Articles | Volume 20, issue 7
https://doi.org/10.5194/hess-20-2611-2016
https://doi.org/10.5194/hess-20-2611-2016
Research article
 | 
04 Jul 2016
Research article |  | 04 Jul 2016

Machine learning methods for empirical streamflow simulation: a comparison of model accuracy, interpretability, and uncertainty in seasonal watersheds

Julie E. Shortridge, Seth D. Guikema, and Benjamin F. Zaitchik

Download

Interactive discussion

Status: closed
Status: closed
AC: Author comment | RC: Referee comment | SC: Short comment | EC: Editor comment
Printer-friendly Version - Printer-friendly version Supplement - Supplement

Peer-review completion

AR: Author's response | RR: Referee report | ED: Editor decision
ED: Publish subject to minor revisions (Editor review) (06 Feb 2016) by Dominic Mazvimavi
AR by Julie Shortridge on behalf of the Authors (15 Feb 2016)  Author's response   Manuscript 
ED: Reconsider after major revisions (24 Feb 2016) by Dominic Mazvimavi
AR by Julie Shortridge on behalf of the Authors (24 Feb 2016)  Author's response   Manuscript 
ED: Referee Nomination & Report Request started (07 Mar 2016) by Dominic Mazvimavi
RR by Anonymous Referee #1 (17 Mar 2016)
RR by Anonymous Referee #2 (03 Apr 2016)
ED: Publish subject to minor revisions (Editor review) (10 Apr 2016) by Dominic Mazvimavi
AR by Julie Shortridge on behalf of the Authors (15 Apr 2016)  Author's response   Manuscript 
ED: Publish as is (29 May 2016) by Dominic Mazvimavi
AR by Julie Shortridge on behalf of the Authors (07 Jun 2016)
Download
Short summary
This paper compares six methods for data-driven rainfall–runoff simulation in terms of predictive accuracy, error structure, interpretability, and uncertainty. We demonstrate that autocorrelation in model errors can result in biased estimates of important values and show how certain model structures can be more easily interpreted to yield insights on physical watershed function. Finally, we explore how model structure can impact uncertainty in climate change sensitivity estimates.