19-41. Development of water monitoring systems based on fusion of ground and multi-sensor satellite data and application of machine learning

Background:

The USGS Water Mission Area (WMA) monitors inland waters of the U.S. and develops tools to improve management and understanding of water resources. The WMA has the goal of operationally producing seamless national observations of water quality and water quantity by integrating ground-based data and multi-sensor satellite data.  These geospatial hydrologic observations will be delivered in near-real-time to meet WMA priorities and requirements including guiding flood and drought planning and forecasting, assessing and forecasting water availability for human and ecological use, and improving models used for water quantity and water quality prediction and decision support.

The current combined availability of earth observation satellite data, advances in data science, and increased high-performance geo-spatial computing resources presents an exciting opportunity to expand WMA’s water observation capabilities.  We are looking for a post-doctoral researcher to help us take on this opportunity and build new observation tools for the estimation of water flows, extent, clarity, temperature, or other water quantity and quality variables.

Description of the Research Opportunity:

We seek a postdoctoral fellow to work with an interdisciplinary USGS team to advance our ability to detect and accurately quantify water properties from orbital assets, including low-latency multispectral and synthetic aperture radar (SAR) satellite imagers. The fellow will work on novel methods to produce integrated hydrologic datasets that combine ground-based observations and remote-sensing observations to reduce bias, capture spatial variability, and improve surface-water characterization. The fellow will collaborate with scientists from multiple USGS mission areas and be mentored by experts in hydrology, data science, remote sensing, and high-performance computing.

The fellow will research machine learning algorithms and techniques for rapidly converting multi-scale data into hydrologic observations.  Data sources include satellite imagery, hydrologic time-series from ground stations and, potentially, optical, thermal, and hyperspectral imagery collected from ground or aerial platforms.  Satellite data available for operational application are from radar and spectral sensors and may be sourced from public or commercial providers.  The fellow will collaborate with other scientists and programmers to implement the new data techniques and research findings into operational water monitoring systems.  

Hydrologic variables of interest to WMA for this research include reservoir storage, river discharge, soil moisture, snowpack, and observable water-quality parameters including turbidity, cDOM, chlorophyll-a, and cyanobacteria. Potential research topics include, but are not limited to development of: novel data assimilation techniques; methods for quantifying observational uncertainty; methods for bridging time gaps between satellite overpasses; methods for fusing data sets of multiple types and resolutions; machine learning algorithms for estimation of a water quantity or quality variable; and methods for incorporating additional data into neural network models.  We encourage applicants to propose research aligning with one of the above topics and with their own expertise and experience.

The Fellow will have access to substantial USGS assets including: recently developed high-performance computing resources, collocated massive collections of multi-spectral, hyperspectral and lidar data; containerized workflows and other analytical tools; national-scale operational hydrologic models of runoff, water quality, and water availability; and collaboration with experts in all of these topic areas.  The Fellow will be organizationally stationed within the Hydrologic Remote Sensing Branch of the Observing Systems Division of the WMA and will receive mentoring from scientists in other USGS offices.  The Fellow will join a WMA project team and support the work of the project with their research.  Potential projects to join include the Remote Sensing of Discharge, Remote Sensing of Water Quality, National Hydro-Geospatial Fabric, and Integrated Water Prediction Community Testbed projects.

The USGS Core Science Systems Mission Area (CSS) recently began operation of two new high-performance computing systems at the Earth Resources Observation and Science (EROS) Center.  These new systems – Denali and Tallgrass – became fully operational in January 2020 and combine large computing power capabilities with collocated storage of a massive satellite data repository.  The Tallgrass system is designed for artificial intelligence (AI) and analytics workloads and is equipped with both the hardware and software to address modern AI, deep learning and machine learning with large GPU compute capability.

Interested applicants are strongly encouraged to contact the Research Advisor(s) early in the application process to discuss project ideas.

Proposed Duty Station: Leetown, WV; Reston, VA; or Location where there is a USGS Office

Areas of PhD: Computer science, artificial intelligence, machine learning, mathematics, engineering, remote sensing, geographic/cartographic sciences, hydrologic sciences, hydraulics, 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 Geographer, Research Hydrologist, Research Physical Scientist, Research Engineer.

(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: Beverly Ledbetter, 916-278-9396, bledbetter@usgs.gov

Apply Here