USGS visual identity mark and link to main Web site

USGS Mendenhall Postdoctoral Research Fellowship Program

17-21. Application of Machine Learning to Whole-Network Data Assimilation for Stream Temperature Observations

Water temperature is a master variable influencing aquatic chemistry, aquatic metabolism, the health of fisheries and industrial processes. The collection and curation of water temperature data supports societally relevant science for the Nation, including, but not limited to, water quality, hydrologic modeling, and fisheries research. These data will continue to grow in volume and complexity as scientific innovations such as sensors and remote sensing push data collection towards an increasing number of sites at higher temporal resolutions.

Given the volume of data collected by USGS and others nationally, modern computational tools and algorithms are critical to assimilate water temperature datasets across the whole network and provide an understanding of data quality.  Machine learning innovation could greatly improve the efficiency of the quality assurance-quality control (QA-QC) process and records computation and processing of large whole network environmental datasets.   An initial focus on the Delaware River Basin is designed as part of the Next Generation Water Observing System.  These approaches could then be shared and applied to other high‐frequency environmental data and used in real-time to improve the quality of our preliminary data products, cost reductions in records processing, and network design.

We seek a Mendenhall Fellow to develop and apply machine learning methods and algorithms for quality QA-QC of large water temperature datasets.  Data assimilation should make use of stream temperature observations from various sources, including in situ observations from USGS and others, as well as remotely sensed water temperature data.  Approaches should include: designing machine learning processes with input from experts currently performing QA-QC; building a dataset for algorithm development; evaluating effectiveness of machine learning techniques compared with existing processes; comparing and contrasting automated QA-QC with manual processes; and testing on a wide range of stations and conditions.

Proposed Duty Station: Reston, VA; Middleton, WI; Lakewood, CO; Richmond, VA; Lawrence, KS.

Areas of PhD: Computer science, mathematics, environmental science, hydrology, limnology, ecology, data science, or civil engineering. (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 Computer Scientist, Research Hydrologist, 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 USGS Human Resources specialist.)

Research Advisors: Brian Pellerin, (703) 648-6865, bpeller@usgs.gov; Jordan Read, (608) 821-3922, jread@usgs.gov.

Human Resources Office Contact: Nina Ngo, nngo@usgs.gov, 703-648-7431


Go back to Summary of Opportunities

U.S. Department of the Interior, U.S. Geological Survey
URL: http://geology.usgs.gov/postdoc/opps/2019/17-21 Pellerin.htm
Direct inquiries to Rama K. Kotra at rkotra@usgs.gov
Maintained by: Mendenhall Research Fellowship Program Web Team
Last modified: 17:07:40 Wed 14 Nov 2018
Privacy statement | General disclaimer | Accessibility