strategies for identifying hydrothermal flow paths and mineral deposits rely upon the same electrical and magnetic character, and other geophysical signals (e.g., gravity) may be used to identify geologic structural features that constrain both types of resource. Although similar techniques are initially applied for energy and minerals assessment, geothermal and mineral studies rapidly diverge in goal-oriented inquiry methodology. Mineral assessments seek to identify what sorts of minerals are present, including zonation patterns where target minerals are enriched. Alternatively, geothermal studies try to identify permeable regions that may be used to efficiently extract heat from the subsurface. In short, mineral and geothermal studies have developed two toolboxes to interrogate similar systems, and it is anticipated that advancements in minerals assessments may add value to energy assessment strategies, and vice versa.
Both Energy and Minerals disciplines employ a range of quantitative methods that loosely fall into the broad category of machine learning (ML). These quantitative methods are multi-variate, and in an ideal world (but often not in practice) they account for correlations between variables to make unbiased estimates of geothermal and mineral resources. ML is an area of recent and rapid expansion of analytical techniques, and the successful applicant will seek to understand which new methods provide benefit to geothermal energy assessments.
The proposed study will advance the understanding of the degree to which ML techniques can be employed to improve resource assessments and geothermal prospectivity maps. The work will apply ML techniques to understand the relationships between highly correlated, noisy, heterogeneous datasets collected at a range of scales to improve resource assessments and prospectivity of geothermal energy.
It is anticipated that the successful applicant will either have (1) extensive quantitative skills and will work closely with subject-matter experts to develop physical intuition (requiring a basic aptitude for understanding physical and conceptual models), or (2) extensive physical understanding and will work closely with ML experts to develop quantitative skills (requiring a basic mathematical aptitude). Programming skills and experience with spatial datasets are desired.
Interested applicants are strongly encouraged to contact the Research Advisor(s) early in the application process to discuss project ideas.
Proposed Duty Station: Portland, Oregon
Areas of PhD: Geology, geophysics, mathematics, engineering, statistics, computer sciences, physical scientists 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 the qualifications for Research Computer Scientist, Research Engineer, Research Geologist, Research Geophysicist, Research Mathematician, Research Physical Scientist, or Research Statistician
(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.)
Human Resources Office Contact: Audrey Tsujita, 916-278-9395, email@example.com