the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Technical note: High-resolution analyses of concentrations and sizes of refractory black carbon particles deposited in northwestern Greenland over the past 350 years – Part 1: Continuous flow analysis of the SIGMA-D ice core using the wide-range Single-Particle Soot Photometer and a high-efficiency nebulizer
Remi Dallmayr
Yoshimi Ogawa-Tsukagawa
Nobuhiro Moteki
Tatsuhiro Mori
Sho Ohata
Yutaka Kondo
Makoto Koike
Motohiro Hirabayashi
Jun Ogata
Kyotaro Kitamura
Kenji Kawamura
Koji Fujita
Sumito Matoba
Naoko Nagatsuka
Akane Tsushima
Kaori Fukuda
Teruo Aoki
Download
- Final revised paper (published on 25 Nov 2024)
- Preprint (discussion started on 23 May 2024)
Interactive discussion
Status: closed
-
RC1: 'Comment on egusphere-2024-1496', Anonymous Referee #1, 04 Jun 2024
General comments
The pre-print manuscript “Technical note: High-resolution analyses of concentrations and sizes of black carbon particles deposited on northwest Greenland over the past 350 years – Part 1. Continuous flow analysis of the SIGMA-D ice core using a Wide-Range Single-Particle Soot Photometer and a high-efficiency nebulizer “ by Kumiko Goto-Azuma et al. describes and examines an instrumental coupling of a CFA feeding a Marin-5 nebulizer and a wide range SP2 to analyze the black carbon profile (size spectrum, number and mass) along a Greenland ice core. This article is the first part of the study, which will be supplemented by a companion paper devoted to the results of the analyses.
The article describes in great detail the analytical system used and provides a serious analysis of its performance. In several paragraphs, the authors compare their system with other existing CFA/Nebulizer/SP2 coupling, attempting to demonstrate that the latter underestimate BC masses for reasons of reduced or unstable nebulizer efficiency, or the size spectrum truncated by the classic SP2. To my knowledge, this comparison reflects parameters that are not considered by the authors, which brings into question some of their conclusions. Further investigation would be required to finalize this work.
Specific comments
Lines 76-78 : While the original SP2 allows incandescence measurements of BC particles with diameters of between 70 and 850-900nm (40-450nm according to DMT), the new SP2-XR model on the market is, according to DMT, suitable for incandescence measurements of BC particles with larger diameters ranging from 50 to 800nm. To date, I know of no research team working on snow and ice samples with this new SP2-XR, but it is used for atmospheric measurements.
You will also note the difference in size range between the data proposed by DMT and that of (Mori et al, 2016) for the original SP2 indicating a factor of 2 in measurable sizes. A specific study as done by (Mori et al, 2016) with an SP2-XR would suggest that the SP2-XR would be able to cover a wider range up to >1.5µm without modification. Unfortunately, this study is not available.
Line 117-118 : Can you explain how you obtained this depth resolution value of 0.3±0.1mm ? Is it only the resolution related to the laser positioning sensor or the resolution once the water is analyzed by the different online instruments ? Is this value defined solely by the fluidics of the CFA system (line, valves, debubbler, …) ?
Line 135 : This is not the focus of this article but can you clarify why your Picarro calibration is only done after the CFA session and not also before the measurements? Have you observed sufficient stability of the instrument if it is running all the time?
Line 151 : This comment concerns the supply of melt water to the nebulizer and then to the SP2. For accurate measurement of BC, the water flow must be controlled as it is proportional to the BC measurement. Peristaltic pumps are not the most stable over time and the flow rate varies according to wear on the tygon tube. What's more, this type of pump does not produce a stable flow, but a pulsed flow, as can be seen on an APEX/SP2 setup. Do you have any clarification on these points, and don't you think it's necessary to add a precise flow measurement before introduction into the nebulizer, using a Sensirion micro flow sensor for example ?
Line 166 + : While the internal mass and size calibration of the SP2 is relatively stable with time if the instrument is not moved, experience has shown that the nebulizer's efficiency is less so. Your method consists of measuring once this efficiency using PSL particles and Aquablack according to the size range. However, whether feeding an ICPMS or, in this case, an SP2, regular calibration of the nebulizer is necessary and this is generally done by analyzing a range of calibration solutions of known concentration on a daily frequency. How can you demonstrate that the Marin-5 model is more stable than the other nebulizers used for this type of experiment ?
In addition to these technical comments, can you provide more information about the post-processing of SP2 data in order to obtain BC mass and size profiles? The SP2 data files are relatively heavy, so some users try to extract them directly from the DMT software, but others turn to the PSI ToolKit.
Line 243 : Figure 2 does not convince me about the stability of the nebulizer, whatever the flow rate or particle size range. For 0.384mL.min-1, for example, for the <2µm section, the efficiency varies from around 27% to 42% (Fig. 2a).
Line 250 + : Signal dispersion. Dispersion tests are carried out using two solutions with different characteristics (in BC, ionic charge and isotopic composition) injected alternately through a valve under the melting head and then circulated to the analytical instruments. This is a good method, but a step is missing to estimate the impact of the melting head on this dispersion. Several parameters are not taken into account. 1) Even if the stratigraphy in the ice samples were perfectly horizontal, mixing would occur between the samples in the center of the ice stick and those on the outside of the inner ring (13mm?) mixed up to the port of the CFA line, 2) the ice strata are not always horizontal in the stick.
If we take into account only the interesting results of your method, this provides the basic parameters on the dispersion of the CFA and the analytical instruments. I'm quite surprised to see that the dispersion lengths (L1 and L2 average) are fairly similar between the instruments. It is known that the large dispersion in the Picarro is linked to a long cavity flush time, but this should be much shorter for the SP2 and ICPMS (to my knowledge closer to 10mm on other configurations). In addition, some studies have used these dispersion parameters to simulate a non-dispersed signal.
L261 : Yes of course the resolution of your CFA is better than these dispersion values, you may indeed observe a signal at a higher frequency, but the values observed will be reduced by this dispersion.
L268 + : Minimal loss of BC. That’s a great information for all BC measured by CFA that should be reproduced elsewhere!
L295 + : BC profile. A rolling average over 10mm is indeed necessary to smooth out the technical characteristics of the CFA, and for an initial assessment of the data. Unfortunately, Figure 6 does not allow this work to be properly appreciated, as it is too crowded. Consideration could be given to adding an enlarged extract of the profile over a short period of a few years in order to appreciate any seasonal variation in the BC, which would be an added value to the use of the CFA and its high resolution. You can save the full profile for future publications.
Lines 315 + : This brings us to the crux of the article, which proposes to demonstrate that configurations other than Nebulizer Marin-5 and WR-SP2 underestimate BC mass concentrations by XX%. It's not just the instrument and the measurement that come into play, but also the data processing. The low size limit of traditional SP2s is well known, which is also why DMT now offers an SP2-XR. Just because there are no measurements taken on sizes above 650 or 850nm does not mean that this part of the size spectrum is not considered. As shown in Figure 7, a Normalized dM/dlogD fit can be used to calculate the total mass (lognormal fit size distribution). This fit does not necessarily require measurements above 650 or 850nm to be correct if most of the peak is covered. To the best of my knowledge, but you can get in touch with the main users, to overcome the problem of the size spectrum being truncated at the top, classic SP2 users use the PSI ToolKit, which proposes the use of this fit in order to extract correct mass values. This last point should change the hasty conclusions of this manuscript.
For users of the U5000T nebulizer, on the other hand, there is a real problem of underestimation coming from instable nebulizer's efficenty.
Technical corrections
As a non-native English speaker, I will not be making any technical corrections to this manuscript.
Citation: https://doi.org/10.5194/egusphere-2024-1496-RC1 - AC1: 'Reply on RC1', Kumiko Goto-Azuma, 28 Jun 2024
-
RC2: 'Comment on egusphere-2024-1496', Anonymous Referee #2, 13 Jun 2024
This manuscript by Goto-Azuma et al. describes a continuous flow ice core analysis system with parameters used for the analysis of the SIGMA-D core from NW Greenland. This manuscript aims to describe the NIPR CFA system (including SP2, ICPMS, Picarro, etc.), conduct a detailed assessment of continuous ice-core BC analysis with the Marin 5 and wide range SP2 system, and introduce the analysis and dating of the SIGMA-D ice core. My overall impression of this manuscript is that while the methods presented here underpin some very interesting BC data from the SIGMA-D core (which are presented in a companion paper), it does not have sufficient novelty or focus to stand as a separate manuscript. I would suggest the authors revisit the purpose of this manuscript and reframe it with a more central goal as I think most of what is included would be more appropriate for the methods section of the science-focused manuscript. Hopefully my suggestions below are useful. I do think the resulting datasets (discussed in the companion paper) are very interesting and appear to be quite robust, but reiterate that I do not think this methods manuscript holds up very well on its own in its current form.
First, the title and abstract indicate the main goal of this manuscript is to present the application of the wide-range SP2 + Marin 5 for continuous ice core analysis. The SP2+nebulizer system has been used for continuous analysis in a number of ice core labs and the details of this specific system for BC measurements in liquid water have been presented previously (Mori at al., 2016), so the assessment of the modified SP2 and Marin 5 nebulizer system is not particularly novel. Most of the other methods presented here (e.g. ice core CFA SP2, Picarro, and ICPMS analysis) are also well-established, with the exception of the BC particle size measurements, and therefore are more appropriate for a methods section of a science-oriented paper in my opinion.
The measurement of BC size distributions throughout the core, though, is quite novel and exciting, but this manuscript lacks detail or justification for this specific measurement. There have not been published long-term reconstructions of BC particles size from Arctic ice cores, largely because as the authors correctly state, “obtaining accurate estimation of the size distribution of BC particles on a routine basis is not easy” (line 67). I agree, and a major reason why is because it is extremely difficult to maintain a stable SP2 response/calibration and nebulizer efficiency throughout an ice core CFA campaign. However, the manuscript did not justify how the authors have overcome these challenges to apply this method to ice core CFA analysis, where it is crucial to demonstrate stability and reproducibility of the method to ensure consistent measurements over weeks and/or months of ongoing analysis. I think more detail is warranted on how the authors ensure a coherent BC size dataset throughout the SIGMA-D analysis, which likely spanned a few months given the stated analysis rate of 6-7 m on one to two analysis days per week. Was the SP2-Marin5 system stable throughout an analysis day, week, month, etc.? Were replicate ice sections analyzed with good agreement? What protocols were used or standards run to ensure a consistent dataset? How was SP2 data processed? Investigating these questions will require presenting some timeseries BC size distribution data, which is omitted entirely in this manuscript despite its emphasis in the title and abstract. Only BC mass and number concentration timeseries are shown and even then, the figures are too small to evaluate the timeseries data.
Other aspects of the BC dataset that would be valuable to assess would be the Marin 5’s performance against the Apex Q, which is more prevalent now for ice/snow analysis than the Cetac U5000AT and also has much better nebulization efficiency for large particles (Wendl et al., 2014). While the Cetac was originally the nebulizer of choice for the SP2 ice core method when it was first introduced (McConnell et al., 2007), I don’t think the Cetac should be the benchmark for the underestimation of BC concentration for a ‘standard’ ice core method anymore since many groups have moved away from it (largely because of its efficiency and stability issues). Lastly, it should be made explicit that many of the findings related to BC concentration underestimation in ice cores presented here apply primarily to Arctic and alpine ice core sites. The choice of nebulizer (Cetac U500AT vs Apex Q, at least) does not seem to impact BC concentrations for Antarctic ice cores sites as much given the much smaller particles and lower BC concs observed at those sites (Arienzo et al., 2016, JGR, Supplemental Fig 1).
Other sections of the manuscript, including the description of the complete CFA setup with the new addition of the ICPMS and preliminary dating of the SIGMA-D ice core, seemed extraneous and distracting to me from the more exciting BC size distribution idea. As I mentioned previously, I think those sections are more appropriate for the methods section of the science focused manuscript as they are largely established methods. Additionally, the dating section did not include enough detail to be compelling (for example the dating section only showed ~3 m of annual layer counting and did not show the tritium ties or volcanic synchronization).
In short, this Part 1 manuscript, which is framed as a BC methods paper by the title and abstract, does not have sufficient novelty or detail to stand alone in its current form. In my opinion, it is better suited to be included as a methods section for the scientific paper unless the manuscript is refocused around the novel BC size distribution method.
Other comments
Line 117: 0.3 +/- 0.1 mm depth resolution seems incorrect- are the units right?
Line 288-289: If the dating section stays, it would be worth including a figure showing volcanic synchronization. What is meant by ‘made adjustment’ prior to 1783? What exactly was adjusted?
Lines 298-300: Are the sporadic peaks attributed to large particles reproducible? What do the BC size distributions look like for those depths? It would be interesting to understand if any meaningful interpretations can be drawn from them. If they are just filtered out of the data and considered noise, then what is there any advantage of using the wide-range SP2 over a standard one?
Lines 329-331: While the CFA system is capable of measuring water isotopes, ICPMS, microparticles, and methane, I don’t think that is demonstrated in this manuscript and distracts from the BC focus.
Does the paper address relevant scientific questions within the scope of ACP?
Yes, ice core BC size distribution and concentration measurements are within the scope of ACP.
Does the paper present novel concepts, ideas, tools, or data?
The BC size distribution method is novel, but the other aspects of the manuscript (dating, ice core CFA analysis) not so much.
Are substantial conclusions reached?
No. I do not think the manuscript reaches substantial conclusions, as the novel aspect of the manuscript (the BC particle mass method) is not well described, and the results/conclusions of the SIGMA-D analysis are discussed in a companion paper.
Are the scientific methods and assumptions valid and clearly outlined? Are the results sufficient to support the interpretations and conclusions? Is the description of experiments and calculations sufficiently complete and precise to allow their reproduction by fellow scientists (traceability of results)?
No. More information needed on the stability and reproducibility of the BC size distribution method over the course of the ice core analysis. The BC size distribution records are not presented in this manuscript making it difficult to assess the method.
Do the authors give proper credit to related work and clearly indicate their own new/original contribution?
Yes
Does the title clearly reflect the contents of the paper? Does the abstract provide a concise and complete summary? Is the overall presentation well structured and clear?
Somewhat. The title and abstract focus on BC measurements, but the manuscript also includes sections about the full NIPR CFA system and SIGMA-D ice core dating that I found distracting.
Is the language fluent and precise? Are mathematical formulae, symbols, abbreviations, and units correctly defined and used?
Yes
Should any parts of the paper (text, formulae, figures, tables) be clarified, reduced, combined, or eliminated?
Yes. Described in comments above.
Are the number and quality of references appropriate?
Yes, though lacks citations to more recently published Arctic BC records.
Citation: https://doi.org/10.5194/egusphere-2024-1496-RC2 -
AC2: 'Reply on RC2', Kumiko Goto-Azuma, 28 Jun 2024
We thank Referee 2 for the very valuable and helpful comments. We would like to revise the manuscript, taking all the comments into consideration.
Our responses to the Referee’s comments are attached.
On behalf of all Co-Authors,
Kumiko Goto-Azuma
-
AC2: 'Reply on RC2', Kumiko Goto-Azuma, 28 Jun 2024
-
RC3: 'Comment on egusphere-2024-1496', Anonymous Referee #3, 16 Jun 2024
This manuscript “Technical note: High-resolution analyses of concentrations and sizes of black carbon particles deposited on northwest Greenland over the past 350 years – Part 1. Continuous flow analysis of the SIGMA-D ice core using a Wide-Range Single-Particle Soot Photometer and a high-efficiency nebulizer“ submitted by Goto-Azuma et al. describes an improved CFA system by coupling single-particle soot photometer and a high-efficiency nebulizer. This technique is suitable to perform high-resolution measurements of black carbon (BC) regarding concentration, as well as size distribution up to 4 μm. The authors applied this technique to analyze the BC particles in an ice core retrieved at the SIGMA-D site from the northwest Greenland. This manuscript is accompanied by a following part focusing on the 350-year BC record of the SIGMA-D ice core. This work has advanced the conventional CFA system, especially concerning to the size distribution of BC that has been less considered before. Therefore, this specific merit deserves a publication in an esteemed journal such as ACP.
I suggest that the authors might discuss the innovative content (e.g., size distribution of BC particles) in more details, and simply the other parts that have been considered thoroughly in previous papers, or to include the other parts in the supplementary material. Accordingly, the introduction should be revised for a concise review of previous works, but focus more on its novelties.
Other comments:
Lines 304-314: The authors claim that a combination of the standard SP2 and a high efficiency nebulizer, and a combination of the standard SP2 and a traditional ultrasonic nebulizer would lead to underestimation of the averaged mass concentration by 12% and 17%, respectively. However, I don’t know if the authors analyzed BC concentration using these two conventional methods. Please provide more details to reach this conclusion.
Line 29: Please take a check on the resolution value of 10-40 mm.
Line 58: The words “have become possible” should be deleted.
Lines 117-118: What's the meaning for saying the depth resolution value of 0.3 ± 0.1 mm?
Lines 186-187: Please explain in more details how to calculate the reproducibility.
Line 224: “± 0:05‰”→“± 0.05‰”
Line 238: “± 0:08‰”→“± 0.08‰”
Citation: https://doi.org/10.5194/egusphere-2024-1496-RC3 - AC3: 'Reply on RC3', Kumiko Goto-Azuma, 28 Jun 2024