Assessment of Variations in Runoff Due to Landcover Changes Using the SWAT Model in an Urban River in Dublin, Ireland
Round 1
Reviewer 1 Report
Overall comment
This is an interesting research that focuses on the changes in land use in Ireland that have changed over the past 30 years and analyzes the changes in runoff using SWAT model.
It would be better to have a consideration at what changes in land use have occured acctually and how they made runoff characteristics change.
Also, even runoff pattern will be changed significantly due to climate change, is the impact of land use changes more significant? It is helpful if you give a comment influence of climate change in Ireland like rainfall pattern.
Individual comment
L91; Period after "...defined" is missing.
L113-114; It is helpful if the average gradients of the upper and the lower reaches of the Dodder river are indicated because it relates to runoff characteristics.
Legend of Figure 3; Please show unit of slope like degree or %.
L149; STOP-4 -> SPOT-4?
Figure 4; According to Figure 4, the change from WETL to RNGB in the upstream area appears to be more remarkable than the urbanization in the downstream area. Could you show what happened historically and its reason? Also, in the table 1, WETL is categorized by Peat bogs, Salt marshes, and Salines, while RNGB is Moors and heathland, Sclerophyllous vegetation and Transitional woodland-shrub. Which is actually applicable in the study area?
Legend of Figure 5; What kind of soil does the symbol mean?
L261-281; I suppose that the parameters of SWAT are set as default values for each land use. Are they tuned or not? Is it possible to reproduce properly the runoff without tuning? Please give any comments how to set of the parameters.
Table 3; What are the criteria of NSE for determining best-fit SWAT model? Some values of the best-fit cases are rather small, so it seems that they are not necessarily fittable. There is also a view that NSE of 0.7 or higher is usually evaluated to be fittable, so please clarify the basis of the criteria.
L427-428, Figure 8; Isn't it possible that underestimation of the simulation is because the model parameters are not set properly? If the simulation can not reproduce the observed results properly, impact-response evaluation between land use change and runoff may not be plausible, then please give a comment on that point.
Author Response
Please see the attachment
Author Response File: 
Author Response.pdf
Reviewer 2 Report
This manuscript assessed the variations in runoff due to the landcover changes. There are several issues should be addressed:
There are no real conclusion. From this manuscript, I see the model performances were increased when considering the dynamic of land cover data. However, the NSE of SWAT modeling for monthly mean runoff were mostly below 0.6, seldom above 0.8. It means even consider the time-varying landcover data, the performances were still not satisfied at monthly time scale and the landcover change had limited impact on runoff simulation.
The results did not contain the quantification of impact of land cover changes on runoff in different period.
There are lots of language, grammer, and punctuation mistakes. For example Line 67, 91, 196.
Figure 7, the y-axis may wrong.
Figure 8, the simulated discharges of different land cover show large difference with observation. This difference was even larger than the difference among the simulated discharges of different land cover. It again suggested the model validation and calibration should be improved.
Further reading please reference to :
Linshan Yang, et al. Identifying separate impacts of climate and land use/cover change on hydrological processes in upper stream of Heihe River, Northwest China. Hydrological Processes, 2017, 31: 1100-1112.
Linshan Yang, et al. Separation of the Climatic and Land Cover Impacts on the Flow Regime Changes in Two Watersheds of Northeastern Tibetan Plateau. Advances in Meteorology, 2017, 6310401.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 2 Report
The revison of the manuscript was significant improved, only two things should be further addressed.
Figure 2, add the best simulated discharge time series, using the best simulated discharge for different parts, in order to present the reliable of the SWAT simulation. Or add a new figure relate to scatter plot between observation and best simulation.
The y-axis shouldn't be NSE or other values in Figure 7? otherwise, what are the represents of the difference of points.
Author Response
Please see the attachment
Author Response File: Author Response.pdf
Round 3
Reviewer 2 Report
I recommend to remove the LC 1990, LC2000, etc, in the top of Figure 7, it would be more reasonable.
Added the units of axis for Figure 8, as well as the statistics of the scatterplot, such as R-square, slope.
Author Response
Comment 1: I recommend to remove the LC 1990, LC2000, etc, in the top of Figure 7, it would be more reasonable.
Response to Comment 1: Please note that the y-axis indicates which landcover map has been used to simulate the SWAT model. On the other hand, the landcover maps at the top indicate for each year which map is expected to provide the best result. Hence we believe the landcover details at the top as well as in the y-axis are needed. The title of Figure 7 has been modified to explain this in the revised manuscript.
Comment 2: Added the units of axis for Figure 8, as well as the statistics of the scatterplot, such as R-square, slope.
Response to Comment 2: The units, slope and R2 have been added in Figure 8.
