Knowledge-Guided Machine Learning for Partitioning Ecosystem Water Fluxes from Flux Tower Data скачать в хорошем качестве

Knowledge-Guided Machine Learning for Partitioning Ecosystem Water Fluxes from Flux Tower Data

Sadegh will present KG-DecompNet, a knowledge-guided machine learning framework developed to partition total evapotranspiration (ET) into its primary components: transpiration (T), surface evaporation (Es), and canopy-intercepted evaporation (Ei). Traditional approaches have faced challenges to separate ET components, especially the dynamic, threshold-based behavior of Ei, leading to likely overestimation of T following rainfall or dew events. KG-DecompNet addresses this by integrating physical constraints into site-level machine learning models trained on multi-year, high-frequency turbulence and meteorological data from 35 NEON sites. Speaker - Sadegh Ranjbar: Sadegh is a Postdoctoral Researcher in Paul Stoy’s lab at the University of Wisconsin–Madison’s Biological Systems Engineering department, with a Ph.D. in Systems Engineering. His expertise is in remote sensing and machine learning applied to ecosystem science. His recent work includes near-real-time monitoring of land surface temperature, carbon fluxes, and water cycle partitioning using geostationary satellite data and knowledge-guided ML models.