Heatwaves Significantly Slow the Vegetation Growth Rate on the Tibetan Plateau

Round 1

Reviewer 1 Report

This work assessed changes in heatwave frequency and duration over the Tibetan Plateau and their impacts on vegetation during the peak growth season (June-August) from 2000 to 2020. Results demonstrate that the increase in frequency and severity of heatwaves slows the growth of vegetation, mainly in August. Although linking anomalous thermal conditions to vegetation greening seems to be an interesting idea, I believe that the manuscript cannot be accepted for publication in its current form. The manuscript suffers from several flaws that should be addressed before making a decision regarding this submission. Overall, my comments are listed as:

• • The authors failed to highlight the main research gap and, accordingly, indicate the main innovative aspects of their research. The Introduction chapter should be revisited to include the main results of some global and regional investigations that assessed changes in extremely high temperatures and their links to vegetation greening.
  • In a cold region like the Tibetan Plateau, I think it is better to rely on absolute temperature thresholds (e.g. 35 or more) to define heatwaves. This is simply because I believe that employing percentile-based thresholds (i.e. 90th percentile) based on the climatology of the whole year may introduce some smoothed and less severe events. This percentile may be used if it is computed only for the JJA season and not the entire year. For example, in Figure 2, are you sure that the Y-axis refers to the frequency of heat waves? In some years, more than 40 heatwaves occurred. Assuming that the definition of a heatwave is based mainly on five days or more of anomalous temperature, this means that we may have at least 200 hot days per year.
  • Abstract: The first few lines of the abstract are wordy. The authors should make good use of the space to highlight their main research methods, findings, and possible implications of the results. Provide the full names of abbreviations like NDVI and EVI. Also, it is recommended to include some quantitative results in the abstract.
  • What is the rationale behind using two different vegetation indices (NDVI and EVI)? Also, how can the response of vegetation to heatwave stress be different between these two metrics?
  • The study area description lacks adequate references.
  • A map showing the distribution of the meteorological stations and topographical gradients is needed.
  • It is recommended to highlight how the response of vegetation growth to heatwaves can vary as a function of the vegetation type. This is important because the plateau has a variety of vegetation communities, as shown in Figure 1.
  • I suppose that the authors have considered a regional series for the whole study domain (Figure 2). I think this approach is problematic given the uneven distribution of stations, as well as the complex topography of the study area. The authors may use a regional series based on the Thiessen polygon method to account for these influences.
  • How did the authors handle the possible impact of snow, cloud cover, and missing values presented in the remote sensing data, which is likely to have an impact in this mountainous region.
  • Another concern is related to how anomalous thermal conditions can impact vegetation differently according to the growth phase. This issue needs further discussion in the text.
  • At which level was the significance of heatwave trends or the correlation between vegetation and heatwaves?
  • Prior to computing trends between vegetation and heatwaves, it is recommended to detrend the series to account for possible trends presented in the data.
  • With almost 25 km resolution, soil moisture data is coarse enough to show how climate changes affect plant growth.
  • Also, this study assumes that the links between air temperature and vegetation are linear, which is definitely not the case. Vegetation growth may be impacted by other variables like irrigation, soil type, slope, etc. This aspect should be discussed more thoroughly in the text.
  • The use of acronyms should be revised carefully at their first appearance.
  • Language is problematic sometimes, as careful editing is needed to fix some syntax problems and improve the style.

Author Response

Dear Reviewers/Editor,

We appreciate your comments on our manuscript titled “Heatwaves significantly slow the vegetation growth rate on the Tibetan Plateau” (ID: remotesensing-1696217). We have fully considered these comments during the revision and have substantially improved the manuscript. We summarize our responses point by point below in blue, and revised the manuscript with “Track Change” mode.

 

 

Response to Reviewer 1

Comments and Suggestions for Authors

Specific comments:

Point 1: The authors failed to highlight the main research gap and, accordingly, indicate the main innovative aspects of their research. The Introduction chapter should be revisited to include the main results of some global and regional investigations that assessed changes in extremely high temperatures and their links to vegetation greening.

Response 1: Thank you for this valuable comment. We have reorganized the introduction to make the research gap more clearly. See line 46 to 53:

As a specific type of extreme high-temperature event, the heatwave has been reported recently by literature on the Tibetan Plateau. However, very few studies have been conducted on the ecological effect of the heatwave on the Tibetan Plateau. We only found one site scale study that reported the heatwave can substantially increase alpine ecosystem respiration on the Tibetan Plateau. Therefore, the response of alpine vegetation on the Tibetan Plateau to heatwave is poorly understood. It is necessary to evaluate the heatwave effect on vegetation more widely and provide valuable information to address the climate change in this region.

 

Point 2: In a cold region like the Tibetan Plateau, I think it is better to rely on absolute temperature thresholds (e.g. 35 or more) to define heatwaves. This is simply because I believe that employing percentile-based thresholds (i.e. 90th percentile) based on the climatology of the whole year may introduce some smoothed and less severe events. This percentile may be used if it is computed only for the JJA season and not the entire year.

Response 2: Thanks for your comment. As the reviewer’s comments, we used percentile-based thresholds to identify heatwaves on the Tibetan Plateau only in the JJA, not for the whole year. We have revised the method section (see section 2.3).

There is no consensus on the definition of a Heatwave at present. The absolute temperature thresholds and percentile-based thresholds are two widely used methods to identify the heatwave. In this study, we choose percentile-based thresholds method because absolute temperature thresholds (e.g. 35 or more) are too high to define heatwaves on the Tibetan Plateau. The maximum daily temperature is around 30 ℃ in the warmest month in the high elevation stations. The China Meteorological Administration reported two heatwaves happened in Lhasa city; they were 30.4℃ and 30.8℃, which is record-breaking temperature in Lhasa station. Therefore, we chose the percentile-based thresholds method and only focused on June, July and August.

Reference:

You, Q.; Jiang, Z.; Kong, L.; Wu, Z.; Bao, Y.; Kang, S.; Pepin, N. A comparison of heat wave climatologies and trends in China based on multiple definitions. Clim. Dyn. 2016, 48, 3975-3989, doi:10.1007/s00382-016-3315-0.

Lei Yang; A.H.; Lu Tongsuo; Liu Lile; Fu Wenxue ;Wei Dong. Characteristics of Air Temperature Variation in Lhasa City over the Past 49 Years. Earth and Environment, 2021, 49, 492-503, doi:10.14050/j.cnki.1672-9250.2021.49.039.

Perkins-Kirkpatrick,S.E.; Lewis,S.C. Increasing trends in regional heatwaves. Nat Commun. 2020, 11, 3357, doi:10.1038/s41467-020-16970-7.

 

Point 3: For example, in Figure 2, are you sure that the Y-axis refers to the frequency of heat waves? In some years, more than 40 heatwaves occurred. Assuming that the definition of a heatwave is based mainly on five days or more of anomalous temperature, this means that we may have at least 200 hot days per year.

Response 3: Thank you for this comment. We did not clarify clearly and led to misunderstand the figure. In Figure 2, the frequency of heatwaves is the sum frequency of all the 64 stations on the Tibetan Plateau. For the entire Tibetan Plateau, only few widespread heatwaves happened in last two decades, as shown in Figure R1. We have revised the Methods and caption of Figure 2 as follows:

A period with at least five consecutive hot days (the maximum temperature is greater than the threshold) was identified as a heatwave event at a single station. The frequency, duration and intensity are the characteristic of the heatwave event. The frequency of heatwaves is the sum of the heatwaves at 64 sites on the Tibetan Plateau in a year. (see the method section)

Figure 2. The frequency of the heatwave at 64 stations in June, July and August on the Tibetan Plateau from 2000 to 2020.

 

Figure R1 the widespread regional heatwaves in Tibetan Plateau during 2000-2020. The x-axis denotes year, y-axis is the occurrence date of the heatwave, and shading represents the number of the sites with heatwave of the average of T_max the heatwave event.

Point 4: Abstract: The first few lines of the abstract are wordy. The authors should make good use of the space to highlight their main research methods, findings, and possible implications of the results. Provide the full names of abbreviations like NDVI and EVI. Also, it is recommended to include some quantitative results in the abstract.

Response 4: Thank you for this comment. We have revised in the abstract to make it more concise.

In recent years, the heatwaves have been reported frequently by literature and the media on the Tibetan Plateau. However, it is unclear how alpine vegetation responds to the heatwaves on the Tibetan Plateau. This study aimed to identify the heatwaves using long-term meteorological data and examine the impact of heatwaves on vegetation growth rate with remote sensing data. The results indicated that heatwaves frequently occur during the peak growing season on the Tibetan Plateau. The heatwaves have no significant trend from 2000 to 2020. The correlation between the intensity of the heatwaves and the vegetation growth rate anomalies is significantly negative on the Tibetan Plateau. Both the normalized difference vegetation index (NDVI) and enhanced vegetation index (EVI) consistently show that the heatwaves slow the vegetation growth rate. This study outlines the importance of heatwave to vegetation growth to enrich our understanding of alpine vegetation response to increasing extreme weather events under the background of climate change.

 

Point 5: What is the rationale behind using two different vegetation indices (NDVI and EVI)? Also, how can the response of vegetation to heatwave stress be different between these two metrics?

Response 5: Thank you for this comment. NDVI and EVI are two widely used indices as the proxy for vegetation status [28,30]. Compared with NDVI, the soil signal is corrected in EVI by introducing the blue band. Meanwhile, NDVI tends to saturate over dense vegetation. Here, we use these two indices to corroborate our results.

Reference:

28. Gao, X.; Huete, A.R.; Ni, W.; Miura, T. Optical–Biophysical Relationships of Vegetation Spectra without Background Contamination. Remote Sens Environ. 2000, 74, 609-620, doi:10.1016/s0034-4257(00)00150-4.
29. Huete, A.; Didan, K.; Miura, T.; Rodriguez, E.P.; Gao, X.; Ferreira, L.G. Overview of the radiometric and biophysical performance of the MODIS vegetation indices. Remote Sens Environ. 2002, 83, 195–213, doi:10.1016/s0034-4257(02)00096-2.

Point 6: The study area description lacks adequate references.

Response 6: Thank you for this comment. We have added the following references and revised the study area description accordingly. (See section 2.1).

 

Point 7: A map showing the distribution of the meteorological stations and topographical gradients is needed.

Response 7: Thank you for this comment. We have revised the Figure 1 and added meteorological stations.

 

Figure 1. The spatial distribution of the meteorological stations on the Tibetan Plateau.

 

Point 8: It is recommended to highlight how the response of vegetation growth to heatwaves can vary as a function of the vegetation type. This is important because the plateau has a variety of vegetation communities, as shown in Figure 1.

Response 8: Thank you for this comment. This is also suggested by the reviewer #2. We have compared the response of different vegetation to heatwave.

In this study, we only focused on the alpine grassland on the Tibetan Plateau, because the most meteorological stations are located in grassland. We have revised the section 2.1 to make this clear.

Meanwhile, we have compared the response difference between alpine meadow and alpine steppe according to the vegetation type map. We have revised the discussion. Different grasslands on the Tibetan Plateau exhibited different response patterns to climate changes [26]. Interestingly, the alpine meadow is more sensitive to heatwaves than the alpine steppe in June (Figure 11), but the opposite is true in July and August. This may be resulted from the different coverage between the two types. Moreover, the growth rate changing mechanism is complex; more factors should be considered [47-50]. Further research is needed to clarify the detailed mechanism of these changes.

 

Figure 11. △VGR in each month in June, July and August on the Tibetan Plateau from 2000 to 2020. The dark green column represent the △VGR of the alpine meadow; the light green column show the △VGR of the alpine steppe.

Reference:

26. Bhattarai, P.; Zheng, Z.; Bhatta, K.P.; Adhikari, Y.P.; Zhang, Y. Climate-Driven Plant Response and Resilience on the Tibetan Plateau in Space and Time: A Review. Plants 2021, 10, 480, doi:10.3390/plants10030480.
27. Cheng, G.; Wu, T. Responses of permafrost to climate change and their environmental significance, Qinghai-Tibet Plateau. J. Geophys. Res. 2007, 112, doi:10.1029/2006jf000631.
28. Yi, S.; Zhou, Z.; Ren, S.; Xu, M.; Qin, Y.; Chen, S.; Ye, B. Effects of permafrost degradation on alpine grassland in a semi-arid basin on the Qinghai–Tibetan Plateau. Environ. Res. Lett. 2011, 6, doi:10.1088/1748-9326/6/4/045403.
29. Piao, S.; Cui, M.; Chen, A.; Wang, X.; Ciais, P.; Liu, J.; Tang, Y. Altitude and temperature dependence of change in the spring vegetation green-up date from 1982 to 2006 in the Qinghai-Xizang Plateau. Agric For Meteorol 2011, 151, 1599-1608, doi:10.1016/j.agrformet.2011.06.016.
30. Ganjurjav, H.; Gornish, E.S.; Hu, G.; Schwartz, M.W.; Wan, Y.; Li, Y.; Gao, Q. Warming and precipitation addition interact to affect plant spring phenology in alpine meadows on the central Qinghai-Tibetan Plateau. Agric For Meteorol 2020, 287, doi:10.1016/j.agrformet.2020.107943.

 

Point 9: I suppose that the authors have considered a regional series for the whole study domain (Figure 2). I think this approach is problematic given the uneven distribution of stations, as well as the complex topography of the study area. The authors may use a regional series based on the Thiessen polygon method to account for these influences.

Response 9: Thank you for this comment. Yes, the distribution of meteorological stations in this study area is sparse and uneven on the Tibetan Plateau. We have three points to respond to the comment. First, in this study, we mainly focus on the vegetation response to the heatwave. The vegetation is mainly distributed on the east part of the Tibetan Plateau, where the meteorological stations are located. Second, if upscale the heatwave to the entire Tibetan Plateau, the reliability of the result is difficult to assess, especially in the west part of the Tibetan Plateau, where almost no meteorological station is located. Third, many previous studies analyze extreme climate events only using the meteorological stations on the Tibetan Plateau (He et al., 2015; You et al., 2016; Li et al., 2021; Cong et al., 2017).

References:

He, S.; Richards, K.; Zhao, Z. Climate extremes in the Kobresia meadow area of the Qinghai-Tibetan Plateau, 1961–2008. Environ. Earth Sci. 2015, 75, 60, doi:10.1007/s12665-015-4784-x.

You, Q.; Jiang, Z.; Kong, L.; Wu, Z.; Bao, Y.; Kang, S.; Pepin, N. A comparison of heat wave climatologies and trends in China based on multiple definitions. Clim. Dyn. 2016, 48, 3975-3989, doi:10.1007/s00382-016-3315-0.

Li, X.; Ren, G.; Wang, S.; You, Q.; Sun, Y.; Ma, Y.; Wang, D.; Zhang, W. Change in the heatwave statistical characteristics over China during the climate warming slowdown. Atmos Res. 2021, 247, 105152, doi:10.1016/j.atmosres.2020.105152.

Cong, N.; Shen, M.; Yang, W.; Yang, Z.; Zhang, G.; Piao, S. Varying responses of vegetation activity to climate changes on the Tibetan Plateau grassland. Int J Biometeorol. 2017, 61, 1433-1444, doi:10.1007/s00484-017-1321-5.

 

Point 10: How did the authors handle the possible impact of snow, cloud cover, and missing values presented in the remote sensing data, which is likely to have an impact in this mountainous region.

Response 10: Thank you for this comment. We have considered the snow and cloud contamination in the remote sensing data during the analysis. Firstly, we calculated the NDVI and EVI from MCD43A4 surface reflectance product. The MCD43A4 product has removed view angle effects and masked cloud and snow contamination in the surface reflectance (Lucht et al., 2010). Secondly, we applied the Savitzky–Golay filter, a time series reconstruction algorithm, to correct inferior values in long-term NDVI and EVI time series in this study (Ma et al., 2006)

References:

Lucht, W.; Lewis, P. Theoretical noise sensitivity of BRDF and albedo retrieval from the EOS-MODIS and MISR sensors with respect to angular sampling. Int J Remote Sens. 2010, 21, 81-98, doi:10.1080/014311600211000.

Ma, M.; Veroustraete, F. Reconstructing pathfinder AVHRR land NDVI time-series data for the Northwest of China. Adv. Space Res. 2006, 37, 835-840, doi:10.1016/j.asr.2005.08.037.

 

Point 11: Another concern is related to how anomalous thermal conditions can impact vegetation differently according to the growth phase. This issue needs further discussion in the text.

Response 11: Thank you for this valuable comment. We have revised the discussion accordingly as follow:

Moreover, the alpine vegetation responds differently to heatwaves in different phenology stages. The △VGR in July is more significant than that in July and August. It is indicated that the alpine vegetation is more sensitive to heatwaves in the early growing season than in the later growing season. Vegetation is fragile and sensitive to the environment in the early growing season [26,44,45], for example, spring phenology is more sensitive to environmental factors than autumn phenology [46]. Meanwhile, vegetation usually grows faster in the early growing season than in the later growing sea-son; therefore, the growth rate may be more sensitive to environmental factors in the early growing season [45].

Reference:

26. Bhattarai, P.; Zheng, Z.; Bhatta, K.P.; Adhikari, Y.P.; Zhang, Y. Climate-Driven Plant Response and Resilience on the Tibetan Plateau in Space and Time: A Review. Plants. 2021, 10, 480, doi:10.3390/plants10030480.
27. Jerry L. Hatfield, J.H.P. Temperature extremes: Effect on plant growth and development. Weather. Clim. Extremes. 2015, 10, 4-10, doi:10.1016/j.wace.2015.08.001.
28. Yu, H.; Xu, J.; Okuto, E.; Luedeling, E. Seasonal response of grasslands to climate change on the Tibetan Plateau. PLoS ONE. 2012, 7, e49230, doi:10.1371/journal.pone.0049230.
29. Yu, H.; Luedeling, E.; Xu, J. Winter and spring warming result in delayed spring phenology on the Tibetan Plateau. Proc Natl Acad Sci U.S.A. 2010, 107, 22151-22156, doi:10.1073/pnas.1012490107.

 

Point 12: At which level was the significance of heatwave trends or the correlation between vegetation and heatwaves?

Response 12: Thank you for the comment. We have revised the manuscript accordingly. The statistical significance of heatwave trends is shown in Figure 2 (P value). The statistical significance of the correlation between △VGR and heatwaves is shown in Figure 8 (P value). We do not directly correlate the NDVI/EVI with heatwaves, because the most of the detected heatwaves are very weak.

 

Point 13: Prior to computing trends between vegetation and heatwaves, it is recommended to detrend the series to account for possible trends presented in the data.

Response 13: Thank you for this comment. We agree with the reviewer’s comment. Trends can result in a false correlation between two variables. This study only calculated the correlation between heatwave intensity and △VGR. We did not directly correlate the NDVI/EVI with heatwaves along the time. So, we think the trends have little impact on our results.

 

Point 14: With almost 25 km resolution, soil moisture data is coarse enough to show how climate changes affect plant growth.

Response 14: Thank you for this comment. We used precipitation and soil moisture to explore how water availability affects alpine grass during the heatwave. The microwave remote sensing-based soil moisture product is widely used in many studies. Although the spatial resolution is coarse, the temporal resolution (daily) is high enough for the analysis in a short period. The precipitation and soil moisture changed consistently when the heatwave happened.

 

Point 15: Also, this study assumes that the links between air temperature and vegetation are linear, which is definitely not the case. This aspect should be discussed more thoroughly in the text.

Response 15: Thanks for the comment. We fully agree with your viewpoint. We have revised the Discussion. The vegetation growth rate anomaly decreased linearly with the heatwave intensity. We think this is partly because most heatwaves are weak on the Tibetan Plateau, and alpine vegetation can recover from these disturbs. From our results, we found that most of the heatwave happened only in a very small part of the Tibetan Plateau. No heatwave happened for all the sites simultaneously on the Tibetan Plateau.

 

Point 16: The use of acronyms should be revised carefully at their first appearance.

Response 16: Thank you for this comment. We have revised this problem throughout the manuscript.

 

Point 17: Language is problematic sometimes, as careful editing is needed to fix some syntax problems and improve the style.

Response 17: Thank you for this comment. We have carefully checked and revised the language of the manuscript.

Author Response File: Author Response.docx

Reviewer 2 Report

Dear authors,

The article is well structured and highlights the results of the study. The Materials and Methods chapter is clearly described, including information regarding meteorological and remote sensing datasets and methods used for: the analysis of the impact of heatwaves on vegetation; for VIs and vegetation growth estimation using remote sensing data.

The Results chapter highlights clearly the trends of heatwaves frequency and their effects on vegetation.

Discussions: the authors explain the results of the study, including the inconsistency in trends between heatwave and extremely high temperatures for the Tibetan Plateau.  The study demonstrates the impact of heatwaves over alpine vegetation by combining stress and water limitation. Are explained clearly the uncertainties involved in the study also: (i) the sparse distribution of the meteorological observations affects the identification of heatwaves; (ii) the missing/gap-filled vegetation index value due to the low quality of remote sensing data (because of the cloud and snow contamination, and terrain complexity); (ii) the definition of a heatwave which is influenced by the unique natural conditions from the Tibetan Plateau.

Conclusions are clear and to the point, in accordance with the results of the study.

In conclusion, the article can be published with major corrections. The suggestions are listed below:

When you talk about vegetation: Is it not clear if the vegetation was analyzed in general or by vegetation types? It is a map with the types of vegetation but I did not understand if the analysis took into account that map or an average of the values of the indices was made on the whole Tibetan Plateau. If it is average, then it was considered that all types of vegetation have the same response to heat waves? It would have been good if the analysis had been done differently.

Please check the article and use the same abbreviation. VIs and Vis are used in the manuscript. Please use only one form.

At first use the abbreviations are detailed. Line 81: Please write in long MODIS at the first time mentioned. Use just the abbreviation in line 95.

Line 45: tem-pe-ra-tu-re. It is not correctly divided.

Line 49: Tibetan Plateau [19..] blank space before the [.

Line 58-50: According to the definition, a heatwave is consecutive days with extremely high temperatures, usually accompanied by a lack of precipitation. 

Is it another definition than the first one in the Introduction?

I suggest moving this sentence next to the first one or to remove it.

Line 61: I suggest using another adjective instead of "discrete".

Figure 1, line 93: The 64 weather stations are not represented in figure 1 as mentioned in this line.

Lines 93-94: The phrase Table A1…. Should be reformulated as follow: Table A1 shows information about each meteorological station (WMO code, name, latitude, longitude and elevation).

Lines 133-134: Has the Savitzky – Golay filter algorithm been applied to identify the vegetation season? It is used to reduce daily fluctuations and extract the beginning, peak, and end of the growing season. It is not clear in this paper what the purpose of this algorithm was because it does not further detail the results obtained with this algorithm. However, it is necessary to better detail what are the stages of the growing season and when they are registered for each type of vegetation.

Line 135: The citations are not mentioned correctly and they are not included in the References.

Line 165: What does the peak growing season mean? How was it identified?

Line 167: Please leave a blank space between Figure and 3.

Line 168: Does the peak of the growing season last 3 months in the study area? Or it is about the vegetation season?

Line 175: 0.64 to 3.24 C/d... I can not see the 3.24 C in Figure 3.

Figure 3: Last 2 numbers have 3 decimals instead of 2, in the month of August. Maybe to add the units at least in the caption (days in fig 3a and degree in fig 3b)

Lines 186-188: At these lines, you mentioned the years when more than half of the total stations on the Tibetan Plateau (64 meteorological stations considered for the present analysis), according to Figure 4. Observing Figure 4, in 2009 and 2011, less than half of the total meteorological stations on the Tibetan Plateau recorded heatwaves. Please correct this.

Lines 244-245: The heatwave intensity is calculated as temperature anomaly during heatwaves multiplied by heatwaves duration. Thus, “°C.d” should be better replaced with “°C*d” or “°C·d”. Should be replaced in Figures 7a and 7c also.

Lines 261-262: In this paragraph, you are explaining the △VGR response to heatwaves in June 2013 and August 2016. So you are talking about Figure 8, not 7. In this regard, Figure 7(a) should be replaced by Figure 8(a) and Figure 7(c) with Figure 8(c), respectively.

Row 291: To be consistent in writing modify “WSDI (Warm spell duration index) with “WSDI (Warm Spell Duration Index)”.

Author Response

Dear Reviewers/Editor,

We appreciate your comments on our manuscript titled “Heatwaves significantly slow the vegetation growth rate on the Tibetan Plateau” (ID: remotesensing-1696217). We have fully considered these comments during the revision and have substantially improved the manuscript. We summarize our responses point by point below in blue, and revised the manuscript with “Track Change” mode.

 

 

Response to Reviewer 2

Comments and Suggestions for Authors

Specific comments:

Point 1: When you talk about vegetation: Is it not clear if the vegetation was analyzed in general or by vegetation types? It is a map with the types of vegetation but I did not understand if the analysis took into account that map or an average of the values of the indices was made on the whole Tibetan Plateau. If it is average, then it was considered that all types of vegetation have the same response to heat waves? It would have been good if the analysis had been done differently.

Response 1: Thank you for this comment. In this study, we only focused on the alpine grassland on the Tibetan Plateau, because the most meteorological stations are located in grassland. We have revised the section 2.1 to make this clear.

Meanwhile, we have compared the response difference between alpine meadow and alpine steppe according to the vegetation type map. We have revised the discussion. Different grasslands on the Tibetan Plateau exhibited different response patterns to climate changes [26]. Interestingly, the alpine meadow is more sensitive to heatwaves than the alpine steppe in June (Figure 11), but the opposite is true in July and August. This may be resulted from the different coverage between the two types. Moreover, the growth rate changing mechanism is complex; more factors should be considered [47-50]. Further research is needed to clarify the detailed mechanism of these changes.

 

Figure 11. △VGR in each month in June, July and August on the Tibetan Plateau from 2000 to 2020. The dark green column represent the △VGR of  the alpine meadow; the light green column show the △VGR of the alpine steppe.

Reference:

26. Bhattarai, P.; Zheng, Z.; Bhatta, K.P.; Adhikari, Y.P.; Zhang, Y. Climate-Driven Plant Response and Resilience on the Tibetan Plateau in Space and Time: A Review. Plants 2021, 10, 480, doi:10.3390/plants10030480.
27. Cheng, G.; Wu, T. Responses of permafrost to climate change and their environmental significance, Qinghai-Tibet Plateau. J. Geophys. Res. 2007, 112, doi:10.1029/2006jf000631.
28. Yi, S.; Zhou, Z.; Ren, S.; Xu, M.; Qin, Y.; Chen, S.; Ye, B. Effects of permafrost degradation on alpine grassland in a semi-arid basin on the Qinghai–Tibetan Plateau. Environ. Res. Lett. 2011, 6, doi:10.1088/1748-9326/6/4/045403.
29. Piao, S.; Cui, M.; Chen, A.; Wang, X.; Ciais, P.; Liu, J.; Tang, Y. Altitude and temperature dependence of change in the spring vegetation green-up date from 1982 to 2006 in the Qinghai-Xizang Plateau. Agric For Meteorol 2011, 151, 1599-1608, doi:10.1016/j.agrformet.2011.06.016.
30. Ganjurjav, H.; Gornish, E.S.; Hu, G.; Schwartz, M.W.; Wan, Y.; Li, Y.; Gao, Q. Warming and precipitation addition interact to affect plant spring phenology in alpine meadows on the central Qinghai-Tibetan Plateau. Agric For Meteorol 2020, 287, doi:10.1016/j.agrformet.2020.107943.

 

Point 2: Please check the article and use the same abbreviation. VIs and Vis are used in the manuscript. Please use only one form.

Response 2: Thank you for this comment. We have unified the abbreviation (only using “VIs”) and revised it accordingly throughout the manuscript.

 

Point 3: At first use the abbreviations are detailed. Line 81: Please write in long MODIS at the first time mentioned.

Response 3: Thank you for this comment. We have revised this problem in the manuscript.

 

Point 4: Line 45: tem-pe-ra-tu-re. It is not correctly divided.

Response 4: Thank you for this comment. We have corrected the “tem-pe-ra-tu-re” as “temperature”.

 

Point 5:  Line 49: Tibetan Plateau [19..] blank space before the [.

Response 5: Thank you for this comment. We have revised it.

 

Point 6: According to the definition, a heatwave is consecutive days with extremely high temperatures, usually accompanied by a lack of precipitation. Is it another definition than the first one in the Introduction? I suggest moving this sentence next to the first one or to remove it.

Response 6: Thank you for recommending We have removed this sentence.

 

Point 7: Line 61: I suggest using another adjective instead of "discrete".

Response 7: Thank you for this comment. We have revised this sentence as follows:

On the Tibetan Plateau, heatwaves are usually of short duration and take place at a small part of the plateau.

 

Point 8: Figure 1, line 93: The 64 weather stations are not represented in figure 1 as mentioned in this line.

Response 8: Thank you for this valuable comment. We have added weather stations in the Figure 1.

 

Figure 1. The spatial distribution of the meteorological stations on the Tibetan Plateau.

 

Point 9: Lines 93-94: The phrase Table A1…. Should be reformulated as follow: Table A1 shows information about each meteorological station (WMO code, name, latitude, longitude and elevation).

Response 9: Thank you for this statement. We have revised in the revised manuscript.

Point 10: Lines 133-134: Has the Savitzky – Golay filter algorithm been applied to identify the vegetation season? It is used to reduce daily fluctuations and extract the beginning, peak, and end of the growing season. It is not clear in this paper what the purpose of this algorithm was because it does not further detail the results obtained with this algorithm. However, it is necessary to better detail what are the stages of the growing season and when they are registered for each type of vegetation.

 

Response 10: Thank you for this comment. We use the Savitzky – Golay filter to correct the inferior values in time series VIs, not to identify the peak growing season. This study only focused on the summer (June, July and August) to examine the heatwave effect on alpine grassland. To make it clear, we have replaced “the peak growing season” with “June, July and August” in the manuscript.

In addition, to correct inferior values in VIs, a time series reconstruction algorithm, the Savitzky – Golay filter (Equation 3), is applied to long-term daily VIs in this study [36] .

References:

36. Ma, M.; Veroustraete, F. Reconstructing pathfinder AVHRR land NDVI time-series data for the Northwest of China. Adv. Space Res. 2006, 37, 835-840, doi:10.1016/j.asr.2005.08.037.

 

Point 11: Line 135: The citations are not mentioned correctly and they are not included in the References.

Response 11: Thank you for this comment. The citations have been corrected.

 

Point 12: Line 165: What does the peak growing season mean? How was it identified?

Response 12: Thank you for this comment. The peak growing season means June, July and August in this work. We only focused on the summer (June, July and August) to examine the heatwave effect on alpine grassland. To make it clear, we have replaced “the peak growing season” with “June, July and August” in the manuscript.

 

Point 13: Line 167: Please leave a blank space between Figure and 3.

Response 13: Thank you for this comment. We have revised it in the revised manuscript.

 

Point 14: Line 168: Does the peak of the growing season last 3 months in the study area? Or it is about the vegetation season?

Response 14: Thank you for this comment. We only focused on the summer (June, July and August) to examine the heatwave effect on alpine grassland. To make it clear, we have replaced “the peak growing season” with “June, July and August” in the manuscript.

 

Point 15: Line 175: 0.64 to 3.24 C/d... I can not see the 3.24 C in Figure 3.

Response 15: Thank you for this comment. We have corrected this problem in the revised manuscript.

 

Point 16: Figure 3: Last 2 numbers have 3 decimals instead of 2, in the month of August. Maybe to add the units at least in the caption (days in fig 3a and degree in fig 3b).

Response 16: Thank you for this comment. We have modified the table and keep two decimal places. See Figure 3 in the revised version.

 

Figure 3. Matrix heatmap for heatwave duration and intensity in June, July and August on the Tibetan Plateau from 2000 to 2020. The matrix heatmap (a) refers to the temporal change of the heatwave duration; the matrix heatmap (b) refers to the temporal change of the heatwave intensity. Each grid cell represents the average duration and intensity of a heatwave at all sites in a given month of a year. The blank grid cell represents no heatwave, and the value is NaN. The color of the matrix heatmap represents the size of the value.

 

Point 17: Lines 186-188: At these lines, you mentioned the years when more than half of the total stations on the Tibetan Plateau (64 meteorological stations considered for the present analysis), according to Figure 4. Observing Figure 4, in 2009 and 2011, less than half of the total meteorological stations on the Tibetan Plateau recorded heatwaves. Please correct this.

Response 17: Thank you for this comment. We have corrected this in the revised manuscript.

Point 18: Lines 244-245: The heatwave intensity is calculated as temperature anomaly during heatwaves multiplied by heatwaves duration. Thus, “°C.d” should be better replaced with “°C*d” or “°C·d”. Should be replaced in Figures 7a and 7c also.

Response 18: Thank you for this comment. We have revised unit in the Figure 7a and 7c.

 

Point 19: Lines 261-262: In this paragraph, you are explaining the △VGR response to heatwaves in June 2013 and August 2016. So you are talking about Figure 8, not 7. In this regard, Figure 7(a) should be replaced by Figure 8(a) and Figure 7(c) with Figure 8(c), respectively.

Response 19: Thank you for this comment. You are right, it should be Figure 7. We have corrected this in the revised manuscript.

Point 20: Row 291: To be consistent in writing modify “WSDI (Warm spell duration index) with “WSDI (Warm Spell Duration Index)”.

Response 20: Thank you for this comment. We have revised “Warm spell duration index” as “Warm Spell Duration Index”.

 

 

 

 

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

The authors have addressed most of the comments that were raised in the previous revision. Before making a final decision on this submission, I have a few concerns to address.

• In the abstract, please include the significance level at which both trends and correlations were evaluated.
• The authors relied on the MODIS BRDF's daily data. It's common for this product's daily version to contain significant numbers of missing values. What percentage of your dataset does it contain? What steps did you take to resolve this situation? The methodology section should address all of these issues in greater detail.
• A section addressing the study's major limitations should be included. This may include the uneven distribution of the meteorological station, the coarse spatial resolution of soil moisture data, etc.
• According to Figure 2, summing the frequency of heatwaves across the entire domain does not make sense, as it's possible that the frequency of heatwaves at specific locations on the plateau has a significant impact on this total. As a result, presenting the average frequency of heatwaves for all available stations is highly recommended.
• I would recommend tracking the onset (1^st occurrence) and end (last occurrence) of heatwaves over time to see if there are any changes in their temporal variability. The ecological response (e.g. resilience, resistance, sensitivity) to such anomalous events is determined by changes in these patterns.
• I am wondering why the scatter points were not presented at the station level (i.e. N=64) in Figure 7.
• Precipitation and soil moisture were found to be inconsistent in Figure 10. Soil moisture increased during the heatwave in August 2016, while precipitation increased (relative to the pre-conditions). It is imperative that this finding and its possible explanations be examined.
• Some minor linguistic errors should be fixed across the whole document. One example is found in L49, "reported" <> "reported that".

Author Response

Response to Reviewer 1

The authors have addressed most of the comments that were raised in the previous revision. Before making a final decision on this submission, I have a few concerns to address.

Many thanks for your approval of our first-round revision. We have carefully revised the manuscript following your comments. We hope that the revised version would be eligible for publication. We summarize our responses point by point below in blue and revised the manuscript with “Track Change” mode on the first-round revision accepted version.

Comments and Suggestions for Authors

Point 1: In the abstract, please include the significance level at which both trends and correlations were evaluated.

Response 1: Thank you for this valuable comment. We have added significance levels for trends and correlations in the Abstract.

Abstract: In recent years, the heatwaves have been reported frequently by literature and the media on the Tibetan Plateau. However, it is unclear how alpine vegetation responds to the heatwaves on the Tibetan Plateau. This study aimed to identify the heatwaves using long-term meteorological data and examine the impact of heatwaves on vegetation growth rate with remote sensing data. The results indicated that heatwaves frequently occur in June, July, and August on the Tibetan Plateau. The average frequency of heatwaves had no statistically significant increasing trends from 2000 to 2020 for the entire Tibetan Plateau. On a monthly scale, the average frequency of heatwaves increased significantly (P < 0.1) in August, while no significant trends were in June and July. The intensity of heatwaves showed a negative correlation with the vegetation growth rate anomaly (△VGR) calculated from the normalized difference vegetation index (NDVI) (r = -0.74, P<0.05) and the enhanced vegetation index (EVI) (r = -0.61, P<0.1) on the Tibetan Plateau, respectively. Both NDVI and EVI consistently showed that the heatwaves slow the vegetation growth rate. This study outlines the importance of heatwave to vegetation growth to enrich our understanding of alpine vegetation response to increasing extreme weather events under the background of climate change.

 

Point 2: The authors relied on the MODIS BRDF's daily data. It's common for this product's daily version to contain significant numbers of missing values. What percentage of your dataset does it contain? What steps did you take to resolve this situation? The methodology section should address all of these issues in greater detail.

Response 2: Thanks for your comments. Yes, the MCD43A4 product did contain a certain number of missing values. The missing values in MCD43A4 data mostly appear in winter due to snow contamination on the Tibetan plateau. Moreover, the cloud fraction is also low when heatwaves happen in summer. Therefore, the proportion of the missing values is very low in June, July and August compared to other periods. In this study, a total of 537 heatwaves were identified at all stations on the Tibetan Plateau in June, July and August from 2000 to 2020. Only 61 of them (about 11.36%) coincided with the MCD43A4 missing and failed to calculate the vegetation growth rate. We excluded these heatwaves in the analysis when MCD43A4 data was missing to make our result reliable. To make this clear in the manuscript, we have added the details in the methodology section.

(a)

(b)

Figure R1. Daily NDVI (a) and EVI (b) calculated from MCD43A4 data at station 52707 from 2000 to 2020.

 

Point 3: A section addressing the study's major limitations should be included. This may include the uneven distribution of the meteorological station, the coarse spatial resolution of soil moisture data, etc.

Response 3: Thank you for this comment. Actually, we have discussed the limitations and uncertainties of this work in the last paragraph of the discussion. We have strengthened the discussion according to your suggestion.

There are some uncertainties involved in this study. Firstly, meteorological observations have a relatively sparse and uneven distribution, resulting in the low representativeness of the identification of heatwaves [41]. Secondly, the quality of the remote sensing dataset is low on the Tibetan Plateau due to the contamination of snow, cloud and complex terrain. Thirdly, the reconstructed vegetation index value can result in uncertainties in the analysis. The different spatial representativeness among station data, MODIS data and coarse resolution soil moisture data can also lead to uncertainties in the study. Fourthly, the definition of a heatwave is uncertain due to the unique natural conditions on the Tibetan Plateau. In some heatwave definitions in the tropic or temperate region, a fixed high-temperature threshold is usually used by combining the temperature 90% percentile. In this work, only the 90% percentile was used and may result in uncertainties when comparing heatwaves on the Tibetan Plateau with another region.

Reference:

41. You, Q.; Fraedrich, K.; Ren, G.; Pepin, N.; Kang, S. Variability of temperature in the Tibetan Plateau based on homogenized surface stations and reanalysis data. Int J Climatol. 2013, 33, 1337-1347, doi:10.1002/joc.3512.

 

Point 4: According to Figure 2, summing the frequency of heatwaves across the entire domain does not make sense, as it's possible that the frequency of heatwaves at specific locations on the plateau has a significant impact on this total. As a result, presenting the average frequency of heatwaves for all available stations is highly recommended.

Response 4: Thank you for this comment. We have revised Figure 2 to show the average frequency of heatwaves. The frequency of heatwaves is the average for all the 64 stations in June, July and August on the Tibetan Plateau, as shown in Figure 2. To make this clearer, we calculated the heatwave frequency for each site (Figure R2). Figure R2 presents the matrix heatmap for the average frequency of heatwaves for each station in June, July and August on the Tibetan Plateau from 2000 to 2020. We have revised the Methods and caption of Figure 2 in the paper as follows:

The average frequency of heatwaves is the average at 64 sites on the Tibetan Plateau in a year.

Figure 2. The average frequency of heatwaves at 64 stations in June, July and August on the Tibetan Plateau from 2000 to 2020.

 

Figure R2. Matrix heatmap for the average frequency of heatwaves for each site in June, July and August on the Tibetan Plateau from 2000 to 2020. Each grid cell denotes the average frequency of heatwaves for each site in a given month from 2000 to 2020. The blank grid cell value is NaN. The color of the matrix heatmap represents the size of the value.

 

Point 5: I would recommend tracking the onset (1st occurrence) and end (last occurrence) of heatwaves over time to see if there are any changes in their temporal variability. The ecological response (e.g. resilience, resistance, sensitivity) to such anomalous events is determined by changes in these patterns.

Response 5: Thank you for this comment. We have extracted the onset and end of heatwaves over time. However, the temporal variability showed diverse trends among these stations. Figure R3 shows the onset and end of heatwaves from 2000 to 2002 at four stations (52602, 52707, 55680, and 56543) on the Tibetan Plateau. The first onset of the heatwaves is advanced, and the last occurrence of heatwaves is extended at station 52707. However, we can not see this phenomenon in the other three stations. Generally, we think this comment is an excellent new idea. The onset/end of heatwave may be advanced/delayed with global warming. The ecological response (e.g., resilience, resistance, sensitivity) to this change is a valuable scientific question to study further in our future works.

Figure R3 The occurrence date of the heatwaves for each site on the Tibetan Plateau during 2000-2020. The x-axis denotes the year, the y-axis is the occurrence date of the heatwave.

 

Point 6: I am wondering why the scatter points were not presented at the station level (i.e. N=64) in Figure 7.

Response 6: Thank you for this comment. Figure 7 shows the correlation between△VGR and heatwave intensity. In the identified heatwaves, most of them are weak heatwaves. To eliminate the random noises, we grouped these heatwaves along with the heatwave intensity. Then we calculated the correlation between heatwave intensity and △VGR.

 

Point 7: Precipitation and soil moisture were found to be inconsistent in Figure 10. Soil moisture increased during the heatwave in August 2016, while precipitation increased (relative to the pre-conditions). It is imperative that this finding and its possible explanations be examined.

Response 7: Thank you for this comment. We have also noticed this interesting phenomenon, but we did not conduct further analysis based on two reasons. First, the soil moisture data quality is unknown on the Tibetan Plateau due to no widespread validation. So, we cannot confirm whether this is a natural phenomenon or a phenomenon caused by data noise. No in-situ soil moisture observation can be used to check this guess. Second, the meteorological station precipitation data is reliable compared with the grid soil moisture dataset. Here, we use these two types of data to examine the role of water availability in vegetation growth rate during the heatwave.

Reference：

Xing, Z.; Fan, L.; Zhao, L.; De Lannoy, G.; Frappart, F.; Peng, J.; Li, X.; Zeng, J.; Al-Yaari, A.; Yang, K. A first assessment of satellite and reanalysis estimates of surface and root-zone soil moisture over the permafrost region of Qinghai-Tibet Plateau. Remote Sens Environ. 2021, 265, doi:10.1016/j.rse.2021.112666.

 

Point 8: Some minor linguistic errors should be fixed across the whole document. One example is found in L49, "reported" <> "reported that".

Response 8: Thank you for this comment. We have revised this problem and rechecked the language throughout the manuscript.

Author Response File: Author Response.docx

Reviewer 2 Report

Accept in present form.

Author Response

Remote Sensing                                                5/12/2022

Dear Reviewer,

 

We appreciate very much for your positive comments on our revised manuscript. According to the constructive comments from Reviewer#1, we have made a thorough revision on all sections of the manuscript.

 

Yours sincerely,

Caixia Dong