Calibrating the Lyne-Hollick filter for baseflow separation based on catchment response time

• Published in: Journal of hydrology (Amsterdam) 2024-07, Vol.638, p.131483, Article 131483
• Main Authors: Nagy, Eszter D., Szilagyi, Jozsef, Torma, Peter
• Format: Article
• Language: English
• Subjects: Baseflow separation; Catchment response time; Digital recursive filter; Kinematic wave theory; Lag time; Lyne-Hollick filter; Time of concentration; Time to equilibrium
• ISSN: 0022-1694
• DOI: 10.1016/j.jhydrol.2024.131483

•Catchment response time proved to be useful to calibrate the Lyne-Hollick filter.•The calibrated parameter values are higher compared to those published before.•Sensitivity analysis showed that using 12 events is recommended for calibration.•The calibration routine is downloadable for Matlab. The physically meaningful separation of baseflow from the observed streamflow time series is still an unresolved problem in hydrology. This study provides a novel method that builds upon catchment response time to formulate an optimization criterion for calibrating the Lyne-Hollick filter. Catchment response time parameters, including time of concentration (Tc) and time to equilibrium (Te), can be derived from observed precipitation and discharge time series. A range for the ratio of these two has already been derived on a physical basis. In the presented approach, the parameter of the Lyne-Hollick filter is calibrated so that the ratio of Tc and Te falls between a physically plausible range. The proposed method was tested in 25 catchments, and the results were compared to those derived in a previous study. The Pearson correlation coefficient improved from 0.654 to 0.862 between Tc and the lag time when applying the proposed calibration procedure.