Sunny-day flooding is becoming common in coastal North Carolina. Sensors revealed 65 flood days per year, and water-quality tests showed fecal contamination up to 100× above closure standards. A new computer model tracks how contaminated floodwaters move, helping identify hotspots and supporting safer water-quality advisories and flood-defense planning.

My research uses artificial intelligence to detect water pollution by analysing DNA traces left by aquatic species. Instead of relying on visual signs or costly expert identification, supervised machine learning reads species patterns to determine water quality. The method is faster, cheaper, and more accurate than traditional analysis.