Predictive Drought Modeling for Del-Co Water
An Academic-Industry Collaboration for Climate-Resilient Infrastructure
To build a more resilient water future, Ohio Wesleyan University (OWU) Data Lab has developed an advanced machine learning model for drought prediction for Del-Co Water Company . By blending deep environmental science with cutting-edge data analytics, this project translates complex climate data into actionable insights for public utility management.
The Collaborative Team
Our project leverages a unique cross-disciplinary framework, bridging academic research and real-world utility operations:
Environmental Domain Experts: Dr. Carolina Barbosa, Dr. Nathan Rowley, and two environmental studies students from the OWU Environment & Sustainability Department. Their team spearheaded the identification of critical climate indicators and environmental variables.
Data Science Architects: James O'Connor and Aryaa Subedi, serving as Data Science Research Assistants from OWU's data analytics program. They under the supervision of Mehwish Abbasi engineered the predictive pipelines and trained the machine learning algorithms.
Utility Partners & Stakeholders: Fred Milan and Frances Wiggins, serving as primary points of contact at Del-Co Water, providing operational guardrails and industry expertise.
Project Framework & Methodology
A major challenge in climate modeling is isolating the data that truly matters. The Environment & Sustainability team identified the core environmental variablesnecessary to detect early-stage drought conditions.
James OConnor and Aryaa Subedi under the supervision of Mehwish Abbasi then ingested these variables to construct the predictive engine. Using advanced machine learning algorithms, they built a model capable of recognizing complex, non-linear climate patterns to forecast water scarcity well before it impacts infrastructure.
Current Status & Next Steps
This initial phase successfully established a robust Proof of Concept (PoC), validating that machine learning can accurately model localized drought conditions for Del-Co Water.
Phase 2: Fall Refinement
Beginning this fall, we are launching the next iteration of the model to enhance its predictive accuracy and operational readiness