Oral Presentation Society for Freshwater Science 2026 Annual Meeting

Harnessing artificial intelligence to automate environmental predictions (135435)

Avni Malhotra 1 , Brieanne Forbes 1 , Stefan Gary 2 , Amy Goldman 1 , Breanna Rivera Waterman 3 , Vanessa Garayburu-Caruso 1 , Etienne Fluet-Chouinard 1 , Sushant Mehan 4 , Michael Bruen 5 , Marcelo Ardon 6 , Bayani Cardenas 7 , Walter Dodds 8 , Christian Lonborg 9 , William H. McDowell 10 , Moussa Moustapha 11 , Allison Myers-Pigg 1 , Peter Regier 1 , Tod Rubin 12 , Hyun Song 13 , Ryan Stewart 14 , Jorge Villa 15 , Nicholas Ward 1 , Tim Scheibe 1 , James Stegen 1
  1. Pacific Northwest National Laboratory, Richland, WA, United States
  2. Parallel Works, Chicago
  3. University of Kansas, Lawrence
  4. South Dakota State University, Brookings
  5. University College Dublin, Dublin
  6. North Carolina State University, Raleigh
  7. The University of Texas at Austin, Austin
  8. Kansas State University, Manhattan
  9. Aarhus University, Aarhus
  10. The University of New Hampshire, Durham
  11. The University of Ebolowa, Ebolowa
  12. The Watershed Project, Richmond
  13. The University of Nebraska–Lincoln, Lincoln
  14. Virginia Tech, Blacksburg
  15. The University of Louisiana at Lafayette, Lafayette

Predicting heterogeneous and non-linear processes remains a fundamental challenge in Earth sciences. Here, we present an artificial intelligence (AI)-guided, closed-loop workflow that iteratively combines predictive modeling with targeted field sampling to rapidly improve environmental predictions. We demonstrate our workflow by predicting oxygen consumption, a key process in rivers, across the contiguous United States (CONUS). Our approach consists of multiple iterative loops of measurements and models, combining distributed participatory field sampling, lab analysis, automated machine learning (ML) predictions, and error and distinctiveness analyses to autonomously guide the next sampling at optimal site locations. Specifically, we used an ensemble of stacked regressors to both maximize predictive performance and quantify predictive uncertainty, which then informed the data acquisition policy. Over 18 iterative loops, we increased the predictive power of sediment oxygen consumption across CONUS rivers by over fifteenfold between the first and last iteration. Relative to our last sampling iteration, our first sampling missed rivers with outlier values and underestimated median oxygen consumption rates by 68%. In addition to identifying areas of high oxygen consumption rates, iterations enabled refinement of laboratory and data handling methods, and engagement with a broad community of field researchers. We conclude that AI-guided iterative loops between targeted sampling and predictive modeling are a powerful and efficient approach for improving predictions of heterogeneous environmental processes. The architecture of our workflow balances model flexibility with interpretability and can be readily applied to other domains where uncertainty- and distinctiveness-aware site selection is crucial.