the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 02 Nov 2023
Modelling the effects of climate and landcover change on the hydrologic regime of a snowmelt-dominated montane catchment
Abstract. Climate change poses risks to society through the potential to alter peak flows, low flows, and annual runoff yield. Wildfires are projected to increase due to climate change; however, little is known about their combined effects on hydrology. This study models the combined impacts of a climate change scenario and multiple landcover scenarios on the hydrologic regime of a snowmelt-dominated montane catchment, to identify management strategies that mitigate negative impacts from climate and/or landcover change. The combination of climate change and stand replacing landcover disturbance in the middle and high elevations is predicted to advance the timing of the peak flow three to five times more than the advance generated by disturbance alone. The modelling predicts that the combined impacts of climate change and landcover disturbance on peak flow magnitude are generally offsetting for events with return periods less than 5–25 years, but additive for more extreme events. There is a dependency of extreme peak flows on the distribution of landcover. The modelling predicts an increasing importance of rainfall in controlling peak flow response under a changing climate, at the expense of snowmelt influence. Extreme summer low flows are predicted to become commonplace in the future, with most of the change in frequency occurring by the 2050s. Low annual yield is predicted to become more prevalent by the 2050s, but then fully recover or become less prevalent (compared to the current climate) by the 2080s, because of increased precipitation in the fall-spring period. The modelling suggests that landcover disturbance can have a mitigative influence on low water supply. The mitigative influence is predicted to be sustained under a changing climate for annual water yield, but not for low flow. The study results demonstrate the importance of a holistic approach to modelling the hydrological regime rather than focusing on a particular component. Moreover, for managing watershed risk, the results indicate there is a need to carefully evaluate the interplay among environmental variables, the landscape, and the values at risk. Strategies to reduce one risk may increase others, or effective strategies may become less effective in the future.
- Preprint
(5995 KB) - Metadata XML
-
Supplement
(3069 KB) - BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on hess-2023-248', Anonymous Referee #1, 13 Nov 2023
Review for manuscript ‘Modelling the effects of climate and landcover change on the hydrologic regime of a snowmelt-dominated montane catchment’ by Smith et al.
Summary
The manuscript by Smith et al. assesses the combined impacts of climate and landcover scenarios on different hydrological signatures in the Penticton Creek watershed in Canada using the hydrological modeling framework Raven, one climate scenario, and 5 landcover scenarios. The authors find changes in snowmelt seasonality, peak flow, and annual flow yield and conclude that the combined impacts of climate and landcover changes offset changes for flood events with return periods shorter than 25 years. They also find that rainfall becomes a more important influencing factor of peak flows under climate change.
General
I think that this contribution tackles a very important question, i.e. what is the join influence of climate and landcover change on hydrological signatures. However, similarly to the existing literature on the topic, it does not go beyond a case study. While the case study is carefully done and changes in different streamflow signature explained in detail for the one watershed under consideration, the generalizability of results is limited given that existing studies showcase the large variability of hydrologic responses to both climate and landcover changes and their interplay. In addition to not being generalizable to other regions, the results are also quite predictable given the existing literature: they point to earlier snowmelt, earlier flood peaks and an increasing influence of precipitation as we move into the future. While I do not see how the current study advances our knowledge related to future changes in streamflow signatures and the interplay between climate and landcover influences beyond the study region, I acknowledge the detailed and well-presented results for the case study watershed.
Other major comments
1. I find the methods descriptions detailed but rather superficial. That is, while the most important steps of the modeling framework are named, many methodological specificities remain unclear. A few examples:
- What is the temporal resolution of the streamflow data used for the analysis? (p.4, l. 98)
- Do the percentage changes in forest cover refer to the entire catchment area or just the forested catchment area (the latter would be more logical in my opinion)? (p.7, l.124-125)
- How were precipitation and temperature interpolated from station data to areal data? (p.7, l. 141)
- Which algorithm was used to estimate the full snowpack energy balance? (p.7, l.144)
- How was the historical streamflow record adjusted for storage changes in Greyback Lake? (p.8, l.152)
2. The climate impact assessment relies on one climate scenario (i.e. GCM and emission scenario combination) only, neglecting uncertainties related to emission scenario and GCM choice. While this limitation is acknowledged in the discussion section, I find that it could be overcome relatively easily by running the model for a few more climate scenarios. Furthermore, the model used for the analysis should be better contextualized within the sample of existing models (see Section 4.4.) by comparing its temperature and precipitation changes to those of other existing models.
3. The authors use a weather generator on the climate simulations to increase sample size (Section 2.2.2.3), which is per-se a good thing. However, it is unclear why these simulations are limited to 100-years given that the focus is among other variables on extreme events, which requires larger sample sizes to separate signal from noise.
Minor comments
- Use superscripts for units such as km2 and m3/s
- The discussion talks quite a bit about risk (Section 4.3). However, the authors do just look at changes in hazard while changes in vulnerability and exposure are not assessed. To avoid confusion, I would therefore use more specific terminology.
Citation: https://doi.org/10.5194/hess-2023-248-RC1 - AC1: 'Reply on RC1', Russell Smith, 02 Feb 2024
-
RC2: 'Comment on hess-2023-248', Anonymous Referee #2, 06 Dec 2023
General comments: Smith et al. applied a modelling approach to evaluate the combined effects of climate and landcover changes on the hydrologic regime of a snowmelt-dominated montane catchment. Based on the modelling results, management strategies that mitigate the negative impacts were identified. The study is regionally focussed on the Penticton Creek catchment, Canada, a catchment that was already in the focus of other studies before (e.g., Winkler et al., 2017; Winkler et al., 2021; Spittlehouse and Dymond, 2022; Smith, 2018, 2022). The authors address an important topic, in particular the interaction of the effects of forest fires (landcover) and climate change on hydrological regimes. Both will become increasingly important in the future.
Overall, the manuscript is well-written and provides insights on annual and seasonal hydrological changes under RCP 8.5 global warming and different landcover conditions. The percentage of disturbed area because of historical wildfires and forest harvesting was considered in landcover scenarios. In total one climate change scenario and 5 landcover change scenarios were parameterized and simulated with the Raven model. Figures and tables clearly show the results of various analyses.
The study is scientifically sound and based on methods that are well established in the scientific community. Although very case specific, the model results are useful for the development of water management strategies.
Individual scientific issues:
- Line 46-47: Are findings for the Penticton Creek comparable? Results from Westra et al. (2013) are not discussed further.
- Line 129: please add classification system for lower, middle, and higher elevations (in meter).
- Table 1: was data from Penticton Airport met. station used in the study, if so, how was it considered?
- Line 141-142: How were data spatially distributed?
- Line 143: What algorithms were used (degree-day method?)
- Line 153: How was Greyback Lake considered in the model. This is not clear from the text.
Section 2.2.2:
- Historical weather data (T, P) from one station located in the upstream area were used for model parameterization. Does the P1 station in the upper part of the catchment dominate the hydrological regime downstream? Should be discussed in uncertainty section.
- One climate change scenario (RCP 8.5) from one GCM (CSIRO) was considered in this study. The selection and (benefit?) of just one climate change scenario is discussed in the uncertainty section but is not based on other scientific results. Different patterns from other GCMs and timing (GCM and RCP dependent) might lead to different results. Here, the authors should explain in more detail why this scenario was selected and the benefit for this study.
- However, synthetic time series of 100 years on air temperature (min, max) and precipitation were prepared to simulate 100 years of river discharge and SWE. How close are the synthetic time periods to the GCM output? OR vice versa, how strong are baseline and future GCM time series biased to (fixed by) observations?
- Regarding climate change, why were future GCM data on solar radiation and wind speed (wind direction) not used although these are key drivers of ET and snow storage processes?
Section 2.2.3:
- Landcover scenarios seem to be static for a given year or vegetation status – is this correct? A 100-year time series of climate change input was combined with one-year landcover condition. If this is assumed, then it should be clearly stated in the text. [unfortunately, the reference Spittlehouse and Dymond is not freely accessible]
- Line 200: What method was used for the interpolation? Please add!
- Line 221ff: Sorry, I've some difficulties in understanding how Fig. 3 was generated, particularly 3b and 3 c. Fig. 3b and c look different compared to the historical wildfires (Fig. 2g). There were wildfires in the outlet region in the past, but not visible in Fig. 3c?
Section 2.3.4:
- Line 252-253: any reference that supports this suggestion? How is net negative snowfall interception considered in the model?
- Line 259: What other model parameters were based on empirical observations? Please add examples to the manuscript or further descriptions to the SI.
Section 2.3.1:
- This sub-section needs more explanation. How were empirical relations developed and for which time period? Although this might be described somewhere else, more details are desirable here.
Section 2.3.2:
- Line 295: ESSFdc1 = ESSFL? Please correct (in SI as well)
- Line 301: this sentence is interesting. I would expect that throughfall increases with increasing precipitation (amount and duration), since more water drains towards the floor. In terms of forest hydrology, wind speed and wind direction are also key drivers of throughfall and its spatial variability. Please explain.
Section 2.3.3:
- Which parameters were calibrated/optimized and how? Single parameters or multiple parameters at the same time? Are these parameters sensitive in variation?
- Line 336: Refer to section on climate change scenarios.
- Line 360: Environmental risk is also impacted by shifts in seasonality.
Section 3.2:
- Table 3: This table shows the mean rainfall (mm). For current conditions, e.g., 10 mm in winter = 10 mm over 3 months period excluding snow? Is the number/unit correct? Differs a lot from numbers for winter as given in Table 4 (although net precip., numbers are much higher)
- Future numbers need to be compared to the baseline. Please add baseline numbers. Incremental change means absolute change?
Section 3.4:
- Regarding peak flow and summer low flow conditions, how were the highest and lowest flow events identified (Figs. 10 and 11)? Do the figures represent a graph of a specific year in which the highest or lowest event took place? Or were the highest or lowest values per month or day selected? (Same for lowest annual discharge, Fig. 12) The authors may think about
- The authors should consider not splitting chapter 3.4 into sub-chapters, as each sub-chapter begins with the same wording and is structured in the same way. The results can be presented more concisely, Figure 12 can be moved to the SI.
Section 4.1:
- Line 667: What is meant by rainstorm size?
Section 4.2:
- Line 705: What does the text “This mitigative influence …” refer to? Are the large burn condition the mitigative influence?
- Line 730: Summer net precipitation is negative and will further decrease. Where does the water come from that drives an increasing ET?
Section 4.3:
- Hydrological risk – I am not sure if “risk” is the best word to be used here. Risk means information on the exposure and vulnerability. A probability of the occurrence of floods and low flows is provided, however, no information is provided on e.g. damages or losses.
- Line 785: How certain are scenario runs for the 2050s?
Section 4.4
- Line 786-794: Input from one GCM and one RCP was used to drive the hydrological model. Best guesses on a lower RCP 4.5 are described under this section without any references or model results. It is most likely that RCP 8.5 leads to higher peak flows and lower summer discharges compared to RCP 4.5 or RCP 2.6. Of course, the argumentation holds true for results expected from lower emission scenarios, but different GCMs may lead to different precipitation patterns and intensities. In terms of identifying management strategies, it is questionable if “a worst-case scenario” is the best choice. I believe it would be advisable to draw conclusions for (sustainable) management from model results driven by the input of a climate ensemble to identify a robust solution.
Citation: https://doi.org/10.5194/hess-2023-248-RC2 - AC2: 'Reply on RC2', Russell Smith, 02 Feb 2024
Status: closed
-
RC1: 'Comment on hess-2023-248', Anonymous Referee #1, 13 Nov 2023
Review for manuscript ‘Modelling the effects of climate and landcover change on the hydrologic regime of a snowmelt-dominated montane catchment’ by Smith et al.
Summary
The manuscript by Smith et al. assesses the combined impacts of climate and landcover scenarios on different hydrological signatures in the Penticton Creek watershed in Canada using the hydrological modeling framework Raven, one climate scenario, and 5 landcover scenarios. The authors find changes in snowmelt seasonality, peak flow, and annual flow yield and conclude that the combined impacts of climate and landcover changes offset changes for flood events with return periods shorter than 25 years. They also find that rainfall becomes a more important influencing factor of peak flows under climate change.
General
I think that this contribution tackles a very important question, i.e. what is the join influence of climate and landcover change on hydrological signatures. However, similarly to the existing literature on the topic, it does not go beyond a case study. While the case study is carefully done and changes in different streamflow signature explained in detail for the one watershed under consideration, the generalizability of results is limited given that existing studies showcase the large variability of hydrologic responses to both climate and landcover changes and their interplay. In addition to not being generalizable to other regions, the results are also quite predictable given the existing literature: they point to earlier snowmelt, earlier flood peaks and an increasing influence of precipitation as we move into the future. While I do not see how the current study advances our knowledge related to future changes in streamflow signatures and the interplay between climate and landcover influences beyond the study region, I acknowledge the detailed and well-presented results for the case study watershed.
Other major comments
1. I find the methods descriptions detailed but rather superficial. That is, while the most important steps of the modeling framework are named, many methodological specificities remain unclear. A few examples:
- What is the temporal resolution of the streamflow data used for the analysis? (p.4, l. 98)
- Do the percentage changes in forest cover refer to the entire catchment area or just the forested catchment area (the latter would be more logical in my opinion)? (p.7, l.124-125)
- How were precipitation and temperature interpolated from station data to areal data? (p.7, l. 141)
- Which algorithm was used to estimate the full snowpack energy balance? (p.7, l.144)
- How was the historical streamflow record adjusted for storage changes in Greyback Lake? (p.8, l.152)
2. The climate impact assessment relies on one climate scenario (i.e. GCM and emission scenario combination) only, neglecting uncertainties related to emission scenario and GCM choice. While this limitation is acknowledged in the discussion section, I find that it could be overcome relatively easily by running the model for a few more climate scenarios. Furthermore, the model used for the analysis should be better contextualized within the sample of existing models (see Section 4.4.) by comparing its temperature and precipitation changes to those of other existing models.
3. The authors use a weather generator on the climate simulations to increase sample size (Section 2.2.2.3), which is per-se a good thing. However, it is unclear why these simulations are limited to 100-years given that the focus is among other variables on extreme events, which requires larger sample sizes to separate signal from noise.
Minor comments
- Use superscripts for units such as km2 and m3/s
- The discussion talks quite a bit about risk (Section 4.3). However, the authors do just look at changes in hazard while changes in vulnerability and exposure are not assessed. To avoid confusion, I would therefore use more specific terminology.
Citation: https://doi.org/10.5194/hess-2023-248-RC1 - AC1: 'Reply on RC1', Russell Smith, 02 Feb 2024
-
RC2: 'Comment on hess-2023-248', Anonymous Referee #2, 06 Dec 2023
General comments: Smith et al. applied a modelling approach to evaluate the combined effects of climate and landcover changes on the hydrologic regime of a snowmelt-dominated montane catchment. Based on the modelling results, management strategies that mitigate the negative impacts were identified. The study is regionally focussed on the Penticton Creek catchment, Canada, a catchment that was already in the focus of other studies before (e.g., Winkler et al., 2017; Winkler et al., 2021; Spittlehouse and Dymond, 2022; Smith, 2018, 2022). The authors address an important topic, in particular the interaction of the effects of forest fires (landcover) and climate change on hydrological regimes. Both will become increasingly important in the future.
Overall, the manuscript is well-written and provides insights on annual and seasonal hydrological changes under RCP 8.5 global warming and different landcover conditions. The percentage of disturbed area because of historical wildfires and forest harvesting was considered in landcover scenarios. In total one climate change scenario and 5 landcover change scenarios were parameterized and simulated with the Raven model. Figures and tables clearly show the results of various analyses.
The study is scientifically sound and based on methods that are well established in the scientific community. Although very case specific, the model results are useful for the development of water management strategies.
Individual scientific issues:
- Line 46-47: Are findings for the Penticton Creek comparable? Results from Westra et al. (2013) are not discussed further.
- Line 129: please add classification system for lower, middle, and higher elevations (in meter).
- Table 1: was data from Penticton Airport met. station used in the study, if so, how was it considered?
- Line 141-142: How were data spatially distributed?
- Line 143: What algorithms were used (degree-day method?)
- Line 153: How was Greyback Lake considered in the model. This is not clear from the text.
Section 2.2.2:
- Historical weather data (T, P) from one station located in the upstream area were used for model parameterization. Does the P1 station in the upper part of the catchment dominate the hydrological regime downstream? Should be discussed in uncertainty section.
- One climate change scenario (RCP 8.5) from one GCM (CSIRO) was considered in this study. The selection and (benefit?) of just one climate change scenario is discussed in the uncertainty section but is not based on other scientific results. Different patterns from other GCMs and timing (GCM and RCP dependent) might lead to different results. Here, the authors should explain in more detail why this scenario was selected and the benefit for this study.
- However, synthetic time series of 100 years on air temperature (min, max) and precipitation were prepared to simulate 100 years of river discharge and SWE. How close are the synthetic time periods to the GCM output? OR vice versa, how strong are baseline and future GCM time series biased to (fixed by) observations?
- Regarding climate change, why were future GCM data on solar radiation and wind speed (wind direction) not used although these are key drivers of ET and snow storage processes?
Section 2.2.3:
- Landcover scenarios seem to be static for a given year or vegetation status – is this correct? A 100-year time series of climate change input was combined with one-year landcover condition. If this is assumed, then it should be clearly stated in the text. [unfortunately, the reference Spittlehouse and Dymond is not freely accessible]
- Line 200: What method was used for the interpolation? Please add!
- Line 221ff: Sorry, I've some difficulties in understanding how Fig. 3 was generated, particularly 3b and 3 c. Fig. 3b and c look different compared to the historical wildfires (Fig. 2g). There were wildfires in the outlet region in the past, but not visible in Fig. 3c?
Section 2.3.4:
- Line 252-253: any reference that supports this suggestion? How is net negative snowfall interception considered in the model?
- Line 259: What other model parameters were based on empirical observations? Please add examples to the manuscript or further descriptions to the SI.
Section 2.3.1:
- This sub-section needs more explanation. How were empirical relations developed and for which time period? Although this might be described somewhere else, more details are desirable here.
Section 2.3.2:
- Line 295: ESSFdc1 = ESSFL? Please correct (in SI as well)
- Line 301: this sentence is interesting. I would expect that throughfall increases with increasing precipitation (amount and duration), since more water drains towards the floor. In terms of forest hydrology, wind speed and wind direction are also key drivers of throughfall and its spatial variability. Please explain.
Section 2.3.3:
- Which parameters were calibrated/optimized and how? Single parameters or multiple parameters at the same time? Are these parameters sensitive in variation?
- Line 336: Refer to section on climate change scenarios.
- Line 360: Environmental risk is also impacted by shifts in seasonality.
Section 3.2:
- Table 3: This table shows the mean rainfall (mm). For current conditions, e.g., 10 mm in winter = 10 mm over 3 months period excluding snow? Is the number/unit correct? Differs a lot from numbers for winter as given in Table 4 (although net precip., numbers are much higher)
- Future numbers need to be compared to the baseline. Please add baseline numbers. Incremental change means absolute change?
Section 3.4:
- Regarding peak flow and summer low flow conditions, how were the highest and lowest flow events identified (Figs. 10 and 11)? Do the figures represent a graph of a specific year in which the highest or lowest event took place? Or were the highest or lowest values per month or day selected? (Same for lowest annual discharge, Fig. 12) The authors may think about
- The authors should consider not splitting chapter 3.4 into sub-chapters, as each sub-chapter begins with the same wording and is structured in the same way. The results can be presented more concisely, Figure 12 can be moved to the SI.
Section 4.1:
- Line 667: What is meant by rainstorm size?
Section 4.2:
- Line 705: What does the text “This mitigative influence …” refer to? Are the large burn condition the mitigative influence?
- Line 730: Summer net precipitation is negative and will further decrease. Where does the water come from that drives an increasing ET?
Section 4.3:
- Hydrological risk – I am not sure if “risk” is the best word to be used here. Risk means information on the exposure and vulnerability. A probability of the occurrence of floods and low flows is provided, however, no information is provided on e.g. damages or losses.
- Line 785: How certain are scenario runs for the 2050s?
Section 4.4
- Line 786-794: Input from one GCM and one RCP was used to drive the hydrological model. Best guesses on a lower RCP 4.5 are described under this section without any references or model results. It is most likely that RCP 8.5 leads to higher peak flows and lower summer discharges compared to RCP 4.5 or RCP 2.6. Of course, the argumentation holds true for results expected from lower emission scenarios, but different GCMs may lead to different precipitation patterns and intensities. In terms of identifying management strategies, it is questionable if “a worst-case scenario” is the best choice. I believe it would be advisable to draw conclusions for (sustainable) management from model results driven by the input of a climate ensemble to identify a robust solution.
Citation: https://doi.org/10.5194/hess-2023-248-RC2 - AC2: 'Reply on RC2', Russell Smith, 02 Feb 2024
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 343 | 106 | 17 | 466 | 37 | 10 | 9 |
- HTML: 343
- PDF: 106
- XML: 17
- Total: 466
- Supplement: 37
- BibTeX: 10
- EndNote: 9
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1