PACE-GLM: Parallel Adaptive Calibration Engine
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.
Related publications
- Street-to-pipe diagnosis of compound rain–tailwater flooding (2026)
- Assessing Hybrid Modeling for Fine-Scale Runoff Prediction (2024)
- Hydrological extremes heighten vulnerability to schistosomiasis (2024)
- A methodological framework for improving the performance of data-driven models (2023)
- Coastal Flooding: Modeling, Monitoring, and Protection Systems (2022)
- Generalization of Runoff Risk Prediction (2022)