Can Artificial Intelligence Help Protect and Restore Rivers?

Hi River Lovers and Protectors,

Artificial Intelligence and its application in river conservation and restoration is an exploding field of research and advocacy. Save The World’s Rivers is evaluating and predicting future opportunities, and considering paths forward. Our initial round of research identifies these:

Top Ten Ways Artificial Intelligence Can Help Protect and Restore Rivers

1. Modeling Alternatives to Hydropower: AI can be used to model alternatives to proposed new, or existing, hydropower to serve the same power needs in the same places. The field of solar electric generation with battery storage is growing rapidly and in most cases is faster, cheaper, easier alternative than hydropower. In addition, replacing aging and environmentally destructive hydropower with solar/battery facilities can be modeled with AI.
2. Identifying The Best Dam Removal Opportunities. AI can be used to analyze myriad factors including social, economic, political, legal, and ecological (biodiversity, flow, fish recovery, nutrient loads, carbon emission/sequestration, etc.) to identify and rank the best site candidates for dam removal. This modeling can be done at the local watershed scale, in a larger watershed like the Colorado River, or in a geographic boundary like the state of California so that advocates can most effectively deploy resources. 
3. Automated Habitat Mapping and Restoration Planning: Healthy rivers depend on good habitat  — fish passages, riparian zones, wetlands, and flow regimes. AI can help analyze satellite and drone imagery to map habitats, invasive species, and degraded areas, as well as design restoration strategies (e.g., optimal placement of wetland buffers, where to remove barriers, daylighting urban waterways, etc).
4. Biodiversity Monitoring and Protection:  AI can “listen to” and “see” wildlife in ways humans cannot. AI can help identify fish species from hydro-acoustic data or underwater video, detect rare or endangered species using environmental DNA (eDNA) pattern recognition, and analyze soundscapes to track birds, amphibians, and mammals tied to river health.
5.  Smart Watershed Management and Flood Forecasting: Rivers depend on the health of the entire watershed. AI enables more precise and proactive planning to model spring runoff, sediment transport, carbon emission/deposition, and beach erosion/restoration. AI can more accurately forecast floods and drought severity. AI can also simulate how land-use changes (logging, agriculture, development) affect rivers to prevents habitat loss, protect communities, and better guide sustainable land-use decisions.
6.  Water Quality Monitoring and Prediction: AI can transform water monitoring from slow, periodic sampling to continuous, intelligent assessment. Using sensor, satellite, drone, and weather data, pollution can be detected early; and harmful algal blooms, chemical spills, or oxygen depletion can be predicted before they occur.
7. Habitat Restoration and Dam Removal Management: 3D river models simulate restoration scenarios (e.g., re-meandering), predict sediment flow, and prioritize revegetation sites, all of which can help model the most efficient and ecologically healthful ways to remove a dam. 
8. Changing Diversion Points and Timing to Optimize Streamflow and Ecological Health. Many Western U.S. rivers have dozens (or more) of small (or larger) dams and diversions points serving agricultural and municipal customers, most of which are 50 – 150 years old and were not created with the ecological health of the river in mind. AI can model changes in the timing of diversions, and diversion points, to optimize the ecological health of a river.
9.  Volitional Fish Migration and ESA recovery. AI can help model and identify the creeks and rivers that will achieve the best ESA recovery scenarios for individual fish species by including  legal and ecological inputs, among others, so that advocates can hi-grade resource deployment, legal strategies, and prioritize recovery goals.
10.  EIS Alternatives Analysis and Legal Briefs: NEPA, FERC, Clean Water Act, and Endangered Species Act reviews can include AI analysis to help highlight important arguments to include in public comment processes that may identify future litigation success and failure. Identifying end-game outcomes can also help inform and optimize negotiation and settlement priorities.

We are continuing to learn, and this list is continuing to evolve. Some of the items on this list are currently being applied, while others are conceptual. As more and more information is available in publicly accessible databases on the internet, and as AI modeling and coding become more informed and cost- and time-efficient, we predict that concepts will increasingly move into application.

We welcome input as well as potential case study opportunities. Contact Gary Wockner, Director, Save The World’s River: gary@savetheworldsrivers.org