Evaluating Multi-Angle Photochemical Reflectance Index and Solar-Induced Fluorescence for the Estimation of Gross Primary Production in Maize

Round 1

Reviewer 1 Report

In this study the authors compare the suitability of two remote-sensing methods (using a photochemical reflectance index (PRI)- based light use efficiency (LUE) model or a sun-induced fluorescence (SIF)-based linear model) to estimate gross primary productivity (GPP) (measured by Eddy Covariance flux) of a maize crop and their sensitivity to environmental factors. They found that PRI was superior in detecting diurnal changes in LUE and GPP, but that SIF was more sensitive to seasonal changes. Using a multi-angle spectra system improved the prediction of canopy LUE in comparison to a single-angle measurement. Relative humidity was the most important environmental factor affecting PRI, while SIF was more affected by PAR.

The study seems to be sound and is relevant as better methods to estimate GPP are important for accurate carbon accounting and measuring the effects of climate change on crops.

General comments:

The objectives of this study are relevant and the findings interesting. The manuscript is well written with only minor edits required (see below).

 

Detailed comments:

Line 63: “is able to be” > “can be”

Line 74: “perplexing impact of canopy structure at sun-view” > “on sun-view”

Line 94: “It’s found” > “It has been found”…

Line 95: “but how the correlation” > “but how the correlation between SIF and GPP responds to environmental variables has rarely been studied…”

Line 102: “the focal goals are” > “the objectives of our study are”

Line 110: “all measurement campaign” > “all measurement campaigns”

Line 118: “from sowing in the mid-June” > “from sowing in mid-June”

Line 199: “Fertilization occurred” > “The crop was fertilized” Can you give some details on what you fertilized it with.

Line 129: Please spell out EC system > Eddy Covariance

Line 140: “Where FPAR were” > “was”

Line 148: Please spell out T_a and RH.

Line 150: >”Q varies with real-time cloudiness of the sky, with a less cloudy sky having greater Q.

Line 155: “And the implement…” > “LAI measurements were taken during twilight and overcast conditions”…

Line 162: “The system is consisted of” > “the system consisted of…”

Line 203: Do you mean: “Days with more than 50% of missing data were excluded from the analysis.”?

Line 221: “The variation of” > “the variations of”

Line 234: “Decreased from the sunrise” > “Decreased from sunrise”

Fig. 3: Make sure the x-axis label is fully visible. At the moment photon is cut off.

Table 1: Can you give the units of GPP in the table please.

Line 296: “Was the secondly most important” > “was the second most important”

Line 345 and 346: “conifer” > “conifer forest”

Line 346: “Simple approaches” > “simple approach”

Line 371: “was the most limited” > “was the most limiting factor for the PRI-based LUE model”

Line 375: > “of the PRI-based LUE model”

Line 381: “then” > “than”

Line 387: >”when vegetation was exposed”

Line 404: >”mathematical”

Line 410: >”to more accurately…”

Line 415: “the PRI-base LUE model were” >”was”

Line 424: >”the most important factor”

 

Author Response

Dear reviewer,

Thanks for your kind comments. In this revised version, we have followed your suggestions very carefully and revised them one by one. In this response letter, we showed our detailed responses to your comments. Please be kind to note that our responses are in italics while our revisions are shown in bold letters.

The response letter was also uploaded as a WORD file. Please see the attachment if you want to download it.

Response:

Thank you very much for your careful reading of our manuscript. We revised all grammar errors and unclear expressions as your suggestions:

• Line 69: “is able to be” > “can be”
• Line 79: “perplexing impact of canopy structure at sun-view” > “on sun-view”
• Line 104: “It’s found” > “It has been found”…
• Line 106: “but how the correlation” > “but how the correlation between SIF and GPP responds to environmental variables has rarely been studied…”
• Line 114: “the focal goals are” > “the objectives of our study are”
• Line 121: “all measurement campaign” > “all measurement campaigns”
• Line 137: “Fertilization occurred” > “The crop was fertilized”
• Line 159: “Where FPAR were” > “was”
• Line 170: >”Q varies with real-time cloudiness of the sky, with a less cloudy sky having greater Q.
• Line 175: “And the implement…” > “LAI measurements were taken during twilight and overcast conditions”…
• Line 183: “The system is consisted of” > “the system consisted of…”
• Line 250: “The variation of” > “the variations of”
• Line 265: “Decreased from the sunrise” > “Decreased from sunrise”
• Line 329: “Was the secondly most important” > “was the second most important”
• Line 384 and 385: “conifer” > “coniferous forest”
• Line 388: “Simple approaches” > “simple approach”
• Line 417: “was the most limited” > “was the most limiting factor for the PRI-based LUE model”
• Line 421: > “of the PRI-based LUE model”
• Line 437: >”when vegetation was exposed”
• Line 455: >”mathematical”
• Line 461: >”for more accurate estimation of GPP”
• Line 466: “the PRI-base LUE model were” >”was”
• Line 475: >”the most important factor”

 

Line 199: Can you give some details on what you fertilized it with.

Response: The detailed fertilizations were added in line 136 to 138.

• The crop was fertilized (40.5 kg N ha^-1 and 103.5 kg P ha^-1) at the time of sowing and was topdressed (225 kg N ha^-1, 37.5 kg P ha^-1, and 37.5 kg K ha^-1) prior to the elongation stage.

 

Line 203: Do you mean: “Days with more than 50% of missing data were excluded from the analysis.”?

Line 381: “then” > “than”

Response: These two confusing sentences were rewritten:

• Line 226: the sentence “Our analysis only included days which available observations from 6:00 to 18:00 occupied more than 50%.” was revised to “Days with less than 50% of available data were excluded from the analysis.”
• Line 428: the sentence “…because the former had a more direct effect on opening and closing of the leaf stomata, then the photosynthetic light use efficiency, not on the reflectance of the leaf and canopy.” Was revised to “…because the former had a direct effect on opening and closing of the leaf stomata and then the photosynthesis, not the reflectance of the leaf and canopy.”

 

Line 129: Please spell out EC system > Eddy Covariance

Line 148: Please spell out Ta and RH.

Response: For the abbreviations of some terms, such as EC (line 147), Ta and RH (line 167), we defined them when they first appear (EC, line 58; Ta, line 95; RH, line 38). And we also spelled out them where you mentioned in this revised version.

 

Fig. 3: Make sure the x-axis label is fully visible. At the moment photon is cut off.

Table 1: Can you give the units of GPP in the table please.

Response: The x-axis label in Figure 3 was revised to be fully visible. And the unit of GPP (μmol CO² m^-2 s^-1) was added in Table 1.

 

Author Response File: Author Response.docx

Reviewer 2 Report

1) Lines 43-44: You used as keywords the terms from the title of your paper. It would be better if you replaced these keywords with the terms that have not been mentioned in the title. This will inmprove further indexing of your paper in the databases.

2) Line 114: It seems like you missed an article before "part".

3) Lines 120-124: It is advisable to use common international names for the phenological stages, as in Nafziger, E. 2013. Corn. Illinois Agronomy Handbook.
Crop Science Extension and Outreach, Urbana
IL, USA.

4) Line 202: What do you mean under "negative" data, and why they were excluded? 

5) Line 206: What values of RSQ (determination coefficient) did you calculate? Did you evaluate Adjusted and Predicted RSQ values, or just "bare RSQ"? 

6) There are no links to the methodology of calculations for statistical criteria in the 2.5 section of the paper, which are necessary to have an understanding of what computation techniques the authors used in their statistical analyses.

7) It is also interesting and valuable information on the volume of data used in the statistical processing, because the study was conducted for a single year, and it is not clear whether sufficient amount of the data pairs was to make reliable statistical evaluation through regression analysis.

8) Line 235: I consider both values of RSQ (0.49 and 0.41) quite low to talk about the strength of connection between the studied indices. The problem might be in the application of a linear model.

9) Table 1: I think that all the values of RSQ (0.44-0.50) are out of one range of determination rate, and it is unfair to talk that some pairs had stronger or weaker connections. If you do not agree, please, refer to the studies in statistics, which will prove your point of view (I mean that RSQ of 0.50 can reliably tell us about higher strength of connection between the features than the RSQ of 0.44 at the same p level).

10) Line 264-265: The discrepancy in the number of data sets makes it impossible to conduct fair comparison between SIF and PRI.

11) Line 317: "The higher the soil temperature, the higher the R2PRI". This is an interesting fact. Do you have an explanation for this phenomenon? Please, provide it, if any.

Author Response

Dear reviewer,

We greatly appreciate your efforts in helping us to improve the paper. Your suggestions truly help us to improve our paper in a great deal. We have addressed your comments as best as we can and have revised our manuscript substantially.

In this revised version, we have followed your comments very carefully and made revisions as much as we can. The remainder of the letter presents our point-by-point response. For your kind information, we repeat your original comments and insert our responses beneath each of your comments.

The response letter was also uploaded as a WORD file. Please see the attachment if you want to download it.

1) Lines 43-44: You used as keywords the terms from the title of your paper. It would be better if you replaced these keywords with the terms that have not been mentioned in the title. This will improve further indexing of your paper in the databases.

Response: Thanks for your kind suggestion. We replace the “gross primary production, photochemical reflectance index, solar-induced fluorescence, and multi-angle observations” that were mentioned in the title with “vegetation photosynthesis, sun-view geometry, and temporal dynamics”.

 

2) Line 114: It seems like you missed an article before "part".

Response: The article “a” was added before “part”. (line 125 on page 3)

 

3) Lines 120-124: It is advisable to use common international names for the phenological stages, as in Nafziger, E. 2013. Corn. Illinois Agronomy Handbook. Crop Science Extension and Outreach, Urbana IL, USA.

Response: Thanks for your advice. We revised the description about the phenological stages of maize as in Nafzigei et al. (2013):

“…The entire growing season of summer maize continued for four months, from sowing in early June [Day of Year (DOY), 155] to harvest in late September (DOY, 267). The phenology of maize is usually divided into two main growth stages: the vegetative (V) stages (emergence to DOY 207) and the ripening (R) stages (DOY 208 to harvest) [45]. The V stages included VE stage (emergence) on DOY 160, V5 stage (5th leaf) on DOY 170, elongation stage on DOY 183, and VT stage (tasseling) on DOY 206; while the R stages experienced R1 stage (silking) on DOY 208, R6 stage (physiological maturity) on DOY 256 till to harvest on DOY 267.” (line 128 to 136 on page 3)

 

4) Line 202: What do you mean under "negative" data, and why they were excluded?

Response: “Negative” data means that the value of the data is less than zero. GPP, SIF and PAR may have negative values due to some possible observation errors. But theoretically, they will never go below zero. Even in dark nights without light, they will only be zero. Thus, we considered their negative values to be erroneous observations and excluded them.

 

5) Line 206: What values of RSQ (determination coefficient) did you calculate? Did you evaluate Adjusted and Predicted RSQ values, or just "bare RSQ"?

Response: In this paper, we used the original coefficient of determination, i.e. bare RSQ, instead of Adjusted RSQ. The Adjusted RSQ is an adjustment for the bare RSQ that takes into account the number of variables in a data set. It’s usually used to include a more appropriate number of variables, thwarting our temptation to keep on adding variables to the data set. But in our study, the number of predicted variables was fixed and limited. Thus, we used the bare RSQ.

 

6) There are no links to the methodology of calculations for statistical criteria in the 2.5 section of the paper, which are necessary to have an understanding of what computation techniques the authors used in their statistical analyses.

Response: Thank you for the comment. We added more details about how we did the statistical analysis in the section 2.5. Firstly, we moved the description of the data used for two models in three time scales shown in Table 1 to this section as follows:

• “…Both models were applied at “daily mean”, “30min”, and “day-by-day” scales. At the “daily mean” scale, we used the daily means of half-hourly values of GPP, APAR, PRI, and SIF from 6:00 to 18:00 of all individual days during the entire growing season in 2018. At the “30 min” scale, we used all the half-hourly values from 6:00 to 18:00 during the growing season. At “day-to-day” scale, we applied two models for each day during the growing season using the half-hourly values from 6:00 to 18:00.” (line 230 to 235 on page 6)

Secondly, we added the details about the data used for random forest analysis.

• “…We calculated the daily mean values of the six above environmental variables and used them as the predictor variables.” (line 238 to 239 on page 6)

 

7) It is also interesting and valuable information on the volume of data used in the statistical processing, because the study was conducted for a single year, and it is not clear whether sufficient amount of the data pairs was to make reliable statistical evaluation through regression analysis.

Response: Thanks for your comment. The volume of data used for regression analysis shown in Table 1 was different from each other. For the “daily mean” scale, the volume of data was tens; for the “30 min” scale, the volume of data was thousands; for the “day-by-day” scale, the volumes of data differed each day, but all ranged from 15 to 24. We added a detailed description of the data in section 2.5.

 

8) Line 235: I consider both values of RSQ (0.49 and 0.41) quite low to talk about the strength of connection between the studied indices. The problem might be in the application of a linear model.

Response: We also considered whether the linear model was suitable or not, and tried several other models such as exponential and power functions. But they all didn’t perform better than a linear model. And we just chose one random day to show the advantage of multi-angle PRI. Besides, in Zhang et al. (2015), most of the correlation coefficients (R) of half-hourly PRI with LUE on individual days were below 0.6.

• Zhang, Q.; Ju, W.; Chen, J.M.; Wang, H.; Yang, F.; Fan, W.; Huang, Q.; Zheng, T.; Feng, Y.; Zhou, Y., Ability of the photochemical reflectance index to track light use efficiency for a sub-tropical planted coniferous forest. Remote Sensing 2015. 7(12): 16938-16962.

 

9) Table 1: I think that all the values of RSQ (0.44-0.50) are out of one range of determination rate, and it is unfair to talk that some pairs had stronger or weaker connections. If you do not agree, please, refer to the studies in statistics, which will prove your point of view (I mean that RSQ of 0.50 can reliably tell us about higher strength of connection between the features than the RSQ of 0.44 at the same p level).

Response: Thank you for the comment. The coefficient of determination (R²), known as “goodness of fit”, is generally used to explain how much variability of one variable can be caused by its relationship to another. In our study, “0.44” and “0.5” meant that 44% and 50% of the dependent variable (GPP) is predicted by the independent variable (LUE_PRI×APAR and SIF, respectively). From this perspective, we thought the PRI-based LUE model performed better than the SIF-based linear model at the “daily mean” scale. Indeed, both 0.44 and 0.5 were not high enough to show that the two models could fully represent the changes in GPP. But some previous studies also showed that the R² value was not so high: the R² of SIF and GPP ranged from 0.28 (using daily values) to 0.46 (using half-hourly values) (Miao et al., 2018); the R² of PRI and LUE was 0.48 (using daily values) (Zhang et al., 2017) and 0.60 (using half-hourly values) (Ma et al., 2020). So we compared these R² values in this way. We are not sure if we fully understand your meaning, if not, we hope that you can make more suggestions in the next revision round for us to improve the manuscript.

• Miao, G.; Guan, K.; Yang, X.; Bernacchi, C.J.; Berry, J.A.; DeLucia, E.H.; Wu, J.; Moore, C.E.; Meacham, K.; Cai, Y., Sun‐induced chlorophyll fluorescence, photosynthesis, and light use efficiency of a soybean field from seasonally continuous measurements. Journal of Geophysical Research: Biogeosciences, 2018. 123(2): 610-623.
• Zhang, Q.; Chen, J.M.; Ju, W.; Wang, H.; Qiu, F.; Yang, F.; Fan, W.; Huang, Q.; Wang, Y.-p.; Feng, Y., Improving the ability of the photochemical reflectance index to track canopy light use efficiency through differentiating sunlit and shaded leaves. Remote Sensing of Environment, 2017. 194: 1-15.
• Ma, L.; Wang, S.; Chen, J.; Chen, B.; Zhang, L.; Ma, L.; Amir, M.,; Sun L.; Zhou, G.; Meng, Z., Relationship between light use efficiency and photochemical reflectance index corrected using a BRDF model at a subtropical mixed forest Remote Sensing, 2020. 12: 550.

 

10) Line 264-265: The discrepancy in the number of data sets makes it impossible to conduct fair comparison between SIF and PRI.

Response: There were more data gaps of PRI because of the instrument failure than that of SIF. So it might result in incompletely consistent volumes of PRI and SIF data. But in each time scale (i.e. “daily mean”, “30 min”, and “day-by-day”), the amount of data used for PRI-based and SIF-based models was not much different as we mentioned in the response of comment 7.

 

11) Line 317: "The higher the soil temperature, the higher the R2PRI". This is an interesting fact. Do you have an explanation for this phenomenon? Please, provide it, if any.

Response: Thanks for your question. We thought this phenomenon might result from the synchronization of water and heat conditions. Besides, the high soil temperature occurred at the rapid growth stage of maize. The higher photosynthetic rate and linear energy distribution at this period might lead to high R²_PRI. We added the explanation in the section 4.3 from lines **-** on page 12 as follows:

“…Soil temperature also had a positive effect on the PRI-based LUE model (Figure 6a and 7f). The high T_s occurred at the late V stages (DOY 195 to 207) with synchronously high RH (Figure 5). Thus, the maize rapidly grown at this stage had a higher photosynthetic rate and linear energy distribution, leading to the good performance of the PRI-based LUE model.” (line 429 to 432 on page 12)

Author Response File: Author Response.docx

Reviewer 3 Report

The manuscript entitled “Evaluating multi-angle photochemical reflectance index and solar-induced fluorescence for better estimation of gross primary productivity in a maize field” by Chen et al. tried to reveal the diurnal and seasonal changes in the correlations between LUE and PRI or SIF in a maize filed. Developing the methodology to estimate GPP by remotely sensed indices is important topic for ecology and agriculture. Especially, the effect of sun and observation angles on PRI value is big issue. Unfortunately, however, this manuscript does not expand our understandings at all and gives us no implication. I would recommend the authors to reanalyze the original data and reconstruct the logics from the beginning. General comments are below;

1. Although R² values are shown in Figs. 4 and 5, it is absolute pointless. Temporal changes or variations in the slope and/or interception of the liner regression might be more meaningful.
2. The author mentioned that PRI and its interpretation of LUE are influenced by angular effects in the Introduction. This point could be one of the most important parts of this study. Thus, the authors should explain more detailed mechanism of the effects in the introduction and analyze their results according to that. In addition, the authors argue that daily mean PRI showed “better performance” than short time PRI, but R² of the liner regression between LUE and daily mean PRI is only 0.49 and it is not enough to say "better" than short-time PRI.
3. Since maize is C4 species, the photosynthetic system of this species is different from C3 species. I recommend the authors to analyze the data with considering the C4-specific characteristics.
4. Because the study site is a crop land, the seasonal variations in GPP, LUE and other data might be strongly affected by crop calendar. The authors should explain about the crop calendar in this site.
5. In general, I find the overall writing, both grammar and content, to be poor. I would suggest a significant edit be made to correct these errors.

Followings are specific comments;

1. “GPP” is generally stand for “gross primary production” not “gross primary productivity”.
2. The author showed air temperature and also soil temperature data. However, the soil temperature is not necessary because the short-term photosynthetic process, PRI and SIF can represent, are affected by leaf or air temperature, not soil temperature.
3. There is no explanation about upper panel of Fig. 1. I guess it is the aerial photograph but if so, the authors should explain it in the figure legend and indicate where the area, the study was done is.
4. It is difficult to distinguish between PRI_s and PRI_can in Fig. 3 (c).

Author Response

Dear reviewer,

Thank you very much for your constructive comments. Combining your comments with those from other reviewers, we have revised our manuscript substantially.

In this response letter, we insert our detailed responses to your comments underneath each of your original comments. Please be kind to note that our responses are in italics while our revisions are shown in bold letters.

The response letter was also uploaded as a WORD file. Please see the attachment if you want to download it.

 

The manuscript entitled “Evaluating multi-angle photochemical reflectance index and solar-induced fluorescence for better estimation of gross primary productivity in a maize field” by Chen et al. tried to reveal the diurnal and seasonal changes in the correlations between LUE and PRI or SIF in a maize filed. Developing the methodology to estimate GPP by remotely sensed indices is important topic for ecology and agriculture. Especially, the effect of sun and observation angles on PRI value is big issue. Unfortunately, however, this manuscript does not expand our understandings at all and gives us no implication. I would recommend the authors to reanalyze the original data and reconstruct the logics from the beginning.

 

General comments are below:

1. Although R² values are shown in Figs. 4 and 5, it is absolute pointless. Temporal changes or variations in the slope and/or interception of the liner regression might be more meaningful.

Response: Thank you for the comment. We agree with you that the slope and interception of the linear regression model are meaningful for the model construction and application. But, the coefficient of determination (R²), known as “goodness of fit”, is generally used to explain how much variability of one variable can be caused by its relationship to another. One of our major objectives for the study was to assess the PRI-based LUE model and the SIF-based linear model and explore the model preference for the estimation of GPP. Thus, the R² values shown in Figure 4 and Table 1 were used to illustrate whether the PRI-based LUE model and the SIF-based linear model could be a suitable model for the estimation of GPP. Besides, the relationships between PRI and LUE was the first step for the PRI-based LUE model, and then the LUE_PRI from PRI needs to be combined with APAR to obtain GPP. While the SIF-based linear model directly used the linear relationship between SIF and GPP. So the slope and interception are of limited help to the evaluation and comparison of the two models. But we will do more research about the temporal changes or variations in the slope and/or interception of the PRI-LUE or SIF-GPP linear regression in our future research.

 

2. The author mentioned that PRI and its interpretation of LUE are influenced by angular effects in the Introduction. This point could be one of the most important parts of this study. Thus, the authors should explain more detailed mechanism of the effects in the introduction and analyze their results according to that. In addition, the authors argue that daily mean PRI showed “better performance” than short time PRI, but R² of the liner regression between LUE and daily mean PRI is only 0.49 and it is not enough to say "better" than short-time PRI.

Response: Thank you for the suggestion. We have conducted a more thorough description and discussion about the angular effect in the introduction and discussion sections, respectively. Firstly, we added a more detailed description of the angular effects in the introduction as follows:

• “…Vegetation canopies are non-Lambertian and exhibit varying degrees of anisotropy, leading to their reflectance changes with viewed angles. Thereby, the PRI signal and its interpretation of LUE can be significantly influenced by angular effects, inclusive of leaf angle distribution and sun-view geometry [16, 22-24]. Multi-angle spectral observations can offer multi-angle information of the vegetation canopy and help to describe the variation in the observed fractions of the canopy with view angles, thereby improving the monitoring and inversion accuracy of crop biochemical parameters for the entire canopy [25, 26].” (line 80 to 86)

Secondly, we revised the section 4.1 and explained the mechanism of the angular effects in more detail:

• “…It was found in this study that the averages of the multi-angle observed PRI within short time better represented the condition of the entire canopy in estimating LUE than the single angle observed PRI (Figure 3), indicating that the utilization of multi-angle observations with tens of measurements evenly distributed at different angles could diminish the angular effects to some extent. Crop canopies are non-Lambertian characterized by bidirectional reflection. Their reflectance changes because of the varied fraction of sunlit/shaded elements of the canopy with the changing angles. The reflectance at 531 nm differed significantly over sunlit and shaded elements of the canopy. The canopy elements whose reflectance differed as a function of illumination intensity lead to the variation in the PRI signal viewed at varied angles [25]. The sunlit elements are more likely to be exposed to excess light levels, causing a conversion of violaxanthin to zeaxanthin and resulting in a lowering of PRI.” (line 365 to 376)

And you might misunderstand Figure 3. Figure 3 illustrated that the average of multi-angle PRI (PRI_can) could better track the variations of LUE than single-angle PRI (PRI_s). The R² between LUE and PRI increased from 0.41 to 0.49, when we replaced PRI_can with PRI_s.

 

3. Since maize is C4 species, the photosynthetic system of this species is different from C3 species. I recommend the authors to analyze the data with considering the C4-specific characteristics.

Response: Thanks for your suggestion. C4 plants are usually more suitable for hot, dry, and stressed environments than C3 plants. In the section 4.2, we combined the characteristics of C4 plants to supplement the interpretation of the performance of the two models as follows:

• “…Besides, maize belongs to C4 plants, which does not have photosynthetic “lunch break” phenomenon under high light at noon. Thus, the distribution of absorbed light energy mainly relies on non-photochemical quenching (NPQ), i.e. heat energy dissipation through the xanthophyll cycle, which may be more conducive to the expression of the diurnal LUE variations by PRI. In contrast, it has been found that the relationship between SIF and GPP appears to be linear under stress conditions precisely due to the reduced SIF values around midday [45].” (line 403 to 409)

 

4. Because the study site is a crop land, the seasonal variations in GPP, LUE and other data might be strongly affected by crop calendar. The authors should explain about the crop calendar in this site.

Response: We revised the introduction of the maize calendar based on our records of the 2018 maize growing season, including the phenological stages, fertilizations, and irrigation. The Detailed revisions are as follows:

• “…The entire growing season of summer maize continued for four months, from sowing in early June [Day of Year (DOY), 155] to harvest in late September (DOY, 267). The phenology of maize is usually split into two main growth stages: the vegetative (V) stages (emergence to DOY 207) and the ripening (R) stages (DOY 208 to harvest). The V stages included VE stage (emergence) on DOY 160, V5 stage (5th leaf) on DOY 170, elongation stage on DOY 183, and VT stage (tasseling) on DOY 206; while the R stages experienced R1 stage (silking) on DOY 208, R6 stage (physiological maturity) on DOY 256 till to harvest on DOY 267. The crop was fertilized (40.5 kg N ha^-1 and 103.5 kg P ha^-1) at the time of sowing and was topdressed (225 kg N ha^-1, 37.5 kg P ha^-1, and 37.5 kg K ha^-1) prior to the elongation stage. The soil was irrigated (45 mm) before the emergence of maize (DOY 158).” (line 128 to 139)

 

5. In general, I find the overall writing, both grammar and content, to be poor. I would suggest a significant edit be made to correct these errors.

Response: We’re sorry for our poor English writing. We have improved the writing of our manuscript as much as we can. We revised the grammar errors and unclear expressions firstly based on suggestions offered by one of the reviewers. And then we revised the manuscript comprehensively with the help of a native English speaker.

 

Followings are specific comments:

 

1. “GPP” is generally stand for “gross primary production” not “gross primary productivity”.

Response: Thanks for your suggestion. After searching the definitions of “gross primary productivity” and “gross primary production”, we figured out that “gross primary productivity” represented the rate of gross photosynthesis (P_g) while “gross primary production” represented the integral of P_g in a period. GPP actually stands for “gross primary production” (Gilmanov et al., 2003). Thus, we replaced “gross primary production” with “gross primary productivity” in full text.

• Gilmanov, T. G., S. B. Verma, P. L. Sims, T. P. Meyers, J. A. Bradford, G. G. Burba, and A. E. Suyker, Gross primary production and light response parameters of four Southern Plains ecosystems estimated using long-term CO₂-flux tower measurements, Global Biogeochem. Cycles, 2003, 17(2), 1071.

 

2. The author showed air temperature and also soil temperature data. However, the soil temperature is not necessary because the short-term photosynthetic process, PRI and SIF can represent, are affected by leaf or air temperature, not soil temperature.

Response: We agreed with your opinion that soil temperature had a little direct effect on the short-term photosynthesis process. However, the environmental variables used for analysis in this study are mainly divided into three categories: light-related variables (i.e. PAR and Q), temperature-related variables, and moisture-related variables. The latter two categories can both be divided into atmosphere-related (T_a and RH) and soil-related (T_s and SWC) variables. Therefore, we retained soil temperature as one of the predictor variables. Moreover, our results showed that T_s did affect the R²_PRI and we offered a possible explanation as follows:

• “…Soil temperature also had a positive effect on the PRI-based LUE model (Figure 6a and 7f). The high T_s occurred at the late V stages (DOY 195 to 207) with synchronously high RH (Figure 5). Thus, the maize rapidly grown at this stage had a higher photosynthetic rate and linear energy distribution, leading to the good performance of the PRI-based LUE model.” (line 429 to 432)

 

3. There is no explanation about upper panel of Fig. 1. I guess it is the aerial photograph but if so, the authors should explain it in the figure legend and indicate where the area, the study was done is.

Response: We re-plotted Figure 1 with marked panel (a), (b), and (c). The panel (a), i.e. the upper panel, was a satellite image of the maize field in our study area from Google Earth, which was stated in the title of Figure 1. We also marked where we did the study, i.e. the flux tower, in the panel (a).

 

4. It is difficult to distinguish between PRIs and PRIcan in Fig. 3 (c).

Response: We re-plotted Figure 3. In the revised Figure 3c, we used hollow and solid circles to distinguish PRI_s and PRI_can. Besides, we replaced the grayscale image with the color image to make the symbols clearer than before.

Author Response File: Author Response.docx