A ‘User-Friendly’ approach to parameter estimation in hydrologic models

The goal of this paper is to analyze the reliability of the Multi-step Automated Calibration Scheme (MACS) over a variety of climatic and hydrologic conditions. The authors developed MACS for the estimation of parameters for hydrologic models; be it for ‘fine-tuning’ of a priori estimates, or for estimating parameters without a priori knowledge of the system. Optimization methods have advanced over the last few decades, and although they are used extensively by the research community, operational hydrologists have been less eager to implement automated calibration procedures. The authors, under cooperative agreement with the National Weather Service (NWS) Hydrology Laboratory have collaborated to develop a progressive calibration strategy, using ‘in-house’ NWS algorithms to optimize parameters for the hydrologic models used in operational streamflow forecasting: the Sacramento Soil Moisture Accounting (SAC-SMA) and the SNOW-17 model. The method, though developed within the NWS forecasting system, can be easily adapted to any hydrologic modeling system. In our current work, MACS has been tested on 20 NWS forecast points (or basins), located in various hydrologic and climatic regimes (five different River Forecast Centers (RFCs)) across the United States. Over half of the basins tested (11) consist of multi-tiered systems in the Western US (i.e. the hydrologic models are run over several elevation zones for one forecast point). The results show comparable, reliable calibrations, similar in quality to the traditional manual techniques, over all of the hydro-climatic regimes used for this study. MACS, generally, produces simulations with desirable performance measures, including improved Nash–Sutcliffe efficiency and lower percent bias. MACS performs well in all regions, even over the complex terrain in the western regions of the United States.