the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 08 Nov 2023
Extrapolation is not enough: Impacts of extreme land-use change on wind profiles and wind energy according to regional climate models
Abstract. Humans change the climate in many ways. In addition to greenhouse gases, climate model inputs thus include a number of other forcings like land-use change. While studies typically investigate the joint effects of all forcings, we here isolate the impact of afforestation and deforestation on winds in the lowermost 350 m of the atmosphere to quantify the relevance of the lower boundary condition for large-scale wind energy assessments. We use vertically resolved sub-daily output from two regional climate models instead of extrapolating near-surface winds with simplified profiles. Comparing two extreme scenarios, we report that afforestation reduces wind speeds by more than 1 m/s in many locations across Europe even 300 m above ground and thus remains relevant at hub heights of current and future wind turbines. We show that standard extrapolation with modified parameters approximates long-term means well but fails to capture essential spatio-temporal details, such as changes in the daily cycle, and is thus insufficient to estimate wind energy potentials. Using adjacent climate model levels to account for spatio-temporal wind profile complexity, we report that wind energy capacity factors are strongly impacted by afforestation and deforestation: they differ by more than 0.1 in absolute terms and up to 50 % in relative terms. Our results confirm earlier studies that land use change impacts on wind energy can be severe and that they are generally misrepresented with common extrapolation techniques.
-
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
(6472 KB)
-
Supplement
(11498 KB)
-
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(6472 KB) - Metadata XML
-
Supplement
(11498 KB) - BibTeX
- EndNote
- Final revised paper
Journal article(s) based on this preprint
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-2533', Anonymous Referee #1, 07 Dec 2023
I have reviewed the paper titled"Extrapolation is not enough: Impacts of extreme land-use change on wind profiles and wind energy according to regional climate models”. I believe the paper needs major revision as more details and analysis are needed to justify what the authors mentioned. In addition, the language needs sharpen to facilitate better understanding for the reader.
- Lines 125-126: remove “uses modeling time steps of 90 seconds, while GERICS uses 240 seconds”. This is one of the many incidents where the sentence is awkward. The language needs more polish.
- More details are needed in Table 1(or describe in text) to show the readers the difference between the two models, such as the boundary layer scheme, surface layer scheme, land surface model and the boundary forcing.
- Figures 1 and 2: can you provide an explanation why the difference in wind speed reduces from lower to higher vertical level?
- Line 167-168“IDL features reductions of similar magnitude, even exceeding 1.5 m/s in large parts of Scandinavia and Iceland near the surface that slowly decay with height (Fig. 2l).” Do you mean Figure 2?
- Lins 178-180 “This uncertainty is a combination of at least two factors, namely (a) how the models (or modelers) translate the scenario into actual boundary conditions and (b) how the models respond to those boundary conditions.” Given the differences in model configuration and run time option, this statement is not accurate. I suggest remove this sentence.
- Section 3.2/Figure 4: why just 3 locations? The result is statistically insignificant. More points are needed for the analysis. For instance, the authors should randomly select at least 10 points per percentile range to get an estimate.
- Figure 5/ Lines 255-258: Again, more data points are need for the analysis.
- Lines 259 to 260: “The higher wind speeds in FOREST as compared to GRASS are a large-scale phenomenon and not an artifact of the selected case study location (see Fig. 6)” Why is it reasonable for the wind speed to be higher in FOREST than GRASS? What larger-scale phenomenon are you referring to?
- Line 284-285: “The higher IDL summer temperatures over FOREST could imply that the boundary layer is more stable than in GERICS, favouring the creation on nocturnal low-level jets” Why higher temperature favors stable boundary layer?
Citation: https://doi.org/10.5194/egusphere-2023-2533-RC1 - AC1: 'Reply on RC1', Jan Wohland, 27 Mar 2024
-
RC2: 'Comment on: Extrapolation is not enough: Impacts of extreme land-use change on wind profiles nad wind energy according to regional climate models', Anonymous Referee #2, 12 Jan 2024
I appreciate the chance to review Extrapolation is not enough: Impacts of extreme land-use change on wind profiles nad wind energy according to regional climate models. Furthermore, I appreciate the exploration of from such extreme model scenarios as ‘all grass’ or ‘all forest’ onshore and how it affects land-atmosphere interactions and vertical wind profiles. However, the method is not reproducible as written requiring me to look elsewhere for answers and then I found a major error, and I also have some reactions to how the reason for the study and the associated results are presented.
The major error relates to the use of the SWT120-3600 turbine specifications. According to the manufacturer’s website (https://en.wind-turbine-models.com/turbines/669-siemens-swt-3.6-120-offshore the hub-height of this wind turbine is either 88 or 90 meters, with a 90-meter hub-height also stated here (https://www.thewindpower.net/turbine_en_79_siemens_swt-3.6-120.php) and here (https://github.com/wind-python/windpowerlib/blob/dev/windpowerlib/oedb/turbine_data.csv). The manuscript states (line 160-1) “This turbine has a hub-height of 120 meters and was chosen following Wohland et al. (2021a) as it represents the median current wind turbine.” In the three web references in this paragraph, the ’120’ refers to the rotor diameter not the hub-height. It remains possible the author is using a difference reference for the hub-height, or it could simply be an honest mistake. Assuming it is in fact an error, figures (Fig. 7&8) and sections of the text related to the capacity factor are wrong and require complete revision.
More related to the presentation of the results, there are several parts I found confusing:
- Title: what do you suggest given that Extrapolation is not enough…?
- It might be assumed from the focus in the manuscript that a model is appropriate, but then, given that you are showing how the swapping of all grass to all forest influences the hub-height wind speeds, what should be expected from large-scale installations of wind turbines over grass or forest; this isn’t the exact point of your paper but it does undermine the applicability of these model simulations to your suggested application (as per the Abstract and Conclusion)
- Abstract (line 11-13): “Our results confirm earlier studies that land use change impacts on wind energy can be severe and that they are generally misrepresented with common extrapolation techniques.”
- As the concluding sentence of the Abstract, what new knowledge will be presented to the reader here?
- ? Are the models analyzed here better or different than the European observations (Vautard et al. 2010 in Nature Geoscience): “In addition, mesoscale model simulations suggest that an increase in surface roughness—the magnitude of which is estimated from increases in biomass and land-use change in Eurasia—could explain between 25 and 60% of the stilling. Moreover, regions of pronounced stilling generally coincided with regions where biomass has increased over the past 30 years, supporting the role of vegetation increases in wind slowdown.”
- Methods
- Are you sure the model heights you use throughout are midpoints of the vertical rather than the top of each model height (Fig. 4&5) At least for WRF, these heights could be post-processed differently from how the model considers altitude.
- Is only the lowest model level influenced by the change to all forest in both models? Presumably the trees could physically extend to more than ~30m tall and therefore into the 2nd lowest model level but the parameterizations and models prevent such complex dynamics?
- Interpolation from hub-height wind speeds from both models from the modeled wind speeds above and below hub-height
- Where is it described how the extrapolation approaches (Eq. 2, Eq. 3) are not enough according to your title?
- I expected to see comparisons of these two approaches to the interpolated model levels, but where are these comparisons?
- 2 includes the displacement height (d) to account for turbulence differences with the atmospheric surface layer but all that seems to be changing is the roughness length (line 146)
- Are you suggesting that displacement height doesn’t need to change in a fully forested onshore scenario poised to estimate wind speeds?
- Supplement includes extra figures but not a more detailed methodology – manuscript suggests the python code will be made available on github upon publication
- Results:
- Section 3.2 (Wind change profiles per season at individual locations)
- Here I was expecting to learn when the extrapolation estimates from either surface (10m) variables or model levels struggled, but it is a comparison between IDL and GERICS
- Section 3.4: Low level jet– agree that surface wind speeds are not appropriate for estimating the LLJ speed or simply the presence of an LLJ (saying nothing about an LLJ being defined by a ‘nose’ that requires slower wind speeds above it and below it; not possible from either wind speed extrapolation approach) – I don’t understand why a discussion about LLJ is in here as a text bridge to how the LLJ may empower or damage wind turbines isn’t made
- Capacity factor plots: Given the enormous research and economic interests in quantifying wake effects, maps of capacity factors (such as Fig. 7a,b,d,e) can easily be mis-interpreted that what is actually being shown is that turbine’s power curve being applied to the winds without including a between-turbine spacing consideration or wake effect consideration – realilzed capacity factors will be much lower even if the wind speeds in the model are perfect everywhere and all the time (and assuming Europe is all grass or all forest)
- Section 3.2 (Wind change profiles per season at individual locations)
- Where is it described how the extrapolation approaches (Eq. 2, Eq. 3) are not enough according to your title?
Citation: https://doi.org/10.5194/egusphere-2023-2533-RC2 - AC1: 'Reply on RC1', Jan Wohland, 27 Mar 2024
- Title: what do you suggest given that Extrapolation is not enough…?
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2023-2533', Anonymous Referee #1, 07 Dec 2023
I have reviewed the paper titled"Extrapolation is not enough: Impacts of extreme land-use change on wind profiles and wind energy according to regional climate models”. I believe the paper needs major revision as more details and analysis are needed to justify what the authors mentioned. In addition, the language needs sharpen to facilitate better understanding for the reader.
- Lines 125-126: remove “uses modeling time steps of 90 seconds, while GERICS uses 240 seconds”. This is one of the many incidents where the sentence is awkward. The language needs more polish.
- More details are needed in Table 1(or describe in text) to show the readers the difference between the two models, such as the boundary layer scheme, surface layer scheme, land surface model and the boundary forcing.
- Figures 1 and 2: can you provide an explanation why the difference in wind speed reduces from lower to higher vertical level?
- Line 167-168“IDL features reductions of similar magnitude, even exceeding 1.5 m/s in large parts of Scandinavia and Iceland near the surface that slowly decay with height (Fig. 2l).” Do you mean Figure 2?
- Lins 178-180 “This uncertainty is a combination of at least two factors, namely (a) how the models (or modelers) translate the scenario into actual boundary conditions and (b) how the models respond to those boundary conditions.” Given the differences in model configuration and run time option, this statement is not accurate. I suggest remove this sentence.
- Section 3.2/Figure 4: why just 3 locations? The result is statistically insignificant. More points are needed for the analysis. For instance, the authors should randomly select at least 10 points per percentile range to get an estimate.
- Figure 5/ Lines 255-258: Again, more data points are need for the analysis.
- Lines 259 to 260: “The higher wind speeds in FOREST as compared to GRASS are a large-scale phenomenon and not an artifact of the selected case study location (see Fig. 6)” Why is it reasonable for the wind speed to be higher in FOREST than GRASS? What larger-scale phenomenon are you referring to?
- Line 284-285: “The higher IDL summer temperatures over FOREST could imply that the boundary layer is more stable than in GERICS, favouring the creation on nocturnal low-level jets” Why higher temperature favors stable boundary layer?
Citation: https://doi.org/10.5194/egusphere-2023-2533-RC1 - AC1: 'Reply on RC1', Jan Wohland, 27 Mar 2024
-
RC2: 'Comment on: Extrapolation is not enough: Impacts of extreme land-use change on wind profiles nad wind energy according to regional climate models', Anonymous Referee #2, 12 Jan 2024
I appreciate the chance to review Extrapolation is not enough: Impacts of extreme land-use change on wind profiles nad wind energy according to regional climate models. Furthermore, I appreciate the exploration of from such extreme model scenarios as ‘all grass’ or ‘all forest’ onshore and how it affects land-atmosphere interactions and vertical wind profiles. However, the method is not reproducible as written requiring me to look elsewhere for answers and then I found a major error, and I also have some reactions to how the reason for the study and the associated results are presented.
The major error relates to the use of the SWT120-3600 turbine specifications. According to the manufacturer’s website (https://en.wind-turbine-models.com/turbines/669-siemens-swt-3.6-120-offshore the hub-height of this wind turbine is either 88 or 90 meters, with a 90-meter hub-height also stated here (https://www.thewindpower.net/turbine_en_79_siemens_swt-3.6-120.php) and here (https://github.com/wind-python/windpowerlib/blob/dev/windpowerlib/oedb/turbine_data.csv). The manuscript states (line 160-1) “This turbine has a hub-height of 120 meters and was chosen following Wohland et al. (2021a) as it represents the median current wind turbine.” In the three web references in this paragraph, the ’120’ refers to the rotor diameter not the hub-height. It remains possible the author is using a difference reference for the hub-height, or it could simply be an honest mistake. Assuming it is in fact an error, figures (Fig. 7&8) and sections of the text related to the capacity factor are wrong and require complete revision.
More related to the presentation of the results, there are several parts I found confusing:
- Title: what do you suggest given that Extrapolation is not enough…?
- It might be assumed from the focus in the manuscript that a model is appropriate, but then, given that you are showing how the swapping of all grass to all forest influences the hub-height wind speeds, what should be expected from large-scale installations of wind turbines over grass or forest; this isn’t the exact point of your paper but it does undermine the applicability of these model simulations to your suggested application (as per the Abstract and Conclusion)
- Abstract (line 11-13): “Our results confirm earlier studies that land use change impacts on wind energy can be severe and that they are generally misrepresented with common extrapolation techniques.”
- As the concluding sentence of the Abstract, what new knowledge will be presented to the reader here?
- ? Are the models analyzed here better or different than the European observations (Vautard et al. 2010 in Nature Geoscience): “In addition, mesoscale model simulations suggest that an increase in surface roughness—the magnitude of which is estimated from increases in biomass and land-use change in Eurasia—could explain between 25 and 60% of the stilling. Moreover, regions of pronounced stilling generally coincided with regions where biomass has increased over the past 30 years, supporting the role of vegetation increases in wind slowdown.”
- Methods
- Are you sure the model heights you use throughout are midpoints of the vertical rather than the top of each model height (Fig. 4&5) At least for WRF, these heights could be post-processed differently from how the model considers altitude.
- Is only the lowest model level influenced by the change to all forest in both models? Presumably the trees could physically extend to more than ~30m tall and therefore into the 2nd lowest model level but the parameterizations and models prevent such complex dynamics?
- Interpolation from hub-height wind speeds from both models from the modeled wind speeds above and below hub-height
- Where is it described how the extrapolation approaches (Eq. 2, Eq. 3) are not enough according to your title?
- I expected to see comparisons of these two approaches to the interpolated model levels, but where are these comparisons?
- 2 includes the displacement height (d) to account for turbulence differences with the atmospheric surface layer but all that seems to be changing is the roughness length (line 146)
- Are you suggesting that displacement height doesn’t need to change in a fully forested onshore scenario poised to estimate wind speeds?
- Supplement includes extra figures but not a more detailed methodology – manuscript suggests the python code will be made available on github upon publication
- Results:
- Section 3.2 (Wind change profiles per season at individual locations)
- Here I was expecting to learn when the extrapolation estimates from either surface (10m) variables or model levels struggled, but it is a comparison between IDL and GERICS
- Section 3.4: Low level jet– agree that surface wind speeds are not appropriate for estimating the LLJ speed or simply the presence of an LLJ (saying nothing about an LLJ being defined by a ‘nose’ that requires slower wind speeds above it and below it; not possible from either wind speed extrapolation approach) – I don’t understand why a discussion about LLJ is in here as a text bridge to how the LLJ may empower or damage wind turbines isn’t made
- Capacity factor plots: Given the enormous research and economic interests in quantifying wake effects, maps of capacity factors (such as Fig. 7a,b,d,e) can easily be mis-interpreted that what is actually being shown is that turbine’s power curve being applied to the winds without including a between-turbine spacing consideration or wake effect consideration – realilzed capacity factors will be much lower even if the wind speeds in the model are perfect everywhere and all the time (and assuming Europe is all grass or all forest)
- Section 3.2 (Wind change profiles per season at individual locations)
- Where is it described how the extrapolation approaches (Eq. 2, Eq. 3) are not enough according to your title?
Citation: https://doi.org/10.5194/egusphere-2023-2533-RC2 - AC1: 'Reply on RC1', Jan Wohland, 27 Mar 2024
- Title: what do you suggest given that Extrapolation is not enough…?
Peer review completion
Journal article(s) based on this preprint
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 476 | 184 | 42 | 702 | 52 | 25 | 28 |
- HTML: 476
- PDF: 184
- XML: 42
- Total: 702
- Supplement: 52
- BibTeX: 25
- EndNote: 28
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
Jan Wohland
Peter Hoffmann
Daniela C. A. Lima
Marcus Breil
Olivier Asselin
Diana Rechid
The requested preprint has a corresponding peer-reviewed final revised paper. You are encouraged to refer to the final revised version.
- Preprint
(6472 KB) - Metadata XML
-
Supplement
(11498 KB) - BibTeX
- EndNote
- Final revised paper