Reviewing the revised manuscript entitled “Synchrony in catchment stream colour levels is driven by both local and regional climate” (Doyle et al., submitted to Biogeoscience)
General comments to the authors:
I was excited to see the responses of the authors to the referee comments and found most of my previous specific comments to be adequately addressed (but see exceptions below). However, I could not find any response to my previous general comments that contained a number of major concerns. I would expect these comments to be specifically addressed before I can give any further recommendations.
Below, I pasted the previously posted general comments and also added a number of new specific comments in reply to the authors response to my previous specific comments.
Previously posted but unaddressed general comments:
“The well-written and well-organized manuscript addresses an important area of research in biogeochemistry / earth system science: how do terrestrial carbon stocks respond to climate change and what is the role of inland waters as a potential export pathway of these stocks? The authors acknowledge the large amount of previous work published on this topic, but in regard to this, should highlight more the knowledge gap and niche that the present manuscript fills. The authors used an interesting approach of cross-wavelet time-series analysis to investigate time scales of magnitudes and directions of correlations. In biogeochemistry, especially peatland biogeochemistry, this analysis is novel and allows new insights in addition to classical correlation analysis. The results and implications of this analysis appear to me a bit drowned among the wealth of other results (which are more straightforward and not very novel, especially in the light of the findings by Ryder et al. 2014). Can the authors highlight the cross-wavelet time-series results more to make a clear stand of how it contributes to new insights? What can we learn from this analysis about the time scales of effects (a good example is given on p. 13, L. 31), what are the problems (as touched upon on p. 15, L. 4-5)?
The two parts of the paper, the analysis of environmental correlates of water color and variability in DOC exports, are interesting in themselves, but what would be even more interesting to many readers was to combine the results of the two analyses: what explains inter-annual and seasonal variability in DOC exports? This is touched upon on p. 16, L. 7-8, but not directly shown. Doesn’t the available data allow a more direct analysis of environmental drivers of DOC export (at a seasonal and annual scale)?
The authors highlight both “local” and “regional” climate in the title. To rigorously support this title, wouldn’t it be helpful to also explicitly analyze how the regional climate affects the local climate (soil temperature) and to what extent this effect is similar across catchments? If such an analysis would reveal that the regional climate affects local soil temperatures independent of catchment characteristics, it may explain the synchrony in water color responses to climate across catchments.
There is a mismatch between stated goals, analyses and conclusions. In the conclusions, the authors highlight that environmental drivers of water color can differ depending on catchment specific characteristics. However, almost all drivers were the same in magnitude and direction across the different catchments (Fig. 4, and see also p. 10, L. 17-22 and p. 11, L. 13-25). Moreover, the authors only tested for differences in DOC loads (but not drivers) across catchments. An analysis of whether drivers differed between catchments would, however, be not very powerful, given that only three catchments were included which, after all were rather similar in key characteristics such as the percentage of streams intersecting peat (Table 1). To strengthen the qualitative discussion of how catchment characteristics modify drivers and magnitudes of DOC loads, it would be valuable to know what range of environmental conditions the studied catchments cover relative to the whole range of Irish blanket peatlands.”
New specific comments to the authors:
I’d like to follow up on one of my previous comments [1]: The authors respond that they did not statistically test for the correlation between DOC loads and climate variability. I agree and understand that 6 years of data is not enough for robust statistical testing. However, in response to another of my previous comments [2], the authors reply that the sentence now reads “‘The results of the study highlighted the role of climate in controlling stream water DOC concentrations, and aquatic carbon export”. I agree with the first part, that inter-annual variability in weather conditions affect stream water DOC concentrations and that this may also imply variability in aquatic carbon export (depending on how variability in discharge relates to variability in DOC concentrations). Yet, I’d like to note that the authors did not test for the effects on aquatic carbon export (as they state in response to comment [1]. First, carbon includes also inorganic or particulate species, but here only dissolved organic carbon is referred to. Hence, "carbon" should be specified. Second, I suggest to make clear that conclusions on the effects of exports are rather speculative, as they were strictly speaking not statistically tested for.
[1] “p. 2, L. 24: it would be interesting for a wide readership to know whether these interannual variations in DOC loads are linked to variability in climate. This remains unclear in the way it is phrased here.”
[2] “p. 2, L. 24: Can the authors specify what is meant with “interaction of catchment conditions and regional meteorological drivers? what characteristics makes DOC export
from a catchment more or less susceptible to environmental drivers? This should be
highlighted here, or at least in the conclusions of the manuscript, if supported by the
data.”
I’d like to clarify my previous comments [3] and [4] that I feel were not addressed as I hoped for (maybe because of language issues). What I suggest is to give contact details for the vendor, i.e. “Orpheus Mini Water Level Logger (OTT HydroMet, Kempten, Germany).”
[3] p. 6, L. 30: please give the location of the vendor of the water level loggers.
[4] p. 9, L. 1: add details on the location of the vendor of the DOC analyzer.
I am not satisfied with the response to my previous comment [5] to include wind speed, radiation and humidity in the model. First, the authors argue that wind speed and air temperature are correlated and that wind speed should correlate with DOC concentrations because of the effect of temperature on DOC concentrations. The authors include air temperature itself as a variable in the model but do not give any reason why wind speed should have an effect in addition to the unique effect of air temperature. I suggest to only include variables in the model for which a plausible mechanistic link to the outcome variable can be expected and to provide such reasoning in the text. Second, the authors did not give any motivation / hypothesis on why humidity and radiation was included. What is the mechanism behind the hypothesized effect of humidity and radiation?
[5] p. 8, L. 6: What is the motivation to include wind speed, radiation and humidity in
the model? Background / hypotheses for testing these variables are not given in the
introduction.
I do not agree with the authors suggestion in response to my previous comment [6] to remove the reference to the Monte Carlo methods from the manuscript. As I understand the functions in the biwavelet package, there is a number of parameters to select when doing the cross-wavelet transform analysis, e.g. the number of Monte Carlo randomizations. Please specify these parameters in order to provide enough detail for the reader to redo the analysis. Which function from the biwavelet package was used? What parameters were chosen?
[6] p. 8, L. 27: Please give more details on the Monte Carlo methods used! |
- Printer-friendly version
- Supplement
