the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 04 Aug 2025
Extreme droughts' impact on Scots pine net primary productivity in the European temperate zone in the period 2002–2023
Abstract. Severe drought episodes significantly influence the productivity of trees in many parts of Europe. This paper shows in detail the influence of the most severe drought events on the productivity of the most important forest-forming tree species in European temperate zone – Scots pine (Pinus sylvestris L.). We identified four months with most severe drought conditions that occurred in Poland in 2002–2023: July 2006, April 2009, August 2015 and June 2019. To quantify trees’ net primary productivity (NPP) we used Moderate Resolution Imaging Spectroradiometer (MODIS) NPP, which was further correlated with temperature, precipitation, evapotranspiration and climatic water balance. The identified summer droughts had considerable effect on pine forest productivity: August 2015 had the lowest NPP of all Augusts in the study period (0.033 kgC∙m-2∙month-1), similarly July 2006 (0.055 kgC∙m-2∙month-1) and June 2019 (0.096 kgC∙m-2∙month-1). Relationship between drought severity and pine’s productivity depends on the time during the year, when the drought occurs. Summer droughts, with their peak in June, July or August, resulted in significantly decreased productivity of trees, while spring droughts, tend to have an initial positive impact on pine’s condition. For summer droughts cases, weather conditions influence the decreased productivity of pine forest for a long time, e.g. a prolonged negative relationship between NPP and temperature for drought cases in June 2019 and July 2006. Such long response of spectral indicator’s value of pine is not clearly visible for droughts occurring either on the beginning (April 2009) or second half (August 2015) of the growing season.
- Preprint
(1749 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 09 Nov 2025)
-
RC1: 'Comment on egusphere-2025-2770', Anonymous Referee #1, 05 Sep 2025
reply
-
AC1: 'Reply on RC1', Oliwier Zając, 11 Sep 2025
reply
- The article “Extreme droughts’ impact on Scots pine net primary productivity in the European temperate zone in the period 2002-2023” by Kulesza et al. investigated the impact of droughts (based on monthly climatic water balance anomalies) on the NPP of Scot pine forests in Poland, between 2002-2023.
- The authors identified particularly 4 summer months 2006, 2009, 2015 and 2019 where forest NPP (detected from MODIS NPP) was significantly affected.
We thank the reviewer for the useful comments, which have significantly contributed to improving our manuscript. Below are our thoughts and considerations for the amendments.
The 2009 case was not a summer case, but a spring case (April). Therefore, in Fig. 8 there was no consistent response across all months, which seems to be an understandable result of the analyses.
- The title mentions „European temperate zone “but the study is focused exclusively on forests in Poland. While the forests studied here are indeed characteristically temperate but the spatial coverage of the study and the number of species (only 1) does not capture the European temperate zone forests. The title is thus misleading.
We appreciate your comments regarding the title. At this point, we would like to stick to our proposal, as our study area and selected species (mentioned in the title) are representative not only on a regional scale but also for the entire country – Poland, one of the largest countries in Central Europe. Furthermore, Scots pine in Poland accounts for 58.9% of the total forest area, being the dominant species in the country (lines 127-128). Our statistical database (20,618 pixels at 500 m spatial resolution) is also crucial in terms of representativeness with respect to the European temperate zone.
- The objectives of the study are not novel. What do we want to learn from finding a drought that has the most negative impact, Is not clear. Also the results cannot be generalized even for this single species, because the response of Scot pines depend on the climatic and soil conditions they are adapted to. The study has not included the complete distribution of Scot pines across Europe and the diversity of local conditions in which they grow.
We appreciate your comments regarding the aims and scope of our study. We would like to address the following comments:
We know that drought and its impact on vegetation have been analyzed before. However, our study adds value by using a long time series of MODIS satellite data for this study area on a national scale. We will provide changes (both at the beginning and in the summary of the paper) to highlight which periods of extreme drought had the greatest potential effect on Scots pine productivity and why this is crucial for the study.
We are aware that the response of Scots pine depends on more than just climatic factors. We agree that the response of Scots pine varies spatially and depends on site conditions. It should be noted, that we wanted to examine productivity variability under changing meteorological conditions. We know that habitat conditions and soil conditions are relatively stable over time, but meteorological factors are more susceptible to change, which may lead to a decrease or increase in the productivity values of the pine.
Our goal was not to fully generalize the response of Scots pine to drought across all site conditions, but to present an approach that allows us to identify the most severe drought episodes and their impact on vegetation dynamics. Therefore, the results should be considered as a case study, not a complete representation of the species' response across its entire range. We therefore emphasize that this study does not claim to universalize the results for the entire species, but rather demonstrates the potential of remote sensing tools combined with the
ERA5-Land reanalysis to detect spatial and temporal patterns of declining forest productivity.Most importantly, we detected 4 cases of extreme drought, which were exceptional in the entire time period covered by the study. This allowed us to compare these cases and determine which one had the greatest impact on pine productivity, as well as to capture the time lag in pine response. This was possible thanks to remote sensing methods, which allowed us to examine the situation 20 years in the past and utilize a large statistical sample nationwide.
- The Abstract ends with reporting some of the result without providing any insight what this finding actually means, what are the implications, why is it relevant.
Thank your for this comment, we will add additional text at the end of the abstract.
- The introduction text is very uninspiring and underdeveloped (there were too many weak sentences to comment out on). The carbon accounting terminology is used wrongly (NPP confused with NEP). The comparisons of the literature is based on handpicked papers that they don’t support fully the statements made (e.g., lines 95 -98).
Thank you for your comment. We will take a look at the introduction to our article and make changes that will enhance its theoretical value. We have already amended the section on the definition of NPP to make it more comprehensive.
- Results and Discussions are mixed.
- No mechanistic understanding is gained from the analysis. It is descriptive based on basic and very limiting method (simple linear regressions to conclude on rather non-linear and complex interaction of meteorological variables NPP response). Which is why they do not find a consistent response across all months in Figure 8.
Our goal was to present the analysis as clearly and transparently as possible, ensuring that the obtained results are understandable and comparable with others in a practical context. We are aware that vegetation responses to meteorological variables are complex and nonlinear processes. However, the use of simpler methods (such as linear regression, Perason correlation, Z-scores) was deliberate – it allowed us to clearly define and detect 4 extreme drought cases and compare them with each other. We know that meteorological factors are not the only factors influencing productivity, but the aim of this study was to focus on them due to their changeable nature and to compare the results for each of the 4 extreme droughts we detected.
Furthermore, despite the limitations of this method, the analysis identified clear temporal and spatial patterns that confirm the usefulness of the approach. The fact that a clear response was not obtained across all months (Figure 8) in our opinion does not indicate a weakness of the method, but rather reflects the true complexity of Scots pine responses across the growing season (summer vs spring drought). In addition to the descriptive presentation of our methodology, we presented a flowchart of each stage of our analysis and developed formulas for calculating individual components (fig. 2).
- The authors mention “A key aspect of the study was to assess the impact of climatic factors on the condition of pine to determine its potential vulnerability to climate change,” but with the analysis they put forward they do not scratch the topic of vulnerability even on the surface (what is their definition of vulnerability? Not clear).
Thank you for your comment. We will include some additional information about this in the introduction.
- “In general, this study indicates an overall impact of meteorological condition on productivity of pine forest.” Yes, we know this. What was the expectation/hypothesis of the authors when they started exploring the data? Not clear.
Yes, we will provide additional text according to the purpose of the study to clarify this.
- Specific comments:
- Line 17: occurred in Poland in 2002-2023 à in the period, or during
Thank you, we have changed that.
- Line 27: the statement is very confusing. The negative effect of temperature on NPP this could also happen independent of the drought. The logic doesn’t make sense to me. Consider revising the statement please.
Thank you, we have made changes to this sentence to make it more clear. This is related to the time-lag correlation analysis (Fig. 9), in which we wanted to investigate the impact of each of 4 detected extreme droughts on NPP and how long this impact lasted.
- Line 37: “main” in terms of what? Coverage? Carbon storage?
Thank you, we provided changes to this sentence.
- Line 46: what is forest stress?
Thank you, we changed that according to the cited literature.
- Line 68: no, there are numerous studies focusing on individual species (depending on the methodology)
Thank you, we have clarified what we wanted to emphasize in this sentence and changed it according to your comment.
- Line 70: typo in a small area
Thank you, we have changed that.
- Line 83: no, net carbon sequestration of the forest depends also on the soil respiration.
Thank you, yes, we added this into the descripiton.
- Line 88: mitochondrial respiration is not the only type of plant respiration. Plants release CO2 also via photorespiration
Thank you, yes we added it to this sentence.
- Line 100: but you don’t identify dieback here. You identify effect on productivity.
Thank you, we changed that.
- Line 105: two decades
Thank you, we changed that.
- Line 109: why is there so much emphasis on the spectral response?
As user’s guide of used MODIS MOD17A2HGF data described:
The derivation of a satellite estimate of terrestrial NPP has three theoretical components: (1) the idea that plant NPP is directly related to absorbed solar energy, (2) the theory that a connection exists between absorbed solar energy and satellite-derived spectral indices of vegetation, and (3) the assumption that there are biophysical reasons why the actual conversion efficiency of absorbed solar energy may be reduced below the theoretical potential value.
That is why we emphasize the importance of spectral response, because the product we use is based on remote sensing indicators. This makes it possible to monitor productivity on a large spatial and temporal scale.
- Line 109: and get to what? What is the goal here? How would the correlation help give new insights? Is there a hypothesis how the response of the trees should be and why?
Thank you, we added extra text according to your comment.
- Line 118: What is multi-annual air temperature? Annual mean?
The multi-annual average air temperature. Thank you we changed that according to cited literature.
- Line 119: last 70 years please report the exact time period
Thank you we changed that
- Line 161-162: but the light use efficiency and APAR are used to derive GPP. How NPP is derived in not clear.
Thank you, yes, we have clarified this by adding additional text.
- Line 171: meteorological elements? You mean drivers?
Yes, We use „elements” throughout the article to ensure consistent use of vocabulary.
- Line 178: goes out of what into what?
We added extra text to clarified this sentence.
- Line 232: What about May?
- Line 236: it is not clear at this point why these four months.
Thank you, we have clarified this by placing the right information in the right order to ensure a good logical progression of the proposed methodology.
- Line 244: but what is the value of a 1:1 correlation when the interaction of the drivers is not captured? It does not provide a useful theoretical understanding.
- Line 248: the method that is used is not sufficient to capture the interaction of drivers
Our primary goal was to first identify extreme droughts and then test whether remote sensing data on Scots pine productivity provided a meaningful answer. Yes, we calculated the correlation individually and reported the correlation coefficients, rather than using models with combined meteorological components. We will add more information on this topic to emphasize the „individual” approach we are taking. Since all meteorological components are interrelated, it is obvious that large correlations will be observed not only for one component, but sometimes also for others. More importantly, we wanted to test the usability of using remote sensing data and analyses to obtain information on the impacts of the four most severe droughts over a large and representative area.
Here we are providing our published series of articles using similar methods and datasets:
Kulesza, K., Hościło, A. 2024. Coherency and time lag analyses between MODIS vegetation indices and climate across forests and grasslands in the European temperate zone. Biogeosciences, 21(10), 2509-2527. DOI: 10.5194/bg-21-2509-2024.
Kulesza, K., Hościło, A. 2024. Temporal Patterns of Vegetation Greenness for the Main Forest-Forming Tree Species in the European Temperate Zone. Remote Sensing, 16(15), 2844. DOI: 10.3390/rs16152844.
Kulesza K, Hawryło P, Socha J, Hościło A. How Reliable Are the Spectral Vegetation Indices for the Assessment of Tree Condition and Mortality in European Temporal Forests? Remote Sensing. 2025; 17(15):2549. https://doi.org/10.3390/rs17152549
- Line 276: please remove dots from units
Thank you, yes, we can changed that.
- Line 277: “The lowest one was observed for December 2021 (-0.004 kgC∙m-2∙month-1)” is this value reliable? Is it within the uncertainty range?
- 5: it is not clear how much is the uncertainty around each of these points. This should be displayed.
According to the user’s guide:
The MODIS MOD17A2HGF product is processed at level 4. Hence, the gap-filled MOD17A2HGF is the improved MOD17, which has cleaned the poor-quality inputs from 8-day Leaf Area Index and Fraction of Photosynthetically Active Radiation (FPAR/LAI) based on the Quality Control (QC) label for every pixel.
We will add this information to the manuscript.
Since these are monthly sums, it is possible that in the winter months the absolute values of NPP may be negative or close to 0 (in units of kilograms), because in these months (e.g. December) there is limited photosynthesis.
- Line 306: the effect of drought and extremely high temperature is mixed. But the study claims to be focused on the effect of drought. If z-scores of temperatures are calculated the extreme heat events should be considered next to drought. Otherwise, how do the authors know for sure that the impact is due to drought if at the same time temperatures were also extremely high?
In our analysis, drought was not identified only based on precipitation deficiencies. Following our methodology, we consider it as the result of the interaction of two key factors: Climatic Water Balance and temperature. In the drought detection process, we considered both of the elements („The months with the highest ZT_dtm.i and lowest ZCWB_dtm,i occurring simultaneously, are indicated as drought events”), ensuring that extreme heat events are not treated independently but are an integral part of the drought characteristics. In other words, the influence of high temperatures was explicitly considered in drought identification, so the results refer to the cumulative impact of precipitation deficiencies and extreme heat, not just one of these factors.
We are providing all the formulas in section 2.5 (Statistical analyses).
- Figure 7: I am not sure what to conclude from the NPP maps three months after the extreme months when NPP response depends on the climatic conditions in these following months, which is not clear in this map.
The purpose of presenting NPP values not only during the drought month but also in the subsequent three months was to capture possible lagged responses of the pine forest to drought. We acknowledge that in the following months NPP also depends on the current climatic conditions. However, this approach allows us to investigate whether we observe prolonged negative impacts on NPP in the 4 detected drought events, and whether these decreased productivity persist beyond the immediate post-drought period. Therefore, Figure 7 is intended to serve as an illustration of the post-drought situation and the potential delayed effects of drought.
We acknowledge that in the following months NPP also depends on the actual climatic conditions, and therefore the interpretation of the maps cannot be purely quantitative. However, this was not our intention. Instead, Figure 7 is meant to provide a qualitative perspective on whether the effects of drought extend beyond the immediate event.
By focusing on the 4 strongest drought episodes, we aimed to illustrate that these extreme events were followed by notable reductions in NPP, in some cases persisting into the subsequent months. This approach allows us to highlight the exceptional strength of these droughts and their potential to trigger delayed pine responses.
In this sense, the maps should be read as an illustration of the post-drought situation, rather than as a mechanistic explanation. Their added value lies in providing a spatially explicit picture of how the forest system recovered or failed to recover after the four most severe drought events we detected.
- Line 362-364: this belongs to Discussion
Thank you, we will move this sentence to Discussion.
- 8: trendlines should not be displayed if the relationship is not significant
Thank you, but we will stick to our version to maintain the graphical consistency of the figure.
- 9: the caption says “All the coefficients are statistically significant (a=0.05)”. But many of the coefficients are below 0.1 for which it is hard to imagine that the coefficients are significant, or at the very least they show no strong response.
Yes, this is due to the large number of homogeneous pine pixels used in the study (20,618 – Lines 151). This influences the number of correlation pairs, which directly contributes to the statistical significance of the analyses. We can emphasize that in the text.
Citation: https://doi.org/10.5194/egusphere-2025-2770-AC1
-
AC1: 'Reply on RC1', Oliwier Zając, 11 Sep 2025
reply
-
RC2: 'Comment on egusphere-2025-2770', Anonymous Referee #2, 31 Oct 2025
reply
The study focuses on the assessment of drought periods impacting net primary productivity (NPP) of Scots pine stands in Poland over the period 2002-2023. NPP proxy is obtained from MODIS and meteorology from ERA-5 reanalysis. Authors use z-scores of air temperature and Climatic Water Balance (CWB) to identify severe drought and examine correlation of NPP standardized anomalies with anomalies in meteorological variables. Authors picked 4 most extreme cases and provided spatial distribution of NPP proxy anomalies. I appreciate the amount of work authors put into the analysis and detailed description of the methods.
My major concerns with the study, as currently presented, are the lack of hypothesis, motivation and no clear way to generalize the results. Indeed, it is important to study drought impacts on forest productivity. But it is not clear what new aspects did we learn here about this topic, and how it could be applied. While computing z-scores seems to be a practical way to find anomalies, the outcomes are unlikely to be robust if temporal or spatial extent of the analysis were modified – again pointing to a lack of clear hypothesis. A connected issue is that severe drought is defined using area-averaged metric and thus does not account for the local conditions and does not allow to distinguish regional differences or areas more suitable for more optimal growth of Scots pine. Selection of the most severe drought events seems to be arbitrary (four cases) and z-scores (relative metric) do not allow to define absolute thresholds (generalization). Considering the chosen journal, I would assume some type of analysis connecting the remote sensing results with ground-based measurements to support the findings (e.g. PL-Tcz seems to be an existing FLUXNET site with Scots pine).
Abstract surprisingly emphasizes absolute values of NPP during drought months without context (percentual deviation from sample mean or similar). Major focus of the abstract is on the difference between spring and summer NPP response to drought. However, I think the authors incorrectly interpret the results. While z-score of T and CWB can be a good predictor of drought in the summer, I think that in April 2009 they detected early spring (unlikely water availability limitation of NPP). That is exceptionally warm conditions that would be expected to promote growth and can actually compensate for reduced summer NPP (see e.g. Wolf et al., 2016; Kowalska et al., 2020)
The lack of well-defined hypothesis leads to an unfocused Introduction. The topics are as follows: 1) importance of forests, 2) stress response, 3) global warming and droughts, 4) utilization of remotely-sensed NPP, 5) importance of drought impact assessment. Especially at L68 authors seem to sell the drawback of their study as an advantage when focusing solely on a single species without going deeper into why. If the dieback is to be studied, the stand dieback occurrence should be compared with the remote sensing (RS) NPP. If this is expected to have relevance for forestry sector, comparison with forest inventories would be expected, etc. Generally Introduction requires a thorough rewrite, considering the definition of specific hypothesis and focusing the chapters on a particular topic. I would also generally suggest language correction after the revisions.
Methods are very well developed, the issue is their design, as described above. In my view, providing mean climatic characteristics for the whole Poland seems a very coarse approach. Also, I would suggest to emphasize that MODIS NPP estimate is not direct measure of NPP, e.g. by using RS as subscript or similar. Considering drought severity identification, I would expect more relevant to use Aridity index (P/ETo) that can help to compensate for local differences in precipitation (the interpretation of CWB = -10 mm would be quite different for site with 1500 mm vs 500 mm annual P). The whole study is carried out in relative terms and actual characterization of T and P is provided only for April 2009. Correlations of ZNPP with time-lagged meteorological anomalies do not seem to be very convincing as it could be simply an outcome of the continuation of drought conditions following the month instead of a demonstration of severe drought legacy effect (this requires much more nuance to tease out).
I appreciate Figs 4 and 5 in the results that show actual NPP(RS) values. Considering Fig. 8, it seems to show that correlations are significant only for ZT and for all variables in April. As ZP and ZETo do not seem to help, I assume they should be removed. Additionally, I would find more interesting the relationship between NPP and T (or CWB) rather than Z metrics. Significant results for April across Z metrics are not surprising, as a slight shift in the start of the active season will greatly affect NPP.
In the Discussions, I think my main question is why do we need yet another metric to identify drought? How would the ZT+ZCWB compare to e.g. SPEI index that is widely used in literature?
In the Conclusions, authors suggest to perform similar study also for other tree species and longer time series. Due to the relative nature of the results, it is unclear on which ground they would be comparable and how exactly does this or similar studies assist with the changes of forestry practices in response to climate change.
Minor comments
L15: Emphasis on unexpected aspects – e.g. forest-forming (trees), spectral (responses)
L18: Not dealing with trees but pixels falling into certain Land-use.
L18: This is not actual productivity but proxy for that productivity
L22: use gC m-2 month-1
L76: Introduction becoming methods from here.
Wolf et al. (2016): https://doi.org/10.1073/pnas.1519620113
Kowalska et al. (2020): http://dx.doi.org/10.1098/rstb.2019.0518
Citation: https://doi.org/10.5194/egusphere-2025-2770-RC2 -
AC2: 'Reply on RC2', Oliwier Zając, 03 Nov 2025
reply
We sincerely thank the reviewer for the thoughtful and valuable comments, which which will significantly improve our manuscript. Our detailed responses and explanations for the revisions are provided below.
- My major concerns with the study, as currently presented, are the lack of hypothesis, motivation and no clear way to generalize the results. Indeed, it is important to study drought impacts on forest productivity. But it is not clear what new aspects did we learn here about this topic, and how it could be applied. While computing z-scores seems to be a practical way to find anomalies, the outcomes are unlikely to be robust if temporal or spatial extent of the analysis were modified – again pointing to a lack of clear hypothesis. A connected issue is that severe drought is defined using area-averaged metric and thus does not account for the local conditions and does not allow to distinguish regional differences or areas more suitable for more optimal growth of Scots pine. Selection of the most severe drought events seems to be arbitrary (four cases) and z-scores (relative metric) do not allow to define absolute thresholds (generalization). Considering the chosen journal, I would assume some type of analysis connecting the remote sensing results with ground-based measurements to support the findings (e.g. PL-Tcz seems to be an existing FLUXNET site with Scots pine).
We appreciate the reviewer’s insightful comment and fully agree that a clearer formulation of the research hypothesis would strengthen the manuscript. We will add this at the end of the introduction. We want to stress that our aim was to examine whether MODIS-derived indicators of Scots pine productivity are sensitive to meteorological conditions associated with drought events. Our initial intention was to perform an exploratory assessment to evaluate the responsiveness of remotely sensed productivity metrics to drought-related weather patterns rather than to define absolute thresholds or spatially uniform criteria. We will clarify this purpose in the introduction and discussion to better reflect the study’s scope.
- Abstract surprisingly emphasizes absolute values of NPP during drought months without context (percentual deviation from sample mean or similar). Major focus of the abstract is on the difference between spring and summer NPP response to drought. However, I think the authors incorrectly interpret the results. While z-score of T and CWB can be a good predictor of drought in the summer, I think that in April 2009 they detected early spring (unlikely water availability limitation of NPP). That is exceptionally warm conditions that would be expected to promote growth and can actually compensate for reduced summer NPP (see e.g. Wolf et al., 2016; Kowalska et al., 2020)
We agree that the abstract should provide a clearer context for the reported absolute NPP values - we will include additional information on this in the abstract. In the revised version, we will describe what these values refer to and we will use z-scores for this purpose to better illustrate the magnitude of anomalies during drought periods.
We thank the reviewer for pointing out the potential misinterpretation of the spring 2009 event. In our approach, the detection of drought conditions was performed sequentially for each month independently. Therefore, the case identified for April 2009 reflects not only exceptionally high temperatures but also relatively low precipitation, as both temperature (T) and climatic water balance (CWB) were taken into account. Each monthly drought event was analyzed in relation to other occurrences within the same month (e.g., April compared to other Aprils).
We agree that the interpretation of this event could have been presented in a more cautious manner, emphasizing the potential nature of the cause rather than a definitive conclusion. We will revise the discussion and abstract accordingly to clarify this point and better reflect the uncertainty and context of the 2009 spring conditions.
- The lack of well-defined hypothesis leads to an unfocused Introduction. The topics are as follows: 1) importance of forests, 2) stress response, 3) global warming and droughts, 4) utilization of remotely-sensed NPP, 5) importance of drought impact assessment. Especially at L68 authors seem to sell the drawback of their study as an advantage when focusing solely on a single species without going deeper into why. If the dieback is to be studied, the stand dieback occurrence should be compared with the remote sensing (RS) NPP. If this is expected to have relevance for forestry sector, comparison with forest inventories would be expected, etc. Generally Introduction requires a thorough rewrite, considering the definition of specific hypothesis and focusing the chapters on a particular topic. I would also generally suggest language correction after the revisions. We acknowledge that the Introduction in the original version lacked focus and a clearly formulated hypothesis. This point was also raised by the first reviewer, and we are currently revising the section accordingly. In the updated version, we will define explicit research hypotheses and restructure the introduction to provide a good theoretical description in the context of NPP. Regarding the sentence in line 68, our intention was to emphasize that many studies focused on very localized areas and thus only analyzed single species, whereas our work provides a national-scale analysis for the dominant species as highlighted in the study area description. We are aware that we should develop this issue further by providing more examples and revising this statement. We recognize that the discussion related to forest dieback was not essential for the current scope of the paper, and in the revised version, we have focused on the theoretical background and interpretation of NPP. The introduction has also been linguistically refined to improve clarity and readability following the reviewer’s suggestion. · Methods are very well developed, the issue is their design, as described above. In my view, providing mean climatic characteristics for the whole Poland seems a very coarse approach. Also, I would suggest to emphasize that MODIS NPP estimate is not direct measure of NPP, e.g. by using RS as subscript or similar. Considering drought severity identification, I would expect more relevant to use Aridity index (P/ETo) that can help to compensate for local differences in precipitation (the interpretation of CWB = -10 mm would be quite different for site with 1500 mm vs 500 mm annual P). The whole study is carried out in relative terms and actual characterization of T and P is provided only for April 2009. Correlations of ZNPP with time-lagged meteorological anomalies do not seem to be very convincing as it could be simply an outcome of the continuation of drought conditions following the month instead of a demonstration of severe drought legacy effect (this requires much more nuance to tease out).
We agree that providing mean climatic characteristics for the entire country represents a generalized view; however, this approach was chosen to ensure methodological consistency. Our main focus was on identifying the relative timing of drought events rather than on the absolute climatic variability across sites.
With respect to drought identification, we used both temperature (T) and climatic water balance (CWB) to detect drought events, as this combination accounts for both thermal and moisture-related stressors that affect NPP dynamics. We acknowledge that indices such as the Aridity Index (P/ETo) could provide an additional perspective by accounting for regional precipitation differences. However, our primary goal was not to assess local conditions or perform fine-scale spatial analyses, but rather to evaluate the sensitivity of MODIS-derived RS-NPP to drought occurrence and, consequently, to examine the response of Scots pine productivity. We will mention this limitation and note the potential use of alternative indices in the revised manuscript.
We also agree that MODIS-derived NPP represents an indirect, model-based estimate of productivity rather than a direct measurement. We already clarified this in new version according to the first reviewer suggestions.
We are aware of the potential concern regarding correlations between ZNPP and time-lagged meteorological anomalies. However, in our study, drought detection was performed sequentially for each month independently, and all months within the study period were analyzed. As a result, the extreme drought conditions did not continue in the subsequent months, and the observed correlations reflect the relative response of MODIS-derived NPP to the specific drought conditions in each drought-detected month. This allows us to find out whether there is a potential prolonged impact of these specific droughts on reduced NPP values. · I appreciate Figs 4 and 5 in the results that show actual NPP(RS) values. Considering Fig. 8, it seems to show that correlations are significant only for ZT and for all variables in April. As ZP and ZETo do not seem to help, I assume they should be removed. Additionally, I would find more interesting the relationship between NPP and T (or CWB) rather than Z metrics. Significant results for April across Z metrics are not surprising, as a slight shift in the start of the active season will greatly affect NPP. We thank the reviewer for the positive feedback on Figures 4 and 5 and for the thoughtful comments regarding Figure 8. We appreciate the suggestion to focus only on the significant variables; however, we prefer to present all results to provide full transparency regarding our methodological approach and the range of data used.
- In the Discussions, I think my main question is why do we need yet another metric to identify drought? How would the ZT+ZCWB compare to e.g. SPEI index that is widely used in literature?
We thank the reviewer for this insightful comment. We use ERA5 data and the selected meteorological variables to investigate how Scots pine productivity responds to individual climatic factors. Following this approach, we also apply the same dataset for drought detection, which ensures a consistent and coherent methodological framework throughout the study. This allows us to link NPP responses directly to drought conditions and provides a seamless and transparent workflow, rather than introducing an entirely separate metric.
- In the Conclusions, authors suggest to perform similar study also for other tree species and longer time series. Due to the relative nature of the results, it is unclear on which ground they would be comparable and how exactly does this or similar studies assist with the changes of forestry practices in response to climate change.
We thank the reviewer for this comment. We will add some information on this topic in the conclusions. Our intention was to emphasize the importance of updating and extending the study in future studies as further drought events occur, allowing us to refine our understanding of Scots pine's response over time. However, due to the relatively large pixel size of
500x500 m, extending this analysis to a larger number of tree species will be difficult, as pixel heterogeneity and the limited number of homogeneous areas for each species would complicate interpretation. This is a general challenge for similar studies in other regions or with other species, and we will briefly discuss these limitations and considerations in the revised conclusions. ·- Minor comments· L15: Emphasis on unexpected aspects – e.g. forest-forming (trees), spectral (responses)· L18: Not dealing with trees but pixels falling into certain Land-use.· L18: This is not actual productivity but proxy for that productivity· L22: use gC m-2 month-1· L76: Introduction becoming methods from here.
We thank the reviewer for these helpful minor comments. We will incorporate the suggested clarifications and corrections in the revised manuscript.
Citation: https://doi.org/10.5194/egusphere-2025-2770-AC2
-
AC2: 'Reply on RC2', Oliwier Zając, 03 Nov 2025
reply
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 1,110 | 56 | 20 | 1,186 | 27 | 29 |
- HTML: 1,110
- PDF: 56
- XML: 20
- Total: 1,186
- BibTeX: 27
- EndNote: 29
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The article “Extreme droughts’ impact on Scots pine net primary productivity in the European temperate zone in the period 2002-2023” by Kulesza et al. investigated the impact of droughts (based on monthly climatic water balance anomalies) on the NPP of Scot pine forests in Poland, between 2002-2023.
The authors identified particularly 4 summer months 2006, 2009, 2015 and 2019 where forest NPP (detected from MODIS NPP) was significantly affected.
The title mentions „European temperate zone “but the study is focused exclusively on forests in Poland. While the forests studied here are indeed characteristically temperate but the spatial coverage of the study and the number of species (only 1) does not capture the European temperate zone forests. The title is thus misleading.
The objectives of the study are not novel. What do we want to learn from finding a drought that has the most negative impact, Is not clear. Also the results cannot be generalized even for this single species, because the response of Scot pines depend on the climatic and soil conditions they are adapted to. The study has not included the complete distribution of Scot pines across Europe and the diversity of local conditions in which they grow.
The Abstract ends with reporting some of the result without providing any insight what this finding actually means, what are the implications, why is it relevant.
The introduction text is very uninspiring and underdeveloped (there were too many weak sentences to comment out on). The carbon accounting terminology is used wrongly (NPP confused with NEP). The comparisons of the literature is based on handpicked papers that they don’t support fully the statements made (e.g., lines 95 -98).
Results and Discussions are mixed.
No mechanistic understanding is gained from the analysis. It is descriptive based on basic and very limiting method (simple linear regressions to conclude on rather non-linear and complex interaction of meteorological variables NPP response). Which is why they do not find a consistent response across all months in Figure 8.
The authors mention “A key aspect of the study was to assess the impact of climatic factors on the condition of pine to determine its potential vulnerability to climate change,” but with the analysis they put forward they do not scratch the topic of vulnerability even on the surface (what is their definition of vulnerability? Not clear).
“In general, this study indicates an overall impact of meteorological condition on productivity of pine forest.” Yes, we know this. What was the expectation/hypothesis of the authors when they started exploring the data? Not clear.
Specific comments:
Line 17: occurred in Poland in 2002-2023 in the period, or during
Line 27: the statement is very confusing. The negative effect of temperature on NPP this could also happen independent of the drought. The logic doesn’t make sense to me. Consider revising the statement please.
Line 37: “main” in terms of what? Coverage? Carbon storage?
Line 46: what is forest stress?
Line 68: no, there are numerous studies focusing on individual species (depending on the methodology)
Line 70: typo in a small area
Line 83: no, net carbon sequestration of the forest depends also on the soil respiration.
Line 88: mitochondrial respiration is not the only type of plant respiration. Plants release CO2 also via photorespiration
Line 100: but you don’t identify dieback here. You identify effect on productivity.
Line 105: two decades
Line 109: why is there so much emphasis on the spectral response?
Line 109: and get to what? What is the goal here? How would the correlation help give new insights? Is there a hypothesis how the response of the trees should be and why?
Line 118: What is multi-annual air temperature? Annual mean?
Line 119: last 70 years please report the exact time period
Line 161-162: but the light use efficiency and APAR are used to derive GPP. How NPP is derived in not clear.
Line 171: meteorological elements? You mean drivers?
Line 178: goes out of what into what?
Line 232: What about May?
Line 236: it is not clear at this point why these four months.
Line 244: but what is the value of a 1:1 correlation when the interaction of the drivers is not captured? It does not provide a useful theoretical understanding.
Line 248: the method that is used is not sufficient to capture the interaction of drivers
Line 276: please remove dots from units
Line 277: “The lowest one was observed for December 2021 (-0.004 kgC∙m-2∙month-1)” is this value reliable? Is it within the uncertainty range?
Fig. 5: it is not clear how much is the uncertainty around each of these points. This should be displayed.
Line 306: the effect of drought and extremely high temperature is mixed. But the study claims to be focused on the effect of drought. If z-scores of temperatures are calculated the extreme heat events should be considered next to drought. Otherwise, how do the authors know for sure that the impact is due to drought if at the same time temperatures were also extremely high?
Figure 7: I am not sure what to conclude from the NPP maps three months after the extreme months when NPP response depends on the climatic conditions in these following months, which is not clear in this map.
Line 362-364: this belongs to Discussion
Fig. 8: trendlines should not be displayed if the relationship is not significant
Fig. 9: the caption says “All the coefficients are statistically significant (a=0.05)”. But many of the coefficients are below 0.1 for which it is hard to imagine that the coefficients are significant, or at the very least they show no strong response.