the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 06 Oct 2025
Matching scales of eddy covariance measurements and process-based modeling – Assessing spatiotemporal dynamics of carbon and water fluxes in a mixed forest in Southern Germany
Abstract. Eddy covariance (EC) measurements are a backbone of ecological research and have provided valuable insights into the variability of carbon and water fluxes in different ecosystems and under varying environmental conditions. Since these measurements are integrative and weighted over changing areas (footprint), species-specific information cannot be easily derived except for extended monocultures. However, EC sites are increasingly established in mixed forest stands which are considered to be more resilient under changing environmental conditions. This imposes the question of how species-specific responses can be derived, and how the magnitude of fluxes originating from temporally varying flux footprints predictions (FFPs) might provide insights into species-specific responses.
At a site in the Black Forest (southwestern Germany), which mainly consists of a mix of mature beech and Douglas fir trees, we investigate how EC flux measurements depend on different FFP areas and how species-specific contributions to gas exchange can be disentangled. We applied an ecosystem model that has been calibrated from EC measurements at various sites with beech- and Douglas fir monocultures, and evaluated it with data of soil water content and soil respiration taken at homogeneous parts of the investigated mixed forest site. Then we compared hourly aggregated measurements of net carbon exchange (NEE) and evapotranspiration (ET) with model simulations under four configurations: (i) pure beech, (ii) pure Douglas fir, (iii) a static weighted average of both species, and (iv) a dynamic weighted average based on footprint variations.
The results show that weighted combinations of the two species generally provide a better match with hourly EC measurements than single-species simulations, while differences between static and dynamic weighting approaches remain relatively small. Nevertheless, specific-species responses to the environment can be significantly different during transitional periods such as autumn and spring when physiological differences between Douglas fir and beeches are most pronounced. We demonstrate that considering these differences is particularly important for gap-filling EC measurements and thus for determining annual carbon and water budgets. We herewith demonstrate that EC measurements over mixed forests provide important model evaluation information and that species-specific modeling is essential for untangling and distributing the underlying species-specific ecosystem dynamics.
- Preprint
(2556 KB) - Metadata XML
-
Supplement
(3798 KB) - BibTeX
- EndNote
Status: open (until 17 Nov 2025)
-
RC1: 'Comment on egusphere-2025-4605', Anonymous Referee #1, 17 Oct 2025
reply
-
CC1: 'Reply on RC1', Rüdiger Grote, 24 Oct 2025
reply
We are happy and grateful for the positive evaluation of the manuscript and fully comply with any request and suggestion.
Regarding the relative high number of gaps, we admit that although the intake tube of the eddy covariance system is principally heated, this heating system was not working in autumn up until we exchanged the tube beginning of December 2024. Before this change, filters were often clogged after fog and rain events which was no problem anymore afterwards.
The reason why we used soil moisture and respiration for evaluation was the availability of separate measurements for Douglas fir and beech dominated plots, respectively. This should indicate that the model is representing fluxes species-specifically in order to be sure that lumped up NEE and ET from eddy covariance measurements are correctly represented for the right reasons. However, we suggest to add a figure with measurements and weighted simulations of NEE and ET in daily resolution throughout the year (see preliminary sketch attached).
We fully comply with the other suggestions, including adding that conventional approaches of gap filling can provide similar results than the applied model. We will address this issue in the final revision of the manuscript.
-
CC1: 'Reply on RC1', Rüdiger Grote, 24 Oct 2025
reply
-
RC2: 'Comment on egusphere-2025-4605', Anonymous Referee #2, 06 Nov 2025
reply
Review of “Matching scales of eddy covariance measurements and process-based modeling - Assessing spatiotemporal dynamics of carbon and water fluxes in a mixed forest in Southern Germany” by Moutahir et al., submitted to EGUsphere, 2025.
General
This is an interesting paper about our capability of measuring and modelling carbon and water fluxes from biodiverse forests in contrast to monocultures. The authors argue that biodiverse forests are said to be more resilient to climatic, environmental and biotic stressors and that hence it is important to distinguish exchange between monocultures and biodiverse forests as well as to identify the contribution of individual species to the total. This is indeed a relevant and timely topic in both measuring and modelling.
The authors use eddy covariance measurements from monocultures to train and evaluate an ecosystem model. Then they apply the model to a mixed forest (63% Beech, 27% Fir) and confront the quality of the results with EC measurements over that forest. In order to quantify the relative contribution of the two species, the authors use short-term flux footprint predictions (FFPs) overlain on a detailed species map.
The objectives, as stated in the introduction, are: 1) to assess the influence of the different species composition on LE and NEE in the mixed forest and 2) to analyse the uncertainty resulting from the assumption of a homogeneous footprint for mixed forest.
This suggests that the paper reports about a methodological question: To what level can we use EC measurements over a mixed forest in combination with footprint analysis to identify the individual contributions of species in a mixed forest. The conclusions, in contrast, focus on the quality of the model effort and less about the capability to measure the contribution of species to the whole.
In principle, I am positive to the aim of the paper and about the setup of the research. It addresses an important question and the methods are carefully chosen. Unfortunately, however, there are some limitations to this study too.
The first is: the paper does not really answer the question how well, in principle, the combination of EC flux measurements and footprint distribution can be attributed to contribution of the two species. Both EC measurements and footprint models come with uncertainties. Suppose the footprint model has an uncertainty in footprint distance of, let’s say, 15%. How will the distribution of ‘detected’ species change? Given such an uncertainty, is it feasible to actually quantify the contribution of species accurately enough? The answer obviously depends on the way the species are mixed. If the species are normally distributed, it will be difficult to separate the effects, whereas it would be much easier of the two species are somewhat clustered (but not too much, for it wouldn’t be a biodiverse forest anymore). In my opinion, the authors should consider the impact of footprint accuracy and species distribution before even trying to answer the research questions.
The second is: the relative contribution of the two species in the mixed forest is nearly the same in all wind directions. Of course this is beyond control of the authors, but it severely limits the applicability of the methods to answer the research questions. The authors do not really go into the variability of the relative contribution with stability and wind speed, perhaps the actual distribution over species is more variable than Fig. 6b suggests.
The third is: The NEE and LE of Beech, Fir and mixed forest do not seem to differ a lot. As a result, the model-observation comparison all tend to yield similar results (Figs. 7-9 and tables 2, 3 and 4). The authors claim that the model with static or dynamic footprint perform better than the pure Beech or pure Fir models. However, I do not really see a major difference. The r2, slope and NSE of the mixed models are indeed somewhat better, but overall, the r2 and NSE are rather small. In fact, the performance of the model to monocultures of Beech and Fir (Figs. S1 and S2) are really poor, particularly for Fir (r2 = -0.26 and 0.24 for the two sites).
So, in principle, this is a great research topic. However, the authors seem to be more focused on applying a designed methodology than to actually answer the research question – if EC measurements in combination with footprint analysis could be a way to identify the contribution of individual species. Additionally, the measurement and model results for the two tree species tend to be rather similar, which makes it difficult to tell if the species attribution method actually works or not. So the conclusions should be negative: we cannot assess the influence of the different species and the uncertainty of resulting form the assumption of homogeneous footprint cannot be identified with the current methods.
Concludingly, although the research question and the methods are interesting and promising, I advise against publication in EGUspheres. Several elements of the study could be publishable though, for example, an analysis of the feasibility of the footprint analysis to attribute to individual species, or the model results after better validation and perhaps going more into differences in fluxes.
A somewhat larger question would be how the authors want to identify the effect of species living together in a ecosystem. If mixed forests are more resilient than monocultures, this must somehow show as a non-linear response to mixing monocultures.
Specific remarks:
Line 83: do you have references to these model studies?
Fig. 1: Could you apply the same colour codes in Fig 1 a and b? E.g. Beech is purple in the left figure and blue in the right one.
Line 106: Could you please go into the way that the tree species are clustered? And does the clustering have effect on the canopy height distribution? Does this affect the eddy covariance measurements?
Line 133: which data quality filter did you apply in the obs-model comparison in Figs 7-9 and tables 2 -4?
Line 157: the method of measuring soil respiration is not clear to me. Did you rotate between plots? And how long did you measure per plot?
Line 160: the model description is not clear to me. Is the model based on an ecosystem or individual trees? Does the model include carbon stocks in stems, branches, leaves, roots and soil? Does the model include allocation to these stocks? And phenology? Does leaf fall contribute to soil carbon stocks?
Line 181: You describe that you studied model performance, but you do not describe how well the model preformed. Could you please?
Line 184: 1.5 years for soil spin up sounds extremely short, particularly for the more stable pools. Please explain.
Line 195; ‘The two initialisations … (Fig1, right)’ à not clear what you mean here. Could you clarify?
Line 208: I do not know these forests too well, but I thought Speulderbos was not a Fir monoculture.
Line 210: Why do you only show daily numbers, not hourly ones, or diurnal cycles? The daily statistics of the different species are quite similar and may hide responses hidden in photosynthetical / stomatal / light responses of the species. A half hourly time-scale would be more appropriate from this perspective, while the footprint is variable at that time-scale too.
Line 214: ‘reasonably well’ à can you quantify?
Line 215, Figs S1 and S2: I do not consider comparisons with r2 - -0.26 and 0.24 ‘reasonably well’, this is an uncorrelated point cloud. Before applying the model to mixed forests, the model should be applicable to monocultures first.
Lines 216-218: Nice that you analysed the model performance. Could you go into the results?
Section 3.2: Overall, fig 6c suggests that the relative contribution of Beech and Fir are quite similar in all wind directions. In this situation, how can you distinguish between species? Don’t you need contrasting contributions to single out the contributions of the species?
Figs. 7-9 and tables 2-3: Would it be an idea to include the statistics information in the figure panels for easier dissemination?
Section 3.3: I would say all obs-model comparison perform similarly. The statistics of the fixed and dynamically weighted model is not considerably better.
Line 401: ‘Overall, … shows the closest agreement’ à That depends on which criteria you apply. I would say all model approaches perform similarly well considering annual NEE and timing, except perhaps the Fir monoculture simulation.
Line 440: ‘…clearly superior…’: The improvement is minor. Please refrain from overselling your results.
Line 417: ‘succesfully reproduced key …’: yes for soil moisture and soil respiration, but not for NEE and LE (Supplementary information).
References: all DOI’s miss a colon in the hyperlink, please change ‘https//’ to ‘https://’
Citation: https://doi.org/10.5194/egusphere-2025-4605-RC2 -
CC2: 'Reply on RC2', Rüdiger Grote, 14 Nov 2025
reply
Response to general remarks:
We acknowledge the criticism expressed and are thankful for the warning not to overstate the results and the advice to appropriately address limitations of the study. The critical issues mostly originate from an uncertain evaluation regarding Douglas fir monocultures, and from the similar contribution of both investigated species within the investigated footprints. While we have to admit that the database for evaluation is quite limited and that the field conditions are not ideal for investigating species-specific footprint contributions, we think that with some modifications, the manuscript can still provide useful and interesting information worth publishing.
In general, we suggest first to modify the objectives in a way to scale down expectations (with impacts on discussion and conclusions), second to put more focus on specific seasons, during which the differences between the tree species are more apparent, and third to elaborate on the evaluations. We also like to point out that despite similar gap-filling results obtained from process-based modelling as well as conventional approaches, simulations allow better analyses of tree species contributions and more reliable flux prediction under changed environmental conditions. In the following three paragraphs we would thus like to address the main general points of criticism, followed by a (mostly) point by point response regarding the specific comments.
First, the uncertainties connected to footprint accuracy and species distribution are not sufficiently addressed. It is certainly true that the question of species contribution to a footprint under field conditions can only be answered with considerable uncertainty, which we will acknowledge in a revised version. Following Kljun et al. (2015) we will argue that any footprint model is most accurate for flat and uniform surfaces and introduce errors in areas with variable topography, non-uniform canopy structures, or rapidly changing meteorological conditions, because it assumes stationary and horizontal homogeneity of the flow over the eddy-covariance integration period. Another uncertainty is to use only footprints containing more than 80% of the cumulative source contribution in the 2 x 2 km domain, which nevertheless can still be assumed sufficient to estimate the main impact areas of the flux measurements (the area containing 100% of the cumulative source contribution is infinite).
Second, the relative contribution of the two species in the mixed forest is similar in all wind directions, limiting the possibilities to determine the influence of different species to flux composition. It is true that the composition of species is similar in all cardinal directions, which hampers – but not prevents – differentiating flux contributions. As suggested, we will therefore additionally investigate the impact of footprint stability and wind speed, and carry out a sensitivity analysis considering various species contributions within a randomly distributed selection of footprints, which will also help quantifying part of the uncertainty addressed in the previous paragraph. In addition, we will further analyse the relation of species abundance and flux contribution with a specific focus on seasonal variability (see also next paragraph).
Third, because of a similarity of flux intensities in beech and Douglas fir, the contribution of each species is difficult to determine, particular if already monocultures could not be represented very well. Indeed, flux intensities are similar for both species in the summer period and also at the annual scale. However, the seasonal differences are much more expressed, which we will demonstrate by adding a seasonally differentiated evaluation, and highlight the importance of specific periods for cumulative annual fluxes. We will therefore put a focus of our analysis on the seasonal variability of species contribution, in particular spring and autumn, and add statistics to demonstrate that at least some footprint variability can be accounted for. It should also be noted that the poor performance of the fir evaluations is at least partly the result of a very high variability of measurements. We will make efforts to filter for outliers which will show that the comparison with data that are more rigorously quality checked will be better. In addition, we will not only show 1:1 regression but also the model performance throughout the year and (selectively) during the day (also see responses to specific questions).
Responses to detailed remarks:
Line 83: do you have references to these model studies?
-> We guess this question refers to the mentioned evaluation. Indeed, beech forests have been addressed with LandscapeDNDC or predecessor models before (Holst et al., 2010; Poschenrieder et al., 2013; Molina-Herrera et al., 2015; Mahnken et al., 2022) and we will add those references to the text. However, for the current exercise, additional data from new beech- and Douglas fir sites for which long-term EC data and good boundary information were available have been used for evaluation. Since these exercises are yet unpublished, the site PIs of Stitna, CZ Republic, and Speulderbos, NL, are also considered as co-authors.
Fig. 1: Could you apply the same color codes in Fig 1 a and b? E.g. Beech is purple in the left figure and blue in the right one.
-> We will happily follow this suggestion
Line 106: Could you please go into the way that the tree species are clustered? And does the clustering have effect on the canopy height distribution? Does this affect the eddy covariance measurements?
-> The tree species are generally clustered in groups although single tree intermixture occasionally occur. In the surrounding of the investigation center, Douglas fir and beech stands occur site by site with different heights but less intermixture. Thus, the region is mostly horizontally structured but vertically homogenous. Although some effects due to different heights of stands within a footprint cannot be excluded, this effect is assumed to be small. We will, however, investigate selected footprints with distinct structure and add this issue to the discussion.
Line 133: which data quality filter did you apply in the obs-model comparison in Figs 7-9 and tables 2 -4?
-> In this study, only CO2 fluxes and ET data with quality flags 0 or 1 were used for further analysis, while data with quality flag 2 were discarded. The REddyProc software was used for further processing of the EC data (Wutzler et al., 2018), including u*-filtering according to Papale et al. (2006). For the comparison with model derived cumulative values, we used gap filling by marginal distribution sampling (MDS) according to Reichstein et al. (2005). (This information has already been given in section 2.2.1 where Eddy covariance measurements are explained).
Line 157: the method of measuring soil respiration is not clear to me. Did you rotate between plots? And how long did you measure per plot?
-> We will add the requested information. A bias during the day was prevented by varying the measurement order of the plots randomly. The measurement time depended on the season and was within 90 – 140 seconds.
Line 160: the model description is not clear to me. Is the model based on an ecosystem or individual trees? Does the model include carbon stocks in stems, branches, leaves, roots and soil? Does the model include allocation to these stocks? And phenology? Does leaf fall contribute to soil carbon stocks?
-> We are sorry for the missing information and will improve the description accordingly. The vegetation module of the model framework is based on trees which represent cohorts of individuals with equal dimensions and homogeneous spatial distribution. The carbon balance in each cohort (in a monoculture there is only one) is based on uptake (photosynthesis), allocation (into leaves, sapwood, fine roots, and structural reserves), and loss (respiration and senescence) processes. All processes are calculated in dependence on environmental conditions with explicit consideration of phenology (based on cumulated temperature sums and chilling requirements). Vegetation and soil modules are in close interrelation with litterfall adding to the soil carbon pools and explicit calculation of decomposition processes supplying heterotrophic respiration.
Line 181: You describe that you studied model performance, but you do not describe how well the model performed. Could you please?
-> It is true that we kept the description of evaluations quite short. In a revised manuscript version, we will add a more comprehensive description of results, including performance over time. We will concentrate particular on the new sites included here for the first time but better acknowledge also previous results.
Line 184: 1.5 years for soil spin up sounds extremely short, particularly for the more stable pools. Please explain.
-> In contrast to models that use pre-runs to simulate site conditions that are in equilibrium with the environment, we rely mostly on explicitly measured conditions. A short spin-up period is nevertheless necessary because the initialization of recalcitrant, intermediate and labile carbon pools in the soil (incl. litter), which can only be estimated from soil carbon and nitrogen content, is inevitably uncertain. The actual period depends on the climate and how close the automated estimates are to the equilibrium conditions, but pre-runs have ensured that the applied spin-up is sufficient for the investigated site.
Line 195; ‘The two initializations … (Fig1, right)’ à not clear what you mean here. Could you clarify?
-> This refers to the site conditions initialized for the Douglas fir and the Beech site. Each site was initialized with tree size and density properties that are representative for the surrounding of the EC flux tower. We use these initializations for all possible footprints which is another source of uncertainty to be discussed in the revision. However, the variable most sensitive to NEE and ET fluxes is leaf area, which is basically the same in all forests of a specific species, provided that the forest is dense and the height is above a threshold value. Since this is generally the case in all of the region, we do not expect any major influence from the stand initialization on the fluxes.
Line 208: I do not know these forests too well, but I thought Speulderbos was not a Fir monoculture.
-> The forest in Speulderbos, also sometimes called Speulder forest, is indeed a pure 2.5 ha sized Douglas fir forest, planted in 1962 (Su et al., 2009). Surrounding forests with various tree species are generally thought to be of minor importance but we will mention this as a source of uncertainty that might be responsible for the relatively large number of outliers in the EC measurements.
Line 210: Why do you only show daily numbers, not hourly ones, or diurnal cycles? The daily statistics of the different species are quite similar and may hide responses hidden in photosynthetic / stomatal / light responses of the species. A half hourly time-scale would be more appropriate from this perspective, while the footprint is variable at that time-scale too.
-> Daily evaluation steps are chosen for easier perception. because sub-daily climate- and EC flux data are not available for all sites. However, it is true that it is useful to know about the reliability of the model at different scales. Therefore, we will demonstrate that the model is able to capture also sub-daily developments for selected fluxes.
Line 214: ‘reasonably well’ à can you quantify?
-> As mentioned before (response to question about L181), we will describe the evaluation results in more detail, also using various statistical measures to elaborate on the relationship between model and simulations.
Line 215, Figs S1 and S2: I do not consider comparisons with r2 - -0.26 and 0.24 ‘reasonably well’, this is an uncorrelated point cloud. Before applying the model to mixed forests, the model should be applicable to monocultures first.
-> We admit that the simple r2 correlation doesn’t look very good for the Douglas fir sites, particularly the site in the Netherlands. This is partly due to EC data that are quite noisy. We checked the data together with the site-PI and can now use an updated data set with additional quality checks. In addition, we can demonstrate that the deviations are particularly high in winter where only few measurements are available but the total contribution of fluxes is small (see also figure attached). The evaluation analysis will therefore be redone.
Lines 216-218: Nice that you analyzed the model performance. Could you go into the results?
-> Thanks again for pointing at this deficit which we will address appropriately in the revision (see also responses to question about L181 and L214)
Section 3.2: Overall, fig 6c suggests that the relative contribution of Beech and Fir are quite similar in all wind directions. In this situation, how can you distinguish between species? Don’t you need contrasting contributions to single out the contributions of the species?
-> As discussed already in the response to general remarks, we will add an investigation of the impact of footprint stability and wind speed, and carry out a sensitivity analysis considering various species contributions within a randomly distributed selection of footprints. In addition, we will further analyse the relation of species abundance and flux contribution with a specific focus on seasonal variability.
Figs. 7-9 and tables 2-3: Would it be an idea to include the statistics information in the figure panels for easier dissemination?
-> This is possible, and we can easily provide them. Most of this information is however, already presented in tables, which we then might shorten on the one hand but on the hand complement with more detailed seasonal analyses.
Section 3.3: I would say all obs-model comparison perform similarly. The statistics of the fixed and dynamically weighted model is not considerably better.
-> The differences between the r2 and bias of the various model settings are indeed small. However, we now have tested various other statistical measures which show that the representation of flux footprints is considerably better when considering both species compared with the assumption of monocultures. This performance difference is particularly apparent in spring and autumn periods (March plus April, and Oct. plus Nov.), during which also the r2 is significantly different (e.g. spring: 0.45 for mixtures, 0.39 for pure beech and 0.32 for pure Douglas fir).
Line 401: ‘Overall, … shows the closest agreement’ That depends on which criteria you apply. I would say all model approaches perform similarly well considering annual NEE and timing, except perhaps the Fir monoculture simulation.
-> Thanks for the suggestion. We will address this point more careful and indicate strong- and weak points of the evaluation, differentiated by species and site.
Line 440: ‘…clearly superior…’: The improvement is minor. Please refrain from overselling your results.
-> Thanks for the reminder. In the revision, we will be more careful not to do so.
Line 417: ‘successfully reproduced key …’: yes, for soil moisture and soil respiration, but not for NEE and LE (Supplementary information).
-> Same as before
References: all DOI’s miss a colon in the hyperlink, please change ‘https//’ to ‘https://’
-> Thanks for the hint. This is now changed in the template for generating references.
Mentioned references
Holst, J., Grote, R., Offermann, C., Ferrio, J. P., Gessler, A., Mayer, H., and Rennenberg, H.: Water fluxes within beech stands in complex terrain, Int. J. Biometeorol., 54, 23-36, https://doi.org/10.1007/s00484-009-0248-x, 2010.
Kljun, N., Calanca, P., Rotach, M. W., and Schmid, H. P.: A simple two-dimensional parameterisation for Flux Footprint Prediction (FFP), Geosci. Model Dev., 8, 3695-3713, https://doi.org/10.5194/gmd-8-3695-2015, 2015.
Mahnken, M., Cailleret, M., Collalti, A., Trotta, C., Biondo, C., D’Andrea, E., Dalmonech, D., Gina, M., Makela, A., Minunno, F., Peltoniemi, M., Trotsiuk, V., Nadal-Sala, D., Sabate, S., Vallet, P., Aussenac, R., Cameron, D., Bohn, F., Grote, R., and Augustynczik, A.: Accuracy, realism and general applicability of European forest models, Glob. Change Biol., 28, 6921-6943, https://doi.org/10.1111/gcb.16384, 2022.
Molina-Herrera, S., Grote, R., Santabárbara-Ruiz, I., Kraus, D., Klatt, S., Haas, E., Kiese, R., and Butterbach-Bahl, K.: Simulation of CO2 fluxes at European forest ecosystems with the coupled soil-vegetation process model “LandscapeDNDC”, Forests, 6, 1779-1809, https://doi.org/10.3390/f6061779, 2015.
Papale, D., Reichstein, M., Aubinet, M., Canfora, E., Bernhofer, C., Kutsch, W., Longdoz, B., Rambal, S., Valentini, R., Vesala, T., and Yakir, D.: Towards a standardized processing of Net Ecosystem Exchange measured with eddy covariance technique: algorithms and uncertainty estimation, Biogeosciences, 3, 571-583, https://doi.org/10.5194/bg-3-571-2006, 2006.
Poschenrieder, W., Grote, R., and Pretzsch, H.: Extending a physiological forest model by an observation-based tree competition module improves spatial representation of diameter growth, Eur. J. Forest Res., 132, 943-958, https://doi.org/10.1007/s10342-013-0730-1 2013.
Reichstein, M., Falge, E., Baldocchi, D., Papale, D., Aubinet, M., Berbigier, P., Bernhofer, C., Buchmann, N., Gilmanov, T., Granier, A., Grünwald, T., Havranokova, K., Ilvesniemi, H., Janous, D., Knohl, A., Laurila, T., Lohila, A., Loustau, D., Matteucci, G., Meyers, T., Miglietta, F., Ourcival, J.-M., Pumpanen, J., Rambal, S., Rotenberg, E., Sanz, M., Tenhunen, J., Seufert, G., Vaccari, F., Vesala, T., Yakir, D., and Valentini, R.: On the separation of net ecosystem exchange into assimilation and ecosystem respiration: review and improved algorithm, Glob. Change Biol., 11, 1-16, https://doi.org/10.1111/j.1365-2486.2005.001002.x, 2005.
Su, Z., Timmermans, W. J., van der Tol, C., Dost, R., Bianchi, R., Gómez, J. A., House, A., Hajnsek, I., Menenti, M., Magliulo, V., Esposito, M., Haarbrink, R., Bosveld, F., Rothe, R., Baltink, H. K., Vekerdy, Z., Sobrino, J. A., Timmermans, J., van Laake, P., Salama, S., van der Kwast, H., Claassen, E., Stolk, A., Jia, L., Moors, E., Hartogensis, O., and Gillespie, A.: EAGLE 2006 – Multi-purpose, multi-angle and multi-sensor in-situ and airborne campaigns over grassland and forest, Hydrol. Earth Syst. Sci., 13, 833-845, https://doi.org/10.5194/hess-13-833-2009, 2009.
Wutzler, T., Lucas-Moffat, A., Migliavacca, M., Knauer, J., Sickel, K., Šigut, L., Menzer, O., and Reichstein, M.: Basic and extensible post-processing of eddy covariance flux data with REddyProc, Biogeosciences, 15, 5015-5030, https://doi.org/10.5194/bg-15-5015-2018, 2018.
-
CC2: 'Reply on RC2', Rüdiger Grote, 14 Nov 2025
reply
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 319 | 46 | 14 | 379 | 27 | 10 | 11 |
- HTML: 319
- PDF: 46
- XML: 14
- Total: 379
- Supplement: 27
- BibTeX: 10
- EndNote: 11
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
General remarks:
The authors present a well-designed single-site study regarding footprint classified EC flux measurements to address tree species specific carbon and water budgets on different temporal scales. Overall, I only suggest minor changes.
The C budgets based on non-gap-filled and gap-filled fluxes differ considerably maybe due to more frequent gaps or fluxes with flag2 during winter. This is somewhat surprising to me because the use of the enclosed LI-7200 should prevent frequent gaps (by the way: is the intake tube of the LI-7200 heated or not?).
Regarding model validation: Why only soil respiration and soil moisture were compared with in situ measurements? It would be nice to see the model performance also for NEE and ET.
The conclusions include the finding that the REddyProc gap-filling showed slightly higher correlation than process-based gap-filling approaches like LandscapeDNDC. So, the statement “the current analysis we could … demonstrate the suitability of process-based models for this task” should maybe complemented by the statement that the model doesn’t perform better than the usual gap-filling, im my point of view.
Some spelling mistakes in the detailed remarks below.
Detailed remarks:
L55: regarding footprint calculation references you should mention Kljun et al. already here:
Kljun, N., P. Calanca, M. W. Rotach, and H. P. Schmid. 2004. A simple parameterisation for flux footprint predictions. Boundary-Layer Meteorology, 112: 503-523.
L94: 7.8782°E instead of 7.8782°W
L180: European instead of Europe
Fig. 4: measured soil moisture instead of measured flux
Fig. 7:Regarding hourly resolution of NEE I would prefer the unit µmol m-2 s-1 instead of kgC ha-1 hr-1
Tab. 3: mm hr-1 instead of kgC ha-1 hr-1
Fig. S8: during the non-growing season instead of during the growing season