the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Technical Note on high-frequency, multi-elemental stream water monitoring: experiences, feedbacks, and suggestions from seven years of running three French field laboratories (Riverlabs)
Abstract. High-frequency and multi-elemental stream water monitoring are acknowledged as necessary to address data limitation in the fields of catchment sciences and freshwater biogeochemistry. In recent years, the development of stream bank analyzers and on-site field laboratories to measure various solutes and/or isotopes at sub-hourly measurement intervals is in progress at an increasing number of sites. This trend should likely persist in the future. Here, we share our experiences of running three French field laboratories (called Riverlabs) over seven years. This technical note gives an overview of the technical and organizational points that we identify as critical in order to provide guidelines for the successful implementation of future projects running such equipment. We therefore share the main stages in the deployment of this tool in the field, the difficulties we encountered and the procedures we used to identify and eliminate their causes. Some of the critical aspects discussed here relate to 1) Supply of the field laboratory: basic functioning of the pumping, filtration and analytical systems, 2) Data quality control and assurance via maintenance services and operations, 3) Data harmonization and coordination of the laboratory components, and 4) Team structure, skills and organization. Our two main conclusions for a successful, long-term functioning of these types of field laboratories are, first, the necessity to adapt several central components of the field laboratory to the local conditions (climate, section, topography, water turbidity, power) and, second, the need of diverse and in-depth technical skills within the engineering team. We believe that sharing these experiences, combined with providing some practical suggestions might be useful for colleagues, who are starting to deploy such or similar field laboratories. These considerations will save time, improve performance and ensure continuous field monitoring.
- Preprint
(1486 KB) - Metadata XML
-
Supplement
(140 KB) - BibTeX
- EndNote
Status: open (until 01 Jun 2024)
-
RC1: 'Comment on egusphere-2024-902', Jennifer Druhan, 01 May 2024
reply
Review of egusphere-2024-902: Technical Note on high frequency, multi-elemental stream water monitoring: experiences, feedbacks and suggestions from seven years of running three French field laboratories (Riverlabs).
This technical note (TN) is intended to review the challenges and considerations that arise in the placement, operation and maintenance of a specialized deployment designed to continuously analyze a flexible and rigorous suite of geochemical parameters in rivers semi-autonomously and in real time. I should clarify at the start that I also run a Riverlab facility, designed by the same company (Extralab). Ours was designed and deployed more recently (2021) and is the only one stationed in North America. From this rather unique perspective I can say that I appreciate and support the intended purpose of this TN, as it is indeed very challenging to appropriately run one of these facilities. However, I am concerned that there’s sparingly little information offered in this TN that is relevant to the deployment we operate. Many of the issues described have been alleviated by updates to the Riverlab technology, while others are a result of the site-specific hydrogeochemical behavior of the Naizin and Strengbach watersheds rather than universally applicable guidance. Finally, there are multiple vital aspects of successful Riverlab operation that are overlooked or omitted here. For these reasons I do not believe that this TN serves its intended purpose of offering experience to benefit future deployments and I recommend that the manuscript be rewritten to reflect the current state of knowledge in working with such technology prior to further consideration for publication. I have detailed the major points this would require below, followed by in-line edits. These are offered in a spirit of collaboration and in hope of ultimately advancing rather than inhibiting this emerging technology.
Jennifer Druhan
Major edits:
- Current state of methodology: The final sentence of the abstract states “These considerations will same time, improve performance and ensure continuous field monitoring”. It is necessary that the paper accomplish these tasks. The challenge faced is that this TN is based on three RiverLab facilities which were deployed between 2015 and 2017. In the interim since, there has been considerable development on the part of the manufacturer to improve a variety of aspects in the overall system design, operation, automation, and management, which are unfortunately missing in this TN. While some of the experiences described in this document do still offer useful guidance, many of the problems discussed in detail here (pump design, flow and pressure optimization, strainer, and intake issues, etc.) have been alleviated or are at least mitigated with new approaches. Unfortunately, this means that the information provided here is largely out of date, specific only to very dilute systems, and thus not constructed in a manner that offers universal guidance to future deployments.
- Missing factors: The TN almost entirely overlooks the challenges associated with data management and software development necessary to automate the RiverLab facilities and ensure preservation, analysis and monitoring of large volumes of data. I think this would be a much more useful and appropriate topic for section 6 than the current “Team structure”.
- Repetition and omission of existing literature: The extent to which this paper overlaps with the Floury et al. 2017 Hydrol. Earth Syst. Sci publication on the Orgeval riverlab is concerning. This prior publication detailed technical aspects of the operation including:
- design, calibration, analytical performance;
- accuracy and instrumental drift of the chromatography system;
- whole-system precision and testing of cross-contamination using a salt test which appears to be the basis of the approach described here;
- laboratory analysis to confirm reproducibility of major ion concentrations;
- a detailed discussion of what is gained through high sampling frequency and improved analytical precision.
In many ways this earlier publication already offers much of what is presented in the current TN, in some cases more successfully. At the same time, the current TN omits any significant use of the Orgeval facility or data, even though it is presented as one of the three deployments included in this note. Relatedly, I must point out that the authors have overlooked a recent publication by Wang et al. (2024) in STOTEN which used the Orgeval dataset, in combination with the earlier Plynlimon data and our own riverlab deployment in Illinois. I was disappointed to see that such a relevant study, which uses data from one of the French deployments, was not listed among the publications used to demonstrate recent advancements based on these novel datasets. I recognize that the paper is from 2024, but it was out months before the Floury et al. (2024) Orgeval paper which the authors do cite, and both papers included Dr. Floury as a coauthor. Given that Dr. Floury is also coauthoring this TN, the omission is rather blatant. Finally, the TN does reference a submitted manuscript by the lead author which is not yet through the review process. This should be removed until it is appropriate to cite as a published study.
In-line edits:
L28: what is meant by “section”?
L34: “over the last two decades” but the oldest citation is from 2012
L36: “with the conviction that concentration data sets with increased resolution will be the key of advancing environmental sciences” I think this is a rather vague appraisal of the relevant Kirchner papers, which develop the need for high resolution chemical records based on frequency analysis. This is not a ‘conviction’ this is a quantitative analysis.
L38: what is a “riverbank side analyzer”?
L38: “can be qualified as the most sophisticated technologies” this is subjective and, in my opinion, incorrect. There are tremendous advances now occurring in a variety of portable and field-deployable riverine geochemical analysis that could be considered “more advanced” than a Riverlab.
L39-40: “avoid some disadvantages…” I have to disagree here as well I don’t think sample storage or delay in analysis are the major issues field-deployed real-time analytical capabilities circumvent.
L44: for the use in
L46-51: I’m failing to see how a summary / reiteration of the major conclusions of the Bieroza paper is relevant to the current scope and purpose of this technical note.
L54: it is very concerning that this statement appears to omit consideration of Floury et al. 2017 paper “The potentiometric Symphony: New progress in the high-frequency acquisition of stream chemical data”, which offers extensive documentation of the technical and operational aspects of the Orgeval riverlab deployment, to the point that aspects of the current TN may be repetitive in respect to this earlier publication.
L55: ‘urging’ is not the correct word here.
L66: “we usually illustrate encountered issues with illustrations” please correct this statement it’s not appropriately constructed.
L76: “and by a weathered layer” -- I’m not sure what this means.
L80: The description of Naizin should include a clear explanation of the extent to which this stream is ephemeral.
L98-101: This is really confusing. Either the three riverlabs are “almost identical” or the latter two have “higher similarity”. I’m pointing this out specifically because the remainder of the TN is so heavily focused on the Naizin and Strengbach systems while Orgeval is almost entirely omitted. If this is the reason why, then perhaps it is more appropriate to remove Orgeval entirely given the existing Floury et al. 2017 publication already describes a lot of what’s to come.
L107: I think it’s important to mention here that this is not the only design option for the pump configuration.
L108: missing turbidity sensor?
L139: “small fishes”?
L154: This is just a consequence of the choice of which submersible pump one deploys. There are plenty that are designed to handle mud.
L171: The information provided here is outdated: These problems associated with variable frequency have been largely dealt with in most current PID offered by the manufacturer.
L184-185: “… the absolute parameter values might not be very informative for other users, which use different PID systems”. I think this is specifically the sort of issue that a rewriting of the TN should seek to alleviate.
L204: this appears to be a site-specific design flaw, rather than universally helpful guidance.
L206-210: My understanding is that the pump is stopped and operations are ceased at Strengbach during prolonged freezes. This seems misleading?
L216: this would imply that standard protocols for collection of field samples and subsequent analysis in a laboratory, many of which involve filtration, are biased?
L224: what exactly constitutes ‘heavy fouling’? this seems quite arbitrary
L233-235: We simply change the filter twice a week. This doesn’t appear to be a critical issue
L255: this seems like a design issue specific to this deployment. I have to say a lot of this section seems like issues specific to this first set of deployments which have been addressed since. It’s thus hard to see how this serves as universal guidance for future deployments
L265-268: this is nonsensical. Of course the concentrations are different after filtration the particulates have been removed
L279: This ‘crucial’ statement is entirely empirical. Direct evidence should be provided that calibration curves are undermined after tubing changes, else this should be removed
L287: this is a simple process. I don’t see why it’s being so heavily emphasized here.
L295: again this is a site-specific design issue. The venting problem is not universal and the newer designs have circumvented this problem
L326: This is totally inappropriate – a TN cannot publish inferences from a conversation with a manufacturing engineer. This must be removed.
L335: I have never seen any evidence that that a lower ion concentration causes faster wear of a pH probe. Please back this up with some references
L338: this statement has been made multiple times through the manuscript with no examples. Please remove.
L340-346: Is this really a problem? It would seem the issue is related to what you are trying to measure and the frequency with which it changes. For example the 8 minute delay based on the salt test at Naizin (L357), you can’t even see this in the IC measurement frequency
L369-375: Again, I’m failing to see whether this really matters – in comparison for example to where in the water column the intake system is situated. This seems quite specific to a small system
L391-392: I find this entirely subjective, based on location of the deployment relative to those who can intervene. This is far too site specific to be offered as general guidance
L420: exactly what ‘waste’ is this referring to?
L434-447: this is entirely too specific to the manufacturer and specific technical support for these deployments. Other options exist, which circumvent these issues. I also don’t think it’s appropriate to publish a TN arguing for a costly maintenance contract with a corporation.
L471-472: This sentence should become an entire section of the TN. The complexity of software and data preservation necessary to run these systems is extraordinary, and in this sense the TN could offer a very helpful and universal guide for new deployments that is distict from the earlier Floury et al. (2017) paper
Section 6 should be removed. This is not useful or appropriate.
L512: hopefully beyond European scale?
L521: I really have to push back here: the idea of a 2-3 year “settling period” is absurd. The instrumentation is free of wear and in its best operational shape immediately after deployment. 2-3 years in many issues begin to arise. This is the most important period to collect data.
Citation: https://doi.org/10.5194/egusphere-2024-902-RC1
Viewed
HTML | XML | Total | Supplement | BibTeX | EndNote | |
---|---|---|---|---|---|---|
144 | 38 | 10 | 192 | 8 | 5 | 6 |
- HTML: 144
- PDF: 38
- XML: 10
- Total: 192
- Supplement: 8
- BibTeX: 5
- EndNote: 6
Viewed (geographical distribution)
Country | # | Views | % |
---|
Total: | 0 |
HTML: | 0 |
PDF: | 0 |
XML: | 0 |
- 1