the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 18 Aug 2026
A High-Resolution Continuous Melter System for Sea Ice and Sediment-Laden Ice Applications
Abstract. Sea ice cores preserve a stratified record of biogeochemical processes that are central to understanding coupled ocean–ice–atmosphere processes. Previously, continuous melter systems were optimized for relatively clean glacial ice and perform poorly on sediment-laden sea ice. Here we present a new high-resolution continuous melter system specifically engineered for sea ice applications. This system combines capabilities of full core processing, inline filtration, and ice depth-resolution control. Further, this configuration allows simultaneous processing of inner and outer meltwater streams, collection of varied grain size fractions, and control for custom user-defined depth resolution during discrete sample collection. Validation experiments using artificial ice cores spanning sea ice densities (~0.7–0.9 g cm⁻³) presented reproducible volume collection, salinity boundaries preservation at sub-centimeter scales, low memory effects (typically <10 %), and high sediment recovery of ~86 % across a 1–106 µm size range. These results show that the new system presented here provides high-resolution discrete sampling of both dissolved and particulate phases, overcoming limitations of existing continuous melter system’s approaches for sediment-rich sea ice cores. With this advancement, new high-resolution analysis for sea ice can be achieved for biogeochemical and paleoclimate reconstructions in polar environments.
- Preprint
(1716 KB) - Metadata XML
-
Supplement
(800 KB) - BibTeX
- EndNote
Status: open (until 04 Nov 2026)
- RC1: 'Comment on egusphere-2026-4356', Matthew Corkill, 18 Sep 2026 reply
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 138 | 63 | 62 | 263 | 59 | 65 | 59 |
- HTML: 138
- PDF: 63
- XML: 62
- Total: 263
- Supplement: 59
- BibTeX: 65
- EndNote: 59
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
The authors present an innovative study on testing a new continuous melting system for sea ice based on glacial ice melter systems. They use artificial ice and solute-bearing (including sediment-analogue particles) aqueous samples to test the sampling at a vertical resolution of 0.5 cm through low-density ice and brine-enriched layers. Carryover and particle collection were also tested.
I think the tests undertaken by the authors are sound and demonstrate that the system could collect samples from sediment-laden sea ice in a laboratory setting, but as no data from natural sediment-laden ice cores is presented here, I think language around validation for this purpose should be softened. For instance, the authors do not yet know how the system will perform when it encounters layers of biology, including those with high concentrations of sticky exopolymers that are common in sea ice and that could cause issues in dead volumes of sample lines and with pre- and downstream filtration.
Nonetheless, the manuscript is very well written and describes the development of a novel sea-ice sampling system that could provide valuable high-resolution measurements important for studying sea-ice biogeochemical processes. The paper also gives a useful overview of existing continuous flow and continuous melting systems.
I’m looking forward to seeing future developments of this system, especially the performance with natural sea ice. Congratulations on the nice instrument.
Matt Corkill
L32: Maybe something like “sits at the interface” instead of “barrier” given fluxes that occur between the ocean and atmosphere through sea ice.
L35: Take care considering sea ice a static archive due to movement of material within sea ice from, e.g., brine convection, gap layer formation, melt-refreeze cycles and freeze concentration altering material (Meiners et al., 2025 doi:10.1002/9781394213764.ch16; Deschepper et al., 2026 doi:10.1016/B978-0-323-85242-5.00076-2).
L46-48: The vertical distributions of brine salinities and brine volume fractions are highly seasonal which I think should be mentioned, e.g., Zhou et al. (2013 doi:10.1002/jgrc.20232).
L49: I think this 35% refers to the seasonal sea-ice zone in one sector of the Southern Ocean, and only sea-ice algae production, not phytoplankton. I think this sentence should be adjusted a little as currently I read it as solutes are released during the melting season to support 35% of the total primary production (including phytoplankton) in the Southern Ocean. See Dalman et al. (2025 doi:10.1029/2024GL113717) for a recent sea-ice algae production estimate.
L51-53: I don’t think the cited review suggests increased nutrient supply due to recent sea-ice extent changes, see Lannuzel et al. (2020 doi:10.1038/s41558-020-00940-4) for an Arctic sea-ice biogeochemistry perspective paper that suggests uncertainty around changes to nutrient supply. Could consider inserting a reference for recent changes to sea-ice extent.
L56-58: There are processes other than sea ice formation and breakup that regulate ocean-ice exchanges, e.g., brine convection.
L65-69: Trace-metal clean ice corers are also used (e.g., Lannuzel et al., 2006 doi:10.1016/j.aca.2005.09.059), and in situ melting systems are being developed (e.g., Corkill et al., 2025 doi:10.1525/elementa.2024.00053).
L69-72: Brine loss is also a big challenge (e.g., Timco and Frederking 1996 doi: 10.1016/0165-232X(95)00007-X; Vancoppenolle et al., 2013 doi:10.5194/tcd-7-3209-2013).
Table 1: Nice table. In the first line there are 3 systems and then only 2 melt rates and depth resolutions presented. Maybe replace one of the slashes in the system column with a different separator to make it a bit clearer? Also, a little inconsistency in spaces around -, / and ~.
L165: “biogenic particles”?
L185-186: The authors could cite existing methods for digesting sea-ice particles here, very aggressive acid mixtures and high temperatures are usually used (e.g., Lannuzel et al., 2014 doi:10.1016/j.marchem.2014.02.006).
L234: I understand the reasoning behind the 2 mm pore space filter, but I wonder if, in the case of large particles being deposited on the mesh early, there could be some leaching/contamination in subsequent samples? Is the filter exchangeable during operation? Looking at the melt head design, I guess it likely is not, I wonder if it could be moved downstream a little and use a change-over system like the inline filtration units. I understand that it is tricky to move a pre-filter downstream, I think leaching from particles loaded on the pre-filter is worth thinking about though.
L250: This manifold and change-over system is really nice! Is it possible to flush offline channels with a cleaning solution, such as 10% hydrochloric acid? This cleaning is often performed between samples when filtering sea ice to measure trace metals.
Figure 3: In the bottom view of the melt head I think the outer channel and cartridge heater threads are labelled as being for thermocouples.
Figure 3: Small technical point for the taper-thread connections for sample lines, it looks like there is a small dead space in between the bottom of the fitting and where the thread in the melt head reduces to 4.76 mm. We tried to mitigate this in a sea-ice melting system by using a bore-through compression fitting and machining the melt head so the sample tube went through the fitting and sat snugly in a recess in the melt head.
L256-258: FMS acronym only expanded in Figure 2 caption. Can the measurement principle of the FMS be included?
Using the flow rate reduction for membrane saturation is really nice.
L291-297: Nice method for preparing low-density ice.
L316: The freezer operating temperature is -15°C, right? I think this is only mentioned in a Section 3 header, maybe also insert in Figure 2? I think it is also worth highlighting in this section that the test was undertaken at -15°C as the ice temperature is super important for not only brine movement but things like salt precipitation. Maybe worth mentioning the caveat that warmer sea ice with a higher brine volume fraction might be more susceptible to interlayer mixing.
Was conductivity of the outer channels also measured? It would be good to see that this is the same as the inner channels (if the ice grew in a semi-homogenous way and not from the outer edges inwards).
Figure 5. Nice figure. Can lines be added to the plot showing the bounds of the fresh and salty layers? I think that information is important to think about things like heat propagating upwards and causing brine to percolate down into the leading fresh layer. I’m also not sure if it has been mentioned yet whether ice is melted top-to-bottom or the other way around, i.e., was the 13 cm sample melted first?
L339-342: Testing for silicate in the future would be really interesting for sea-ice diatoms!
L366: Consider adding something like “(the equivalent of carryover in 0.5-1.5 cm for 0.5 cm resolution ice sampling)”.
L371-372: Was the analytical baseline DI water? Is a negative residual realistic? Think about removing this statement on considering positive and negative residuals together if the negative residual is below the detection limit.
L380: Consider spelling out Limit Of Detection and Quantification somewhere.
L383-387: I think it is worth adding a statement about these being upper-range values for sea ice, similar to the statement for macronutrients. It would be interesting to see lower values tested in the future as, e.g., Fe concentration as much as ~3 orders of magnitude lower than the spike concentration have been reported in sea ice (Lannuzel et al., 2016 doi:10.12952/journal.elementa.000130), and I think down at those concentrations is where it gets tricky.
L406: The 5-20% deficit could prove a challenge for particle carryover and should be mentioned in the discussion, I think.
L410: Can a brief explanation be added for why 1 µm was selected?
L451-453: ice types for sea ice could include ice with gap layers filled with brine or seawater due to thermodynamic processes or ice deformation (e.g., Ackley et al., 2008 doi:10.1029/2008GL033644; Fernández-Méndez et al., 2018 doi:10.3389/fmars.2018.00075), so while I agree this system is really nice for ice cores, maybe all ice types is a bit too broad.
L463-473: I agree that the conductivity data looks very promising, but the initial collection and preservation of the ice core is crucial and I think the authors should mention it. For example, in spring/summer a warm core and very porous core could lose a considerable amount of brine during coring and extraction, furthermore, if stored on its side brine could accumulate in one side of the core and partially freeze in storage. So, brine could be underrepresented overall, and the inner and outer melt streams could be quite different.
T2 separated brine-enriched layers with very impermeable fresh layers which would inhibit downward percolation from higher brine-enriched layers. I still think T2 was a very nice test, but the impermeable fresh layers may have contributed to preventing interlayer mixing in the ice core while it was melting. In natural ice cores brine-enriched layers will not necessarily be separated by impermeable layers so I think this should be mentioned.
L493-497: Consider adding a table in the main text showing limits of detection compared to ranges of values from the literature. I think it is very interesting for the reader if they’re considering using this method.
For dissolved iron, it looks like the LOD is ~0.5 ppb or ~9 nM, which is approaching the upper range for many Southern Ocean studies (Lannuzel et al., 2016 doi:10.12952/journal.elementa.000130). LODs close to median values for Mn, approaching upper for Cu and far above for Co (Duprat et al., 2021 doi:10.1525/elementa.2021.00032). I think the wording around being comparable to or below lower range values should be adjusted. Again, the table I mentioned above would really help, in that case comparison could be made with sediment-rich Arctic studies where these LODs are likely much more applicable.
L563: Possibly reword to include something like “validated under laboratory conditions” seeing as no data from natural sea ice is presented here.