Automation of gridded HEC-HMS model development using Python : initial condition testing and calibration applications

Abstract

The US Army Corps of Engineers’s (USACE) Hydrologic Engineering Center-Hydrologic Modeling System (HEC-HMS) rainfall-runoff model is widely used within the research community to develop both event-based and continuous rainfall-runoff models. The soil moisture accounting (SMA) algorithm is commonly used for long-term simulations. Depending on the final model setup, 12 to 18 parameters are needed to characterize the modeled watershed’s canopy, surface, soil, and routing processes, all of which are potential calibration parameters. HEC-HMS includes optimization tools to facilitate model calibration, but only initial conditions (ICs) can be calibrated when using the gridded SMA algorithm. Calibrating a continuous SMA HEC-HMS model is an iterative process that can require hundreds of simulations, a time intensive process requiring automation. HEC-HMS is written in Java and is predominantly run through a graphical user interface (GUI). As such, conducting a long-term gridded SMA calibration is infeasible using the GUI. USACE Construction Engineering Research Laboratory (CERL) has written a workflow that utilizes the existing Jython application programming interface (API) to batch run HEC-HMS simulations with Python. The workflow allows for gridded SMA HEC-HMS model sensitivity and calibration analyses to be conducted in a timely manner.