Modeling of CO₂ Efflux from Forest and Grassland Soils Depending on Weather Conditions

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

The results are very important for predicting CO₂ emissions from soil or soil respiration in forests and grasslands. However, there are some questions and comments as follows:

1.      Introduction: The introduction presents limited literature on the "state of the art" in studies related to forests and grasslands in your research areas. The authors have cited several studies that were conducted in areas completely different from where your work was carried out. There is a significant body of research from temperate regions that has not been considered in this section. Additionally, please explain the relationship or correlation between precipitation and soil moisture, as soil moisture is a very important environmental factor controlling soil respiration but the manuscript show the precipitation.

2.      Results: The correlation results between mean monthly temperature (Tsoil, Tair) and monthly precipitation for normal, wet, and dry years (Table 2) indicate no correlation between precipitation and Tsoil or Tair during the dry season. I believe this could affect the model during the dry period. Please explain how this issue is addressed in the analysis of your model.

3.      The symbols in Figure 4 do not match the representation in the picture.

4.      Discussion: Please explain and compare the main factors controlling soil respiration at both sites, as this is crucial for the analysis of your model.

5.      Do you have a method to check the uncertainty of your results after obtaining them from the model?

Comments for author File: Comments.pdf

Author Response

The results are very important for predicting CO₂ emissions from soil or soil respiration in forests and grasslands. However, there are some questions and comments as follows:

Dear reviewer, we would like to thank you for taking the time to review our manuscript and your constructive comments and suggestions. Please find the detailed responses below and the corresponding highlighted changes in the re-submitted manuscript.

Comments 1: Introduction: The introduction presents limited literature on the "state of the art" in studies related to forests and grasslands in your research areas. The authors have cited several studies that were conducted in areas completely different from where your work was carried out. There is a significant body of research from temperate regions that has not been considered in this section. Additionally, please explain the relationship or correlation between precipitation and soil moisture, as soil moisture is a very important environmental factor controlling soil respiration but the manuscript show the precipitation.

Response 1: We added the related research investigating modeling of soil respiration in the temperate region. This allowed us to highlight the difference between our ensemble-based approach and the standard approach focusing mainly on a single model scheme.

With regard to the precipitation ~ soil moisture relationship we agree that it is the soil moisture that should directly affect the CO2 emissions. However because the direct measurement of soil moisture are not always available, it is customary to use precipitation as its proxy on the monthly scale. In the dynamic models such as RothC or Century the moisture effect is acquired from the balance between the precipitation and potential evaporation. We added the following text to the introduction:

LINES 81-85: “On the other hand, in the dynamic models such as RothC or Century the soil moisture effect is acquired from the balance between the precipitation and potential evaporation. Due to the higher soil-moisture ~ precipitation correlations on the longer than daytime scales [48] it is customary to use precipitation as a proxy to moisture on the monthly scale [20,49,50]..”

Comments 2: Results: The correlation results between mean monthly temperature (Tsoil, Tair) and monthly precipitation for normal, wet, and dry years (Table 2) indicate no correlation between precipitation and Tsoil or Tair during the dry season. I believe this could affect the model during the dry period. Please explain how this issue is addressed in the analysis of your model.

Response 2: Thank you for requesting this explanation.  To address it, we added the new Discussion section (LINES 437-489: 4.1. Weather conditions affect the modeling quality), which describes the issue and the respective model performances. We also applied the ANOVA analysis to check the significance of the weather-specified differences in the obtained R² and RMSE for both forest and grassland model sets.

Comments 3: The symbols in Figure 4 do not match the representation in the picture.

Response 3: Thank you for noticing the inconsistencies. The correct unites in Precipitation is mm, now consistent. We also added the labels a and b to the figure (Now Figure 3).

Comments 4: Discussion: Please explain and compare the main factors controlling soil respiration at both sites, as this is crucial for the analysis of your model.

Response 4: We have already listed the main factors, controlling soil respiration, in the following paragraph of the Introduction:

LINES 63-71: “One way to estimate CO2 efflux is the modeling which is an important alternative to direct chamber measurements [17,18]. At the same time, it's possible to consider the simple empirical modeling use temperature (T) and precipitation (P) [15,19] and complex dynamic models based on simulation of processes occurring in soil [5,19-21].

Both CO2 efflux measurements and applications of empirical CO2 efflux models highlight a strong dependency of CO2 efflux on the temperature [8,19,22-25], moisture [19,26-27], precipitation [18,19, 28-30] and throughfall [31-32], change of ground water level in soil [33], allocation of above-ground biomass [34-35], quality and amount of soil organic carbon (SOC) stored in soils [35,36-40], soil properties [11,40,41], and microbial community composition [42].”

And we also added the following paragraph to the Discussion section (LINES 473-482): 4.1. Weather conditions affect the modeling quality:

“The results obtained in this research confirm that the dependence of CO2 emission on temperature is dominant [8,20,37], and the dependence on humidity/precipitation is second after it [20,28-30]. The influence of vegetation as a factor regulating the hydrothermal conditions of soils for our sites is revealed in the differences between forest and grassland only in some years (see Table S3 in the Supplementary materials). For both ecosystems, the best modeling results in both R² and RMSE are shown in wet years and the worst in dry years, while in years with normal moisture conditions differ by vegetation type. For grassland site, the model results in normal conditions are similar as in dry years, whereas for forest site ones are somewhat worse than in wet years and better than in dry years.”

Comments 5: Do you have a method to check the uncertainty of your results after obtaining them from the model?

Response 5: Yes, the uncertainty is being checked with the RMSE, which is the square root of the variance between the measured and modeled values. The RMSE are presented in the Supplementary materials Tables S2, S3, and S4.

Reviewer 2 Report

Comments and Suggestions for Authors

Manuscript Number: soil systems-3371551Titled: " Modeling of CO₂ efflux from forest and grassland soils depending on weather conditions" aims at the site-specific differences and their effects on the empirical CO₂ efflux modeling based on the experimental data of the 25-year field observations for the forest and grassland ecosystems. They found that the SCLISS modeled Tlit and Mlit provide a good alternative to the direct atmospheric measurements and can be a good initial temperature and moisture data for the CO₂ efflux modeling, when the direct Tsoil and moisture observations are not available on site.

Although based on 25 years of field experiment data, I don’t think using the moisture content and temperature of litter or sod instead of soil temperature and moisture is a good alternative method. There is no significant difference in difficulty between obtaining data of soil temperature and moisture content and data of litter moisture content and temperature. This makes this research work lack of significance.

Furthermore, there has been considerable work estimating soil respiration. In this study, for example, in line 369-376, they found that for both forest and grassland ecosystems, the parameterization with litter or sod horizon temperature and moisture provides better results than the parameterization with deeper-horizon soil temperature and moisture. It sounds obvious, that it feels lacking in Originality.

 

Given these problems in the study, I do not recommend its publication in Soil Systems.

Author Response

Comments 1:  Manuscript Number: soil systems-3371551 Titled: " Modeling of CO₂ efflux from forest and grassland soils depending on weather conditions" aims at the site-specific differences and their effects on the empirical CO₂ efflux modeling based on the experimental data of the 25-year field observations for the forest and grassland ecosystems. They found that the SCLISS modeled Tlit and Mlit provide a good alternative to the direct atmospheric measurements and can be a good initial temperature and moisture data for the CO₂ efflux modeling, when the direct Tsoil and moisture observations are not available on site.

Although based on 25 years of field experiment data, I don’t think using the moisture content and temperature of litter or sod instead of soil temperature and moisture is a good alternative method. There is no significant difference in difficulty between obtaining data of soil temperature and moisture content and data of litter moisture content and temperature. This makes this research work lack of significance.

Furthermore, there has been considerable work estimating soil respiration. In this study, for example, in line 369-376, they found that for both forest and grassland ecosystems, the parameterization with litter or sod horizon temperature and moisture provides better results than the parameterization with deeper-horizon soil temperature and moisture. It sounds obvious, that it feels lacking in Originality.

Response 1: Dear reviewer, Thank you very much for taking the time to review our manuscript and your concerns. We would like to notify that we look at them the carefully. Here we present the set of now improved statements that justify the significance and originality of the presented research.

Our work is bases in the previous research, of course, and one of the novelties is the ensemble approach and the technique to correctly estimate the R0 coefficients of the empirical models involved into the ensembles.

We pointed out in the introduction: (LINES 61-64) “Due to the high spatial heterogeneity of CO₂ effluxes and several practical limitations it is impossible to allocate sufficient resources to cover large areas by field measurements [5]. One way to estimate CO₂ efflux is the modeling which is an important alternative to direct chamber measurements [16,17].” We also acknowledged that the direct data of the soil temperature and moisture is not always available and the meteorological data is being used instead. We suggest overcoming this issue (LINES 75-81) “If such measured data sources are not available on a site, they can be modeled using the meteorological conditions and the budget model schemes such as SCLISS (Soil CLImate Statistical Simulator) [41]. The soil temperature and moisture modeling could be applied for the selection of the appropriate soil depth and the respective horizon that could be important to identify the CO₂ sources of CO₂ efflux and the proper CO₂ efflux modeling [23,42].” 

These roses an applicability questions: (LINES 86-88) “Researchers highlight an arisen issue of an appropriateness of models to be used 85 only within their area of applicability [47] and on the correct spatial and temporal scales 86 [48].”   

We also highlight the novelty of our approach: (LINES 88-94) “Empirical model research mostly focuses either on some specific model or on the disconnected sets of models [25, 33, 36-37]. The novelty of the presented approach to the empirical modeling is to apply the ensembles of CO₂ efflux models to investigate their applicability to the weather specific conditions depending on soil and ecosystem types. Our previous research showed that the ensemble approach allows for the direct model intercomparison to investigate their conditional performances in attempt to find the better models in ensembles [43,44,53].” (LINES 102-104) “For this study, we apply the ensemble of the empirical CO₂ efflux models to the forest and grassland ecosystems on the similar Entic Podzol located in the same temperate-continental climate zone”

Then this goes with our hypotheses: (LINES 106-111) “We hypothesize that the level of humidity during the year (wet, normal, or dry) will affect the accuracy of modeling and quantification of parameters of models applied. We also propose that the interannual variability in respirations of trees and grasses due to the differences in temperature, moisture, and root allocation can highlight a role of the soil horizons involved in the CO₂ efflux process [42] and impact the simulation results as well.”

So our application of SCLISS model looks appropriate: (LINES 111-114) “Using the modeling of the litter (for forest) or sod (for grassland) horizon and soil temperature (Tlit_m, Tsoil_m) and moisture (Mlit_m, Msoil_m) with the SCLISS model we will evaluate which of the modeled values provide an appropriate substitution of the measured soil temperatures and precipitations.”

So we think that this work provides important information on the usage of the weather specific models and about ways how to use the modeled soil temperature and moisture data for the empirical modeling of the CO₂ efflux.

Reviewer 3 Report

Comments and Suggestions for Authors

The authors modeled CO₂ efflux based on key environmental variables, including soil (0-5 cm depth) and air temperature and amount of precipitation. They analysed experimental data from 25 years of field observations, and generated models for normal, wet, and dry years. These models were parameterized for forest and grassland ecosystems located on similar Entic Podzols (Arenic) within the same bioclimatic coniferous-deciduous forest zone.

To achieve this, they applied weighted non-linear regression to estimate CO₂ efflux for the different climatic conditions. Model resolution was enhanced by adjusting the slope and intercept in the linear comparison between measured and modeled data, specifically through changes in R₀ — the CO₂ efflux at a soil temperature (Tsoil) of 0 °C. The statistical methods appear to have been carefully applied to this large dataset. However, the results section was challenging to follow, making it difficult to interpret the findings.

I suggest to avoid the complex scatterplots (Figure 5 and 6) and focusing only on the regression outputs. Additionally, I recommend shortening the results section to highlight only the most important findings.

Comments on the Quality of English Language

English is not bad, but it could be improved. Sometimes the authors use the present tense instead of the past tense, which could affect the consistency of the text.

Author Response

The authors modeled CO₂ efflux based on key environmental variables, including soil (0-5 cm depth) and air temperature and amount of precipitation. They analysed experimental data from 25 years of field observations, and generated models for normal, wet, and dry years. These models were parameterized for forest and grassland ecosystems located on similar Entic Podzols (Arenic) within the same bioclimatic coniferous-deciduous forest zone.

To achieve this, they applied weighted non-linear regression to estimate CO₂ efflux for the different climatic conditions. Model resolution was enhanced by adjusting the slope and intercept in the linear comparison between measured and modeled data, specifically through changes in R₀ — the CO₂ efflux at a soil temperature (Tsoil) of 0 °C. The statistical methods appear to have been carefully applied to this large dataset. However, the results section was challenging to follow, making it difficult to interpret the findings.

Dear Reviewer, we would like to thank you for the throughout attention and constructive comments. We tried to address them all and added a subsection to cover raised issues with the hypothesis.  Please find the detailed responses below and the corresponding highlighted changes in the re-submitted manuscript

Comments 1: I suggest to avoid the complex scatterplots (Figure 5 and 6) and focusing only on the regression outputs.

Response 1: Thank you for the suggestion. We tried to simplify these scatter plots (Now Figures 4,5) to focusing on the regression outputs and not on the initial data points or formulas. Now they look more consistent and easier to understand.

 

Comments 2: Additionally, I recommend shortening the results section to highlight only the most important findings.

Response 2: Thank you for the recommendation. To put better focus on the main results, we moved the first sub-section from the Results to the Discussion (LINES 437-489: 4.1. Weather conditions affect the modeling quality). There we added an extra discussion on the effects of the correlations to the model performances.

Comments 3: English is not bad, but it could be improved. Sometimes the authors use the present tense instead of the past tense, which could affect the consistency of the text.

Response 3: We looked through the manuscript to improve quality of writing and addressed this issue.

Reviewer 4 Report

Comments and Suggestions for Authors

 

The authors need to think carefully about what is novel from their research.

I am not exactly clear what is novel.

Much work has been done on temperature and soil water content on CO2 efflux already.

 

Hypothesis

“We hypothesize that the level of humidity/aridity during the year (wet, 95 normal, or dry) will affect the accuracy of modeling and quantification of parameters of 96 models applied.”

This should be rewritten in a more specific manner. Otherwise it is difficult to accept/ reject the hypothesis.

What is being tested exactly? Humidity or aridity?

Is the accuracy or the quantification of parameters?

 

Line 97 is also not clear

“We also propose that the interannual variability in activities…”

What is meant by activities?

 

Line 98/99

“C allocation by roots can highlight a role….”

What does this mean?

 

I think it is OK to have a hypothesis, but it must be clear.

Also it could help to establish specific objectives of what the study is trying to achieve?

This could follow the hypotheses.

It sounds like you need to have more than one hypothesis.

 

Section 2.3.

Why is this section needed?

The model is already published?

Why do you need to describe these detaiuls?

This seems like repeating information.

 

Table 2

P values need to be written clearly and using conventional numbers.

 

Fig. 2

This is hard to follow.

What is a? b? c? d?

You must label your figs properly.

Also what are the dotted lines? What they represent?

Very noisy figs and hard to follow.

 

Fig. 4

You now have prec. In units of cm.

Why?

In fig. 2 you had units of mm?

Please be consistent with units.

Also where is the (a) and (b) on fig. 4?

What soil depth is fig. 4 for?

No indication of the soil depth???

 

Table 3

Spread?

What is this? Please use correct terms?

 

Line 316 and 318

You have given low correlation values?

These must be explained.

Given all the data you collected why is the r value so low??

 

Line 326

Water?

What is this?

Soil water content? Please be clear.

 

Fig. 7

This is very hard to read and must be changed.

Too many points and the equations are hard to read.

Also where are the statistics for these figs.

e.g. R squared and P value

 

only show the important results.

Put other results in the appendix.

 

 

Line 425

Why is the soil exposed to sunlight?

This makes me wonder if the experiment/ measurements were flawed somehow.

This makes me questions all of the results.

 

Line 464

If you rewrite your hypothesis you will be able to write the discussion in an easier and clearer way.

Writing

“The results obtained supported partly our first hypothesis that……”

Is not helpful to anyone !

Author Response

Comments 1: The authors need to think carefully about what is novel from their research.  I am not exactly clear what is novel. Much work has been done on temperature and soil water content on CO2 efflux already.

Response 1: Dear reviewer, we would like to express our thanks for the throughout and constructive review. It helped us to make the hypotheses clearer and improve the Discussion section (LINES 437-489: 4.1. Weather conditions affect the modeling quality) to address them in full. Please find the detailed responses below and the corresponding highlighted changes in the re-submitted manuscript.

 Comments 2: Hypothesis “We hypothesize that the level of humidity/aridity during the year (wet, 95 normal, or dry) will affect the accuracy of modeling and quantification of parameters of 96 models applied.” This should be rewritten in a more specific manner. Otherwise it is difficult to accept/ reject the hypothesis. What is being tested exactly? Humidity or aridity? Is the accuracy or the quantification of parameters?

Response 2: Thank you for the clarification question! We left the “level of humidity” for different years to make it more coherent.  We focused on the R0 value change after an application of our procedure – this is quantification. The accuracy would be for the parameter uncertainty estimations, which is not our current task and could be addressed in further research. We also added the subsection in the Discussion to fully address the hypothesis.

 Comments 3:  Line 97 is also not clear “We also propose that the interannual variability in activities…” What is meant by activities?

Line 98/99 “C allocation by roots can highlight a role….” What does this mean?

Response 3: To clarify these issues we rephrased the sentence:

LINES 108-111: “We also propose that the interannual variability in respirations of trees and grasses due to the differences in temperature, moisture, and root allocation can highlight a role of the soil horizons involved in the CO₂ efflux process [42] and impact the simulation results as well.”

 Comments 4: I think it is OK to have a hypothesis, but it must be clear. Also it could help to establish specific objectives of what the study is trying to achieve? This could follow the hypotheses. It sounds like you need to have more than one hypothesis.

Response 4: Yes there are actually two hypotheses stated together:

LINES 106-111: “We hypothesize that the level of humidity during the year (wet, normal, or dry) will affect the accuracy of modeling and quantification of parameters of models applied. We also propose that the interannual variability in respirations of trees and grasses due to the differences in temperature, moisture, and root allocation can highlight a role of the soil horizons involved in the CO₂ efflux process [42] and impact the simulation results as well.”

 

Comments 5: Section 2.3. Why is this section needed? The model is already published? Why do you need to describe these detaiuls? This seems like repeating information.

Response 5: Even though the models have already been published, we think this information is important for understanding what was used in the ensemble of investigated models. We would like to keep this section intact.

Comments 6: Table 2 P values need to be written clearly and using conventional numbers

Response 6: Thank you, this is fixed. (Now Table 3 in Discussion).

Comments 7: Fig. 2 This is hard to follow. What is a? b? c? d? You must label your figs properly. Also what are the dotted lines? What they represent? Very noisy figs and hard to follow.

Response7: Thank you for noticing these inconsistencies. We added the proper labels and highlighted the lines. We also added the respective explanation into the figure caption (Now Figure 10 in Discussion).

Comments 8: Fig. 4 You now have prec. In units of cm. Why? In fig. 2 you had units of mm? Please be consistent with units. Also where is the (a) and (b) on fig. 4?

Response 8: Thank you for noticing the inconsistencies. The correct unites in Precipitation is [mm], now consistent. We also added the labels a and b to the figure.

Comments 9: What soil depth is fig. 4 for? No indication of the soil depth???

Response9: Thank you for noticing the inconsistencies.  We added for the modeled soil layer depth: (LINES 237-238) “where Msoil_m is defined as the average moisture in the soil layer up to 1 m deep.” (Now Figure 3)

Comments 10:  Table 3 Spread? What is this? Please use correct terms?

Response 10: Thank you for pointing to the incorrect term usage. To correct, we changed “Spread” term to “Range”, which is the difference between the extreme values. (Now Table 2)

Comments 11: Line 316 and 318 You have given low correlation values? These must be explained. Given all the data you collected why is the r value so low??

Response 11: This related comparison is conducted for the 0 < Tsoil < 1 ^oC temperature range only (see Table S1 in Supplementary materials). There cold season measurements are characterized by the low respiration values, reduced at low temperatures, and their high variability due to large amount of obstacles in the CO2 pathways such as snow cover. These all make the presented correlations low.

Comments 12: Line 326 Water?  What is this? Soil water content? Please be clear.

Response 12: Yes, following your suggestion, we corrected this to “Soil water content”

Comments 13: Fig. 7 This is very hard to read and must be changed. Too many points and the equations are hard to read. Also where are the statistics for these figs. e.g. R squared and P value

Response 13: Thank you for the suggestion. Yes we agree that the figures are overcrowded.  We cleared it and replaced the Figures 6 and 7 with the simple ones focusing on the respective regressions. The respective R2 and RMSE values are in the Table S2 of the Supplementary materials.

Comments 14: Line 425 Why is the soil exposed to sunlight? This makes me wonder if the experiment/ measurements were flawed somehow. This makes me questions all of the results.

Response 14: Thank you. This is the lack of the proper explanation on our side. The soil is below the grass cover, but it is less protected than if it was under the forest canopy. We changed the statement as following:

LINES 609-611: “For the grassland ecosystem, however, the model results for normal conditions worsen toward dry conditions, signaling water demand and temperature differences when the soil is less protected by the overlaying vegetation cover.”

Comments 15: Line 464 If you rewrite your hypothesis you will be able to write the discussion in an easier and clearer way. Writing “The results obtained supported partly our first hypothesis that……” Is not helpful to anyone !

Response 15: Thank you for pointing to this inconsistency. We kept the hypothesis intact and added the subsection to the Discussion to fully address the hypothesis statement. Now there are two following subsections in the Discussion that address the hypothesis:

LINES 437-489: 4.1. Weather conditions affect the modeling quality;

LINES 492-518: 4.2. Distributions of the measured CO₂ efflux relevant for the weather-specific R₀ estimations