Carbon sequestration in different urban vegetation types in Southern Finland

Thoelix et al.

General comments

The study employs three models of different detail to analyze carbon fluxes at three sites of different green infrastructure. The models are initialized, parameterized, and partly evaluated with measured data. From the description, the evaluation seems to be a bit poor in cases, particularly regarding the grassland site, which makes the comparison of different plant types rather difficult. To my feeling, the drought stress, which might be an important driver for carbon dynamics is not convincingly reflected in the models. Nevertheless, the simulations seem to be able to represent the overall carbon fluxes for the tree sites and results might also give an indication about the grassland dynamics. Still, I would be happier when the discussion would be more carefully formulated. Despite deficiencies, I think that the analysis has some merit and may serve for an improved measurement setup as well as model developments.

Specific comments

Abstract

P1L1: ‘in sequestration potentials’ instead ‘in the sinks’

P1L6: delete ‘momentary’

P1L9: replace ‘concerning mature trees and that in urban areas’ with ‘because’ (and adjust to ‘includes’ in the next line)

P1L12: do you mean ‘for the same vegetation type’ instead ‘across a studied urban vegetation area’?

P1L16: replace ‘key’ by ‘single’

Introduction

P2L27: Some more recent reviews would be e.g. Ferrini et al. 2020 or Cuthbert et al. 2022

P2L29ff: I see what you would like to say. Still, I think it would be favorable if the argumentation would be clearer that carbon fluxes are on the one hand indicators for other ecosystem services and on the other hand easy targets for physiological-oriented models. And that therefore the models need to be employed in order to also judge the development of ecosystem services.

P2L54: replace ‘utilized’ with ‘used to estimate carbon sequestration’

P3L62: replace ‘validation’ by ‘evaluation’ (also check throughout the manuscript)

Methods

P10L250: What is meant with ‘momentary photosynthesis’ is a simulated assimilation rate that is based on half-hourly climate data which are supposed to be representative for this period, correct? TER and NEE are calculated for the same periods but are not called ‘momentary’. Please proof me wrong but I think this term inconsistent, not exactly correct, and superfluous.

P10L264: Is it correct that the models have some light absorption algorithms that consider that parts of the tree canopies are always shaded? Could you indicated that such a process is applied in all models to account for self-shading? Otherwise photosynthesis rates would be of course too high.

P11L272: What does ‘set to a fixed value of 415ppm’ mean. Is it possible to regulate the CO2 concentration in such a small measurement chamber? Did you check the ambient CO2 concentration? As it is known that CO2 concentrations in cities are often way higher than that, it seems necessary to confirm that this is not the case or explain how the models are able to address this issue.

P11L285: Soil respiration was measured at 8 plots at each of two sites? How were they distributed over the sites to ensure representativeness? What about the grassland site – are the models evaluated with it before/ at other places?

P12L303: In order to directly compare measurements and simulations, it would be favorable, if you could derive mm m-2 h/day-1 values from the sapflow measurements. Could this be done?

Results

P13L333: If you have simulations and data for 2020 too, please show them also in the graphs.

P16L366ff: If I understand correctly, it is not possible to judge if simulated transpiration rates are higher or lower than reality. The comparison as it is in Fig. 5 is only indicating that the dynamics are possibly related. Thus, formulations need to be chosen very carefully. Please revise.

P19L407ff: Does LPJ Guess calculate soil respiration? In this case, not only RH (to compare with JSBACH) but also the full value should be shown in Fig. 7 and be discussed. I would also suggest to put in a 1:1 line.

P20L442: This is not very well formulated. While single peaks of emission in summer are indeed very high with LPJ Guess, I daresay that the vegetation period is still a larger sink than the winter time.

P20L443: similar over the target area – what area? Or over the years?

P21L445: parenthesis?

Discussion

P22L83ff: The first paragraph is more or less a repetition of parts in the introduction. Consider deleting.

P23L493ff: replace ‘absolute values’ by ‘measurements’, ‘simulated correctly’ by ‘realistically represented’ and ‘mainly correct’ by ‘reasonably close to measurements’ or similar. I am also struggling with the expression of ‘systematic discrepancies. Do you mean that the model structure and output sometimes doesn’t match the scale and targets of measurements? Consider rephrasing.

P24L504ff: Don’t the models consider interception? Your assumption implies that canopy interception is underestimated although the leaf area is more or less correctly met. This doesn’t seem very likely. Also, I cannot imagine that stem water refilling might be responsible for the long-term water deficit. On the other hand, I could imagine that soil properties, in particular preferential flows (in addition to runoff) might represent a considerable uncertainty.

P25L514: replace ‘from vegetation types observed in detail’ by ‘because of a different composition or structure of the actual vegetation’

P25L519: replace ‘future studies would benefit from’ with ‘measurements would particularly been improved by’

P25L520: replace ‘for’ by ‘to measure’

P25L523: replace ‘problematic to definitively determine’ by ‘not possible to decide’ or similar

P25L526: replace ‘is challenging’ by ‘has its drawbacks’ or similar. In addition, it might be argued while some targets e.g. drought related phenology was best met in LPJ, others, such as the carbon exchange during spring, was better represented by the other models. And thus, every model has its strong- and weak points.

P27L593ff: Did you check radiation? Years might be particularly good if cloudiness is low and radiation might be high.

P28L620ff: Consider rephrasing. What do you actually recommend? An empirical adjustment of phenological pattern for cities? A better initialization based on measurements? A better representation of varying environmental impacts such drought into the models?

Mentioned references

Cuthbert, M. O., Rau, G. C., Ekström, M., O’Carroll, D. M., and Bates, A. J.: Global climate-driven trade-offs between the water retention and cooling benefits of urban greening, Nature Commun., 13, 518, https//doi.org/10.1038/s41467-022-28160-8, 2022.

Ferrini, F., Fini, A., Mori, J., and Gori, A.: Role of Vegetation as a Mitigating Factor in the Urban Context, Sustainability, 12, 4247, https//doi.org/10.3390/su12104247, 2020.

The article compares 3 models to analyze carbon sequestration in urban vegetation (irrigated and non-irrigated lawns, park trees, and urban forests). The study considered various parameters

such as soil moisture and temperature, sap flow, leaf area index, momentary photosynthesis, soil respiration, and net ecosystem exchange. Evaluation of all these parameters and their presentation makes this article a bit complicated. Repetition at many places in the Discussion makes it too lengthy.

 The Abstract sufficiently represents all aspects of the research work.

The introduction is properly written with clear aims and objectives. In some places ( P2L28-28), a few unrelated topics may be removed.

Materials and Methods is somewhat lengthy. It should be precise and easy to understand.

Observation represents all aspects of the study. In P10L250 is Momentary photosynthesis GPP? Make it clear.

 The results of the study are clearly presented. In P13L338 it is written “The soil was moister in 2020 than 2021”, but data for 2020 is not graphically presented as to compare with 2021.

The discussion contains repetition at places.

References are sufficiently provided.