PACE-GLM: Parallel Adaptive Calibration Engine
Team
PACE-GLM dynamically reallocates idle computing threads to active model agents during distributed calibration. Validated on a high-dimensional WRF-Hydro model of the Murderkill River watershed, it reduced wall-clock calibration time and improved model performance.