Data Science for scaling water research Jordan S Read, USGS Office - - PowerPoint PPT Presentation
Data Science for scaling water research Jordan S Read, USGS Office of Water Information U.S. Department of the Interior U.S. Geological Survey Water Issues at Continental Scales Water quality and quantity are changing at a scale never seen
U.S. Department of the Interior U.S. Geological Survey
Jordan S Read, USGS Office of Water Information
and we are only beginning to understand the implications
A network of… Integrity
impoundments
The national hydrography dataset
https://nhd.usgs.gov
levels, discharge, reservoir and lake levels, surface-water quality, and rainfall
National Water Information System
https://waterdata.usgs.gov
we use water (1950-2010)
resolution of reporting
National Water Information System
https://water.usgs.gov/watuse
The water quality portal
https://www.waterqualitydata.us
resolution multispectral data
USGS Landsat
https://landsat.usgs.gov/
Understand water use trade-offs for energy, the environment, and human health
Understand water use trade-offs for energy, the environment, and human health
Understand water use trade-offs for energy, the environment, and human health
With familiar tools
infrastructure monitoring
Understand water use trade-offs for energy, the environment, and human health
infrastructure monitoring
Understand water use trade-offs for energy, the environment, and human health
infrastructure monitoring
Understand water use trade-offs for energy, the environment, and human health
infrastructure monitoring
Understand water use trade-offs for energy, the environment, and human health
infrastructure monitoring
Understand water use trade-offs for energy, the environment, and human health
Understand water use trade-offs for energy, the environment, and human health
infrastructure monitoring
Can we afford to leave data on the table?
Understand water use trade-offs for energy, the environment, and human health
infrastructure monitoring
Do we need more people at the table?
Understand water use trade-offs for energy, the environment, and human health
Domain Technology Domain Technology
Data poor Data rich
Understand water use trade-offs for energy, the environment, and human health
Domain Technology Domain Technology
Data poor Data rich Domain Scientists Data Scientists
image: The Economist
Use of Scientific Knowledge
High Low
Use of Data
Low High
Theory-based Models Machine learning models
Adapted from Karpatne et al. 2017
ignore systems understanding
Use of Scientific Knowledge
High Low
Use of Data
Low High
Theory-based Models Machine learning models Theory-guided Data Science Models
Adapted from Karpatne et al. 2017
Use of Scientific Knowledge
High Low
Use of Data
Low High
Domain Scientist Data Scientist
How does Data Science function at a science agency?
Practical near-term limits to scaling
technology science
Practical near-term limits to scaling
USGS Water enterprise data systems emphasis
USGS Water Data Science emphasis
Visualizations Tools Training Research
Computational thinking Computational practice
Imagination and engagement
Visualizations Training Research Tools https://owi.usgs.gov/vizlab
Visualizations Training Research Tools Outreach Instruction Advising Collaboration
Imagination and engagement
Visualizations Training Research Tools
Reproducible, transparent, and efficient
Sustaining a community of tool builders
Visualizations Training Research Tools
Reproducible, transparent, and efficient
Visualizations Training Research Tools
Reproducible, transparent, and efficient
Process-based models Big compute Machine learning
Theory-guided Data Science Models
Climate change effects on lake temperature and fish habitat for ~11,000 temperate lakes
Visualizations Training Research Tools
Reproducible, transparent, and efficient
for science agencies
Jordan Read jread@usgs.gov