the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 25 Sep 2023
Timing of spring events changes under modelled future climate scenarios in a mesotrophic lake
Abstract. Lakes experience shifts in the timing of processes as a result of climate warming, and especially relative changes in the timing of events may have important ecological consequences. Spring in particular is a period in which many key processes that regulate the ecology and biogeochemistry of lakes occur, and also a time which may experience significant changes under influence of global warming. In this study, we used a coupled catchment-lake model forced by future climate projections to evaluate changes in the timing of spring discharge, ice-off, the spring phytoplankton peak, and the onset of stratification, in a mesotrophic, temperate lake. All these events showed a clear trend towards earlier occurrence with climate warming, with ice cover tending to disappear at the end of the century in the most extreme climate scenario. Moreover, relative shifts in the timing of these springtime events also occurred, with the onset of stratification tending to advance slower than the other events, and the spring phytoplankton peak and ice-off advancing faster in the most extreme climate scenario. The outcomes of this study stress the impact of climate change on the phenology of processes in lakes and especially the relative shifts in timing during spring. This can have profound effects on food-web dynamics as well as other regulatory processes, and influence the lake for the remainder of the growing season.
-
Notice on discussion status
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
-
Preprint
(819 KB)
-
Supplement
(1646 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(819 KB) - Metadata XML
-
Supplement
(1646 KB) - BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-1679', Anonymous Referee #1, 31 Oct 2023
General Comments
- Overall, this is a well-written and important contribution to our understanding of changing ecosystem functioning in lakes. It provides new insight by both developing projections of biogeochemical variables like phytoplankton dynamics and a novel comparison of the relative change in timing of important limnological events. The methods and results are clearly presented and the research is well contextualized. I suggest a few improvements below to better present the research in the context of other studies and research within the field of limnology.
Specific Comments
- If you’re not limited by words already, consider adding a sentence in the abstract that states how well your model did during training/validation to add support to the validity of your projections
- In the first few sentences of the intro, can you add some language to make it crystal clear whether the studies you are citing demonstrated already observed changes or projections in timing of processes? On a glance, I think most of the studies you cite are observed already and adding a short paragraph that more thoroughly summarizes findings from other projection studies would help highlight the novelty of your approach (including phytos and catchment loading and comparing relative changes in annual timing events across multiple variables)
- I think you are using spring ‘metrics’, ‘events’, ‘processes’ interchangeably to refer to your four response variables in the intro—might be good to choose one and stick with that
- You introduce some really good, but new, content in the last paragraph of the intro (line 59 on) about why relative differences in the timing of spring events matters. I wonder if you could make this its own paragraph before you introduce your study and hypotheses so that you can expand a bit more on why relative shifts in timing matter—this is the key finding from your study so it should be emphasized heavily in the intro
- You might want to add a citation in the introduction somewhere to Adrian et al. 2012 who discuss how changes in climate drivers during key time periods are critical to informing overall ecosystem function: Adrian, R., Gerten, D., Huber, V. et al.Windows of change: temporal scale of analysis is decisive to detect ecosystem responses to climate change. Mar Biol 159, 2533–2542 (2012). https://doi.org/10.1007/s00227-012-1938-1
- Line 42: I suggest remove ‘in this study’ phrasing and focus on why these metrics are important generally in this paragraph before you emphasize the details of your study specifically
- Methods, line 97-102: can you provide reference to any other studies which use biogeochemical process models and have similar R2 for reproducing observations? I’m not implying that the fit isn’t good enough, just that comparing to what others have done would be helpful to justify some of the lower R2 values
- Line 108: can you provide a date range for the historical record of ice-off dates? Also in this section, can you report the bias for your 2C threshold for simulating ice-off for comparison since you report the error using the ice module?
- Line 145: include a citation for your workflow here as well?
- Line 148-149: a sentence similar to this would add strength to the abstract in demonstrating that your model performed well against observations. I would suggest adding the years of this calibration/validation time period here (not necessary for abstract though I think)
- Results, line 170: maybe just me, but I’m not familiar with the term ‘shoal’. Could you rephrase as ‘increase’ or ‘decrease’?
- Table 1: Is there a way you could visualize this rather than providing a table (but perhaps keep table in SI)? I’m envisioning something similar to Figure 2 where you show the difference between the value at the beginning of the simulation (intercept) and the mean value at the end of the projection time period based on Sen’s slope? This would allow you to highlight the directionality and magnitude of average change
- Figure 2: can you make the font size overall a bit larger? It is necessary for me to zoom in quite a bit to read it as is. Would also suggest adding panel labels if this is a journal requirement. Instead of the purple square, maybe could you make the red diamonds open for years with a bad fit, filled for years with a good fit? The square is a little distracting (not a major issue though)
- Line 215: is there a figure you can reference to support this? As it’s written, it’s unclear if you mean under current conditions or under future projections
- Line 245: this is really interesting. Did you calculate chlorophyll-a concentrations later in the growing season or just spring? I am wondering if there is an antecedent effect for later in the year which could have broader implications for additional bloom events and could be useful to add to the discussion
- Line 272: I think you should emphasize that this is especially true for biological responses like chla (there are studies looking at multiple connected hydrodynamic processes, Ayala et al. 2020, Barbosa et al. 2021, Feldbauer et al. 2022, Desgue-Itier et al. 2023, Wynne et al. 2023, etc.)
- I think the study could benefit from more discussion of the implicit assumptions from focusing on spring event timing as your response variables (e.g., instead of summer, winter, or fall events). You do a good job justifying why spring is important (and I believe it), but I think you could add context which highlights other research which shows that antecedent conditions during other time periods (e.g., winter-time dynamics, storm events) are important for year-round functioning and adding some context to acknowledge this would be helpful in the discussion. Some potential citations
- Cavaliere et al. 2021 https://doi.org/10.1029/2020JG006165
- Adrian et al. 2012 https://doi.org/10.1007/s00227-012-1938-1
- Thayne et al. 2021: https://aslopubs.onlinelibrary.wiley.com/doi/full/10.1002/lno.11859
Citation: https://doi.org/10.5194/egusphere-2023-1679-RC1 - AC1: 'Reply on RC1', Jorrit Mesman, 08 Dec 2023
-
RC2: 'Comment on egusphere-2023-1679', Xiangzhen Kong, 03 Nov 2023
Review report for manuscript titled “Timing of spring events changes under modelled future climate scenarios in a mesotrophic lake”
The present study elaborated to investigate the future climate impacts on the spring hydrological and ecological processes (i.e. spring discharge, ice-off, spring phytoplankton peak, onset of stratification) in a typical temperate lake Erken. The findings have critical implications because these processes were rarely evaluated simultaneously and their different sensitivities to climate change may result in different change paces or rates, and eventually lead to profound consequences on lake ecosystem in the future. The paper is well-prepared and concisely written. I have a few major and minor concerns and would like to recommend publication if the authors can address them properly during the revision.
Major comments:
- The manuscript repeatedly emphasizes the ecological consequences of the different rate of advancing among the four investigated events, but these are not actually evaluated and subject to inferences and speculation, which may be attributed to the limitation in the model. This can be compensated, and manuscript can be improved, if the author can add a conceptual diagram in the discussion section, which summarizes the findings from the study (i.e. different advancing rate among the processes, increasing gap between stratification onset and other processes, and the potential ecological consequences from literature, for example, increasing magnitude of winter diatom blooms, see e.g., Hebert et al., 2021 e2114840118 PNAS , or Kong et al, 2021, 190, 116681 Water Research). Please consider this suggestion during the revision.
- It is confusing to learn that the model did not catch the actual dynamics in certain years (Fig. 2). Despite the reason of the methodology or definition of the events, it would be necessary to provide an acceptable explanation for these ‘bad’ years not only in the supplements but also in the main text. For example, are these bad years featured by hydrological or climatic extremes? or there were malfunctions of the sampling infrastructure?
- The mixture terminology of ‘processes’ and ‘events’ should be reconciled. Are there any differences? If not, please avoid switching terms and be consistent throughout the text. It would facilitate reading if all the ‘events’ changed to ‘processes’, or vice versa.
Minor comments
- Abstract, please specify which ‘process’ are referring here at the very beginning (e.g. eco-hydrological processes).
- Line 45, if these processes are well acknowledged to be interlinked and occurs in causality and order already, what is the rationale to study them together? I think it should be further stressed that these processes have different sensitivity to climate change, and may response asynchronously in the future with changing orders and causal linkage. As a result, we must evaluate them together.
- Line 52-53, what are the four processes? Please specify, or define them earlier with a clear name and use this name thoughout the text.
- Please summarize the main hypotheses and/or research questions with bullets by the end of introduction section.
- Please increase the font size in Figure 2.
- Figure 4, if I understand correctly, the color represents the ‘ratio of the slope’, rather than the slope itself. Please correct the title of the legend bar to avoid any confusion.
- Line 212, it is intriguing to see that the onset of stratification is always later than the Chl-a peak event, even at the very beginning of the simulation in 1985 (Fig. 3). Conventionally, as already stated in the introduction, the onset of stratification is a prerequisite for the spring phytoplankton bloom (Line 49). Are there any observations in Lake Erken, that the current or previous spring phytoplankton blooms were already earlier than the onset of stratification since 1985? Are these species diatom, according to field data and model predictions? Overall, it is necessary to add a few more explanations here.
Citation: https://doi.org/10.5194/egusphere-2023-1679-RC2 - AC2: 'Reply on RC2', Jorrit Mesman, 08 Dec 2023
-
RC3: 'Comment on egusphere-2023-1679', Love Raman Vinna, 06 Nov 2023
Review of “Timing of spring events changes under modelled future climate scenarios in a mesotrophic lake. By Jorrit P. Mesman, Inmaculada C. Jiménez-Navarro, Ana I. Ayala, Javier Senent-Aparicio, Dennis Trolle, Don Pierson”
In this manuscript the authors use a one way coupled drainage area to lake model setup to investigate future climate impact on spring processes including ice-off, 50% cumulative spring discharge, spring phytoplankton bloom and stratification onset. The bold and novel model setup include stream flow, nutrients and temperature (SWAT+, LOADEST, air2water) coupled to lake physics and biogeochemistry (GOTM, WET). The important findings of the authors show how the occurrence of important spring processes are occurring earlier in a future warmer climate. The manuscript is in a good order but would benefit from extra clarity, sliming down and expansion as my points hereunder show.
1 This manuscript continue and analyze deeper the effect of climate from the work done in Jiménez-Navarro et al. (2023). The reader needs to clearly understand what is the difference between the two works, both in regard to which questions are being addressed here as well as be given all relevant information for spring processes. This point runs throughout the rest of this review.
2 The method description need to be expanded and put in line with Jiménez-Navarro et al. (2023). Among other things I cannot see how many parameters was used in air2stream, which is not a statistical model, it is a semi deterministic model (a hybrid process-based and data-driven model). Additionally, more detailed information regarding the GOTAM-WET model coupling is required. One of the things I miss is how transparency in the lake is modeled/treated. Do the biological model adjust lake transparency, and how do this affect spring bloom and stratification onset? And how do the coupled model preform at deeper depth? The reader can now only see what happens at 3 m depth.
3 the authors struggle with model correctness, needing to use a surface temperature threshold for ice-off despite having an ice module and need to explain discrepancies in stratification onset and chlorophyll spring peak. I ask myself how this can be and have some points here which might enlighten the manuscript. First do the grid resolution compared to measurement resolution affect the results? From Figure S5 timing of stratification it looks like the vertical lines denoting stratification onset do not match the data and should in fact be earlier for the measurements (red line crossing green threshold before timing of stratification). Is this due to a too short window for continuing stratification, is there an error in the script, or do the resolution play a role? As for data. The one way coupled catchment and lake model setup was calibrated from 2000 to 2015 for the lake part and from 2007 to 2015 for the river part. Is the difference in calibration period affecting the results? Looking at Figure 2 for Ice-off this looks to be the case. And how do you deal with the 2000 to 2006 period in regard to river input into the lake model? Building on this, can the less than ideal model correctness be explained by the location of measurements in and above the lake? Lake measurements come from a station at the deepest point in the lake ca 400 m from the eastern shore. This distance might be far enough away for near-shore processes to play a role, but are the location representative for the overall lake physics covering the central parts of the lake? Additionally but not required for this manuscript, it would have improved the results if the complete time frame of available data was considered for calibration, validation (if deemed necessary) could have been carried out in the start and not the end of available measurements see ex. Shen, H., Tolson, B. A. & Mai, J. Time to Update the Split‐Sample Approach in Hydrological Model Calibration. Water Resour. Res. 58, (2022). https://doi.org/10.1029/2021WR031523.
4 Lake processes are heavily dependent on local atmospheric conditions, so to for the drainage area processes. The authors used five GCM models which by their global nature are course resolved. The GCMs are bias correction toward local measurements in Jiménez-Navarro et al. (2023), but if I understand Supplementary table E 4th column (RMSE) this bias correction is almost nonexistent. Taking the difference between GCM INM-CM5-0 and measurements as an example, mean air temperature RMSE (Root Mean Square Error) drops from the unbiased comparison of 5.712 oK to 5.687 oK after bias correction and for Wind Speed from 4.283 to 2.588 m/s, and improvement with <1% and ~40 % respectively. The bias correction of precipitation, a key input to the drainage area model, looks to have failed. Now I might misunderstand how the Bias correction results are shown, but this illustrate my first point. Can we trust that the calibration is still valid using the climate models as input? Additionally, the reader needs to know why these climate models and scenarios were selected. I suspect because they cover the extreme ranges of for example temperature, precipitation, wind speed etc.. Furthermore since the setup is used for projecting climate effects, is the time frame (for drainage area and lake) long enough so that the models capture the climate trend (which is small compared to seasonal variations)? It would help the reader to see how the trends during the setup/calibration period are in the model compared to measurements.
5 Through the analysis of trends from the climate simulations, the authors treat the climate scenarios as constant change over time ex. Fig 3. This is not correct, in fact the gradient for each scenario change over time, especially for SSP 245. I suggest dividing the model output into 30 year chunks while conducting the analysis, or look at the amount of change from a reference to a far future period.
Citation: https://doi.org/10.5194/egusphere-2023-1679-RC3 - AC3: 'Reply on RC3', Jorrit Mesman, 08 Dec 2023
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-1679', Anonymous Referee #1, 31 Oct 2023
General Comments
- Overall, this is a well-written and important contribution to our understanding of changing ecosystem functioning in lakes. It provides new insight by both developing projections of biogeochemical variables like phytoplankton dynamics and a novel comparison of the relative change in timing of important limnological events. The methods and results are clearly presented and the research is well contextualized. I suggest a few improvements below to better present the research in the context of other studies and research within the field of limnology.
Specific Comments
- If you’re not limited by words already, consider adding a sentence in the abstract that states how well your model did during training/validation to add support to the validity of your projections
- In the first few sentences of the intro, can you add some language to make it crystal clear whether the studies you are citing demonstrated already observed changes or projections in timing of processes? On a glance, I think most of the studies you cite are observed already and adding a short paragraph that more thoroughly summarizes findings from other projection studies would help highlight the novelty of your approach (including phytos and catchment loading and comparing relative changes in annual timing events across multiple variables)
- I think you are using spring ‘metrics’, ‘events’, ‘processes’ interchangeably to refer to your four response variables in the intro—might be good to choose one and stick with that
- You introduce some really good, but new, content in the last paragraph of the intro (line 59 on) about why relative differences in the timing of spring events matters. I wonder if you could make this its own paragraph before you introduce your study and hypotheses so that you can expand a bit more on why relative shifts in timing matter—this is the key finding from your study so it should be emphasized heavily in the intro
- You might want to add a citation in the introduction somewhere to Adrian et al. 2012 who discuss how changes in climate drivers during key time periods are critical to informing overall ecosystem function: Adrian, R., Gerten, D., Huber, V. et al.Windows of change: temporal scale of analysis is decisive to detect ecosystem responses to climate change. Mar Biol 159, 2533–2542 (2012). https://doi.org/10.1007/s00227-012-1938-1
- Line 42: I suggest remove ‘in this study’ phrasing and focus on why these metrics are important generally in this paragraph before you emphasize the details of your study specifically
- Methods, line 97-102: can you provide reference to any other studies which use biogeochemical process models and have similar R2 for reproducing observations? I’m not implying that the fit isn’t good enough, just that comparing to what others have done would be helpful to justify some of the lower R2 values
- Line 108: can you provide a date range for the historical record of ice-off dates? Also in this section, can you report the bias for your 2C threshold for simulating ice-off for comparison since you report the error using the ice module?
- Line 145: include a citation for your workflow here as well?
- Line 148-149: a sentence similar to this would add strength to the abstract in demonstrating that your model performed well against observations. I would suggest adding the years of this calibration/validation time period here (not necessary for abstract though I think)
- Results, line 170: maybe just me, but I’m not familiar with the term ‘shoal’. Could you rephrase as ‘increase’ or ‘decrease’?
- Table 1: Is there a way you could visualize this rather than providing a table (but perhaps keep table in SI)? I’m envisioning something similar to Figure 2 where you show the difference between the value at the beginning of the simulation (intercept) and the mean value at the end of the projection time period based on Sen’s slope? This would allow you to highlight the directionality and magnitude of average change
- Figure 2: can you make the font size overall a bit larger? It is necessary for me to zoom in quite a bit to read it as is. Would also suggest adding panel labels if this is a journal requirement. Instead of the purple square, maybe could you make the red diamonds open for years with a bad fit, filled for years with a good fit? The square is a little distracting (not a major issue though)
- Line 215: is there a figure you can reference to support this? As it’s written, it’s unclear if you mean under current conditions or under future projections
- Line 245: this is really interesting. Did you calculate chlorophyll-a concentrations later in the growing season or just spring? I am wondering if there is an antecedent effect for later in the year which could have broader implications for additional bloom events and could be useful to add to the discussion
- Line 272: I think you should emphasize that this is especially true for biological responses like chla (there are studies looking at multiple connected hydrodynamic processes, Ayala et al. 2020, Barbosa et al. 2021, Feldbauer et al. 2022, Desgue-Itier et al. 2023, Wynne et al. 2023, etc.)
- I think the study could benefit from more discussion of the implicit assumptions from focusing on spring event timing as your response variables (e.g., instead of summer, winter, or fall events). You do a good job justifying why spring is important (and I believe it), but I think you could add context which highlights other research which shows that antecedent conditions during other time periods (e.g., winter-time dynamics, storm events) are important for year-round functioning and adding some context to acknowledge this would be helpful in the discussion. Some potential citations
- Cavaliere et al. 2021 https://doi.org/10.1029/2020JG006165
- Adrian et al. 2012 https://doi.org/10.1007/s00227-012-1938-1
- Thayne et al. 2021: https://aslopubs.onlinelibrary.wiley.com/doi/full/10.1002/lno.11859
Citation: https://doi.org/10.5194/egusphere-2023-1679-RC1 - AC1: 'Reply on RC1', Jorrit Mesman, 08 Dec 2023
-
RC2: 'Comment on egusphere-2023-1679', Xiangzhen Kong, 03 Nov 2023
Review report for manuscript titled “Timing of spring events changes under modelled future climate scenarios in a mesotrophic lake”
The present study elaborated to investigate the future climate impacts on the spring hydrological and ecological processes (i.e. spring discharge, ice-off, spring phytoplankton peak, onset of stratification) in a typical temperate lake Erken. The findings have critical implications because these processes were rarely evaluated simultaneously and their different sensitivities to climate change may result in different change paces or rates, and eventually lead to profound consequences on lake ecosystem in the future. The paper is well-prepared and concisely written. I have a few major and minor concerns and would like to recommend publication if the authors can address them properly during the revision.
Major comments:
- The manuscript repeatedly emphasizes the ecological consequences of the different rate of advancing among the four investigated events, but these are not actually evaluated and subject to inferences and speculation, which may be attributed to the limitation in the model. This can be compensated, and manuscript can be improved, if the author can add a conceptual diagram in the discussion section, which summarizes the findings from the study (i.e. different advancing rate among the processes, increasing gap between stratification onset and other processes, and the potential ecological consequences from literature, for example, increasing magnitude of winter diatom blooms, see e.g., Hebert et al., 2021 e2114840118 PNAS , or Kong et al, 2021, 190, 116681 Water Research). Please consider this suggestion during the revision.
- It is confusing to learn that the model did not catch the actual dynamics in certain years (Fig. 2). Despite the reason of the methodology or definition of the events, it would be necessary to provide an acceptable explanation for these ‘bad’ years not only in the supplements but also in the main text. For example, are these bad years featured by hydrological or climatic extremes? or there were malfunctions of the sampling infrastructure?
- The mixture terminology of ‘processes’ and ‘events’ should be reconciled. Are there any differences? If not, please avoid switching terms and be consistent throughout the text. It would facilitate reading if all the ‘events’ changed to ‘processes’, or vice versa.
Minor comments
- Abstract, please specify which ‘process’ are referring here at the very beginning (e.g. eco-hydrological processes).
- Line 45, if these processes are well acknowledged to be interlinked and occurs in causality and order already, what is the rationale to study them together? I think it should be further stressed that these processes have different sensitivity to climate change, and may response asynchronously in the future with changing orders and causal linkage. As a result, we must evaluate them together.
- Line 52-53, what are the four processes? Please specify, or define them earlier with a clear name and use this name thoughout the text.
- Please summarize the main hypotheses and/or research questions with bullets by the end of introduction section.
- Please increase the font size in Figure 2.
- Figure 4, if I understand correctly, the color represents the ‘ratio of the slope’, rather than the slope itself. Please correct the title of the legend bar to avoid any confusion.
- Line 212, it is intriguing to see that the onset of stratification is always later than the Chl-a peak event, even at the very beginning of the simulation in 1985 (Fig. 3). Conventionally, as already stated in the introduction, the onset of stratification is a prerequisite for the spring phytoplankton bloom (Line 49). Are there any observations in Lake Erken, that the current or previous spring phytoplankton blooms were already earlier than the onset of stratification since 1985? Are these species diatom, according to field data and model predictions? Overall, it is necessary to add a few more explanations here.
Citation: https://doi.org/10.5194/egusphere-2023-1679-RC2 - AC2: 'Reply on RC2', Jorrit Mesman, 08 Dec 2023
-
RC3: 'Comment on egusphere-2023-1679', Love Raman Vinna, 06 Nov 2023
Review of “Timing of spring events changes under modelled future climate scenarios in a mesotrophic lake. By Jorrit P. Mesman, Inmaculada C. Jiménez-Navarro, Ana I. Ayala, Javier Senent-Aparicio, Dennis Trolle, Don Pierson”
In this manuscript the authors use a one way coupled drainage area to lake model setup to investigate future climate impact on spring processes including ice-off, 50% cumulative spring discharge, spring phytoplankton bloom and stratification onset. The bold and novel model setup include stream flow, nutrients and temperature (SWAT+, LOADEST, air2water) coupled to lake physics and biogeochemistry (GOTM, WET). The important findings of the authors show how the occurrence of important spring processes are occurring earlier in a future warmer climate. The manuscript is in a good order but would benefit from extra clarity, sliming down and expansion as my points hereunder show.
1 This manuscript continue and analyze deeper the effect of climate from the work done in Jiménez-Navarro et al. (2023). The reader needs to clearly understand what is the difference between the two works, both in regard to which questions are being addressed here as well as be given all relevant information for spring processes. This point runs throughout the rest of this review.
2 The method description need to be expanded and put in line with Jiménez-Navarro et al. (2023). Among other things I cannot see how many parameters was used in air2stream, which is not a statistical model, it is a semi deterministic model (a hybrid process-based and data-driven model). Additionally, more detailed information regarding the GOTAM-WET model coupling is required. One of the things I miss is how transparency in the lake is modeled/treated. Do the biological model adjust lake transparency, and how do this affect spring bloom and stratification onset? And how do the coupled model preform at deeper depth? The reader can now only see what happens at 3 m depth.
3 the authors struggle with model correctness, needing to use a surface temperature threshold for ice-off despite having an ice module and need to explain discrepancies in stratification onset and chlorophyll spring peak. I ask myself how this can be and have some points here which might enlighten the manuscript. First do the grid resolution compared to measurement resolution affect the results? From Figure S5 timing of stratification it looks like the vertical lines denoting stratification onset do not match the data and should in fact be earlier for the measurements (red line crossing green threshold before timing of stratification). Is this due to a too short window for continuing stratification, is there an error in the script, or do the resolution play a role? As for data. The one way coupled catchment and lake model setup was calibrated from 2000 to 2015 for the lake part and from 2007 to 2015 for the river part. Is the difference in calibration period affecting the results? Looking at Figure 2 for Ice-off this looks to be the case. And how do you deal with the 2000 to 2006 period in regard to river input into the lake model? Building on this, can the less than ideal model correctness be explained by the location of measurements in and above the lake? Lake measurements come from a station at the deepest point in the lake ca 400 m from the eastern shore. This distance might be far enough away for near-shore processes to play a role, but are the location representative for the overall lake physics covering the central parts of the lake? Additionally but not required for this manuscript, it would have improved the results if the complete time frame of available data was considered for calibration, validation (if deemed necessary) could have been carried out in the start and not the end of available measurements see ex. Shen, H., Tolson, B. A. & Mai, J. Time to Update the Split‐Sample Approach in Hydrological Model Calibration. Water Resour. Res. 58, (2022). https://doi.org/10.1029/2021WR031523.
4 Lake processes are heavily dependent on local atmospheric conditions, so to for the drainage area processes. The authors used five GCM models which by their global nature are course resolved. The GCMs are bias correction toward local measurements in Jiménez-Navarro et al. (2023), but if I understand Supplementary table E 4th column (RMSE) this bias correction is almost nonexistent. Taking the difference between GCM INM-CM5-0 and measurements as an example, mean air temperature RMSE (Root Mean Square Error) drops from the unbiased comparison of 5.712 oK to 5.687 oK after bias correction and for Wind Speed from 4.283 to 2.588 m/s, and improvement with <1% and ~40 % respectively. The bias correction of precipitation, a key input to the drainage area model, looks to have failed. Now I might misunderstand how the Bias correction results are shown, but this illustrate my first point. Can we trust that the calibration is still valid using the climate models as input? Additionally, the reader needs to know why these climate models and scenarios were selected. I suspect because they cover the extreme ranges of for example temperature, precipitation, wind speed etc.. Furthermore since the setup is used for projecting climate effects, is the time frame (for drainage area and lake) long enough so that the models capture the climate trend (which is small compared to seasonal variations)? It would help the reader to see how the trends during the setup/calibration period are in the model compared to measurements.
5 Through the analysis of trends from the climate simulations, the authors treat the climate scenarios as constant change over time ex. Fig 3. This is not correct, in fact the gradient for each scenario change over time, especially for SSP 245. I suggest dividing the model output into 30 year chunks while conducting the analysis, or look at the amount of change from a reference to a far future period.
Citation: https://doi.org/10.5194/egusphere-2023-1679-RC3 - AC3: 'Reply on RC3', Jorrit Mesman, 08 Dec 2023
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 341 | 122 | 45 | 508 | 43 | 11 | 12 |
- HTML: 341
- PDF: 122
- XML: 45
- Total: 508
- Supplement: 43
- BibTeX: 11
- EndNote: 12
Viewed (geographical distribution)
| Country | # | Views | % |
|---|---|---|---|
| United States of America | 1 | 149 | 30 |
| Germany | 2 | 72 | 14 |
| New Zealand | 3 | 28 | 5 |
| Sweden | 4 | 28 | 5 |
| China | 5 | 24 | 4 |
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
- 149
Cited
1 citations as recorded by crossref.
Jorrit P. Mesman
Inmaculada C. Jiménez-Navarro
Ana I. Ayala
Javier Senent-Aparicio
Dennis Trolle
Don Pierson
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(819 KB) - Metadata XML
-
Supplement
(1646 KB) - BibTeX
- EndNote
- Final revised paper