Repo for machine learning models for regional prediction of hydrologic forcing functions. Includes probabilistic seasonal high flow regions for CONUS, and prediction of high flow metrics for selected regions.
After correlation screening, the remaining climatic attributes tend to represent 1 per season. That's good we're filtering out redundant months, but it suggests that we could do even better if our attributes were seasonal climate instead of monthly.
After correlation screening, the remaining climatic attributes tend to represent 1 per season. That's good we're filtering out redundant months, but it suggests that we could do even better if our attributes were seasonal climate instead of monthly.