peer review report 2 on assessing the role of uncertain precipitation estimates on the robustness of hydrological model parameters under highly variable climate conditions

Journal of Hydrology: Regional Studies (2017) 94–95 Contents lists available at ScienceDirect Journal of Hydrology: Regional Studies journal homepage: www.elsevier.com/locate/ejrh Peer Review Report Peer review report on “Assessing the role of uncertain precipitation estimates on the robustness of hydrological model parameters under highly variable climate conditions” Original Submission 1.1 Recommendation Minor Revision Comments to Author: The authors aim to investigate how using various precipitation products influence the robustness of parameter calibration in a non-stationary climate environment in several African river basins Use of a differential split-sample test challenges calibration efforts, as parameters calibrated under certain climate conditions are transferred to another, largely different climate condition Hydrologic performance of different precipitation products is evaluated along with a sensitivity analysis of model parameters for each precipitation product The authors recommend transposing calibrated parameters very cautiously if it is a necessary practice, since there was little consistency in calibration results between the two climate conditions examined While the conclusion of parameters not transferring well to drastically different conditions is hardly surprising, the work presented in the manuscript is interesting and suitable for publication in Journal of Hydrology: Regional Studies as it provides a thorough examination of how precipitation uncertainty can influence calibration efforts This highlights an idea that is largely assumed to be true, but often overlooked in formal experiments Furthermore, the selected pairing of satellite-based observations in a poorly gauged region is nice to see, since there is a natural tendency to focus on data rich regions Overall, the manuscript is well written and fairly easy to follow and the comments I have are merely for consideration and should not entail significant effort Therefore, I recommend it be accepted for publication with minor revision Minor comments: 1) The title doesn’t quite capture the purpose of the article The hydrological model itself is not non-stationary and effort wasn’t focused on allowing the model to change with non-stationary conditions Something to the effect of “Assessing the role of uncertain precipitation estimates on the robustness of hydrological model parameters in a non-stationary environment” may reflect the purpose better (although this is getting lengthy) 2) Section 2.2.1: I am curious as to why the authors didn’t use the available high resolution versions of the precipitation data (for example, km PERSIANN-CSS and CMORPH at km as mentioned) and aggregate to the model 0.1 degree rather than resampling the coarse precipitation data It seems that regridding CMORPH to 0.25 degree and back down to 0.1 degree (for example) adds an extra layer of uncertainty Why not have the common grid be 0.1 degree from the start since that is what the model is ingesting? 3) Same section: could a statement be added about how the resampling was performed at the end of this section? 4) Line 92–93: This sentence would benefit from some rewording DOI of the original article:http://dx.doi.org/10.1016/j.ejrh.2016.09.003 2214-5818/$ – see front matter http://dx.doi.org/10.1016/j.ejrh.2016.12.053 Peer Review Report / Journal of Hydrology: Regional Studies (2017) 94–95 95 5) Line 119: Soil properties from the tă his sentence is a fragment 6) Lines 126–128: The authors mention all datasets are resampled (or in this case upscaled) to 0.1 degree, what is the original resolution of the SRTM? 7) Line 214–217: Fig shows region A2 as having out of correlation coefficients being insignificant Is it still an acceptable choice to use this basin? Perhaps the authors can further justify the decision of its inclusion 8) Line 230: “the” instead of “he” 9) Paragraph starting at Line 291: The summary of the sensitivity analysis is nice, but it would be great if the authors could provide some comment as to why they think these particular parameters (GwLoss, PPrefFlow and CalEvap) are so influential and why the interaction of parameters in model runs for some precipitation products is weak and others are strong 10) Line 330: is months sufficient for a warm-up period for this study area (i.e has there been a sufficient wet/dry cycle to minimize the impact of memory of initial conditions)? 11) Line 613: again, discussion on possibilities as to why these parameters are sensitive would be good 12) Bullet point starting on Line 628: This is a rather strong statement against use CMORPHV1.0, GSMap and PERSIANN in Southern Africa, especially since other uncertainties (particularly those related to model assumptions/structure) are not really accounted for in this study A qualifying statement that this is the case for the LISFLOOD model in particular or at least an acknowledgement of unaccounted for uncertainties is probably in order 13) Line 647: “too” instead of “to” Ph.D Andrea Thorstensen The Henry Samueli School of Engineering, Civil and Environmental Engineering, zotcode 2175, University of California, Irvine Irvine, CA 92697-2175, United States ... use this basin? Perhaps the authors can further justify the decision of its inclusion 8) Line 23 0: ? ?the? ?? instead of “he” 9) Paragraph starting at Line 29 1: The summary of the sensitivity analysis... impact of memory of initial conditions) ? 11) Line 613: again, discussion on possibilities as to why these parameters are sensitive would be good 12) Bullet point starting on Line 628 : This is a rather.. .Peer Review Report / Journal of Hydrology: Regional Studies (20 17) 94–95 95 5) Line 119: “Soil properties from the tă his sentence is a fragment 6) Lines 126 – 128 : The authors mention all