Skip to content
Field Notes

The research side of permaculture.studio.

Agriculture & Soil6 min read

Five years of glacial rock flour on a Danish field: what the numbers do and do not show

A five-year Danish field trial found glacial rock flour dissolved and raised soil carbon, but its largest removal figure rests on an assumption about the ocean.

What the research shows

  • In a five-year field trial on a sandy agricultural soil in Denmark, an estimated 56 percent of glacial rock flour applied at 50 tonnes per hectare dissolved. Field trial
  • Treated plots showed a significant increase in soil organic carbon, equivalent to 18.3 tonnes of CO2 per hectare, concentrated in the mineral-associated organic matter fraction. Field trial
  • Greenlandic glacial rock flour contains high concentrations of silicate and trace metals, including iron and manganese. UnclassifiedThe study design could not be determined automatically.
  • In a high-Arctic glacierised catchment, sediment-bound iron and silicon yields were several times higher than the dissolved fluxes of the same elements. UnclassifiedThe study design could not be determined automatically.
  • The authors estimate in-field CO2 uptake of 3.2 to 4.9 tonnes of CO2 per hectare, rising to 6.39 tonnes only if downstream transport and re-equilibration with the marine carbonate system is assumed. Field trialpartially supported
  • Whether glacier discharge increases or decreases marine primary production depends on glacier type and which resource limits growth, so downstream effects on the ocean are not uniformly a carbon sink. Unclassifiedpartially supportedThe study design could not be determined automatically.

Each of these was checked against the abstracts of the papers it cites. See the receipts.

Five years of glacial rock flour on a Danish field: what the numbers do and do not show

A five-year field trial in Denmark found that about 56 percent of Greenlandic glacial rock flour spread on a sandy agricultural soil dissolved, and that treated plots held more soil organic carbon than controls. Both are field measurements. The paper notes that long-term field constraints on weathering rates and carbon sequestration have been limited 7.

The carbon accounting is where the headline number gets ahead of the evidence. The figure most likely to circulate is 6.39 tonnes of CO2 removed per hectare, but it holds only if the dissolved weathering products are carried downstream and re-equilibrate with the alkaline marine carbonate system 7. Within the field boundary, the authors estimate 3.2 to 4.9 tonnes of CO2 per hectare, a range that accounts for how dissolved carbon specates in an acidic soil solution 7. The organic carbon gain, equivalent to 18.3 tonnes of CO2 per hectare, is a separate measured stock change, and the study does not establish what caused it 7.

The mechanism, briefly

Enhanced weathering rests on a simple reaction. Silicate minerals react with carbonic acid — CO2 dissolved in water — releasing metal cations plus bicarbonate. The bicarbonate is soluble, eventually reaching the ocean, where it can precipitate as carbonate and hold carbon on geological timescales. The reaction is limited by mineral surface area, which is why grinding rock into fine powder is the intervention.

Glacial rock flour is what you get when an ice sheet does the grinding. It is ultra-fine sediment formed beneath glaciers. In Greenland, meltwater carries roughly 1.28 gigatonnes of suspended sediment into the ocean each year 5. That material contains high concentrations of silicate and trace metals including iron and manganese 5.

Work on a glacierised catchment in SW Spitsbergen found that sediment-bound iron and silicon yields were several times higher than the dissolved fluxes of the same elements, and that the style of chemical weathering varied sharply between different subglacial outflows from the same glacier. Sulphide oxidation and carbonate dissolution dominated in a channelised outlet; silicate mineral weathering mattered more in smaller outflows 6. Rock flour is not uniform. What it does depends on what was ground up and how the water moved through it.

What the trial actually did

The experiment applied 50 tonnes of Greenlandic glacial rock flour per hectare to a sandy agricultural soil in Denmark and followed it for five years 7. Fifty tonnes per hectare is 5 kilograms per square metre. That is a lot of material to move, spread and pay for, and any practical assessment has to include those costs.

The authors used a soil mass balance approach within a randomized block design. The abstract describes this as the first implementation of mass balance methods in that framework 7. That cuts both ways: the approach is new to this setting, which is a reason for interest and caution at once. A single well-designed trial is not a replicated body of evidence.

Over five years, an estimated 56 percent of the applied flour dissolved 7. The authors describe this as substantial weathering under field conditions in a temperate climate. It remains one site, one soil, one source of rock flour, one climate.

Two carbon pools, two different kinds of claim

Inorganic carbon. The weathering reaction consumes CO2. Accounting for carbonate speciation in an acidic soil solution — where some dissolved carbon can degas rather than stay as bicarbonate — the authors estimate 3.2 to 4.9 tonnes of CO2 per hectare captured in the field 7. The higher figure of 6.39 tonnes per hectare requires assuming the weathering products are transported downstream and re-equilibrate with the more alkaline marine system 7. That assumption is standard in enhanced weathering accounting and not unreasonable. It is also not something this experiment measured. It is a downstream calculation attached to a field measurement.

Organic carbon. Treated plots showed a significant increase in soil organic carbon stocks, equivalent to 18.3 tonnes of CO2 per hectare 7. That is larger than either inorganic figure. The gains were primarily in the mineral-associated organic matter fraction 7.

This is the most interesting result and the one with the least explanation. Mineral-associated organic matter is generally the more persistent fraction of soil carbon, so finding gains there rather than in coarse particulate matter matters for whether the carbon sticks. But the abstract reports where the carbon ended up, not why it arrived. Several mechanisms could produce the pattern. Faster plant growth on treated plots would add carbon. Fresh mineral surfaces could bind dissolved organic matter that was previously mobile. Microbial processing could shift. The reported result does not distinguish between these.

So the honest reading is narrow: treated plots ended five years with more carbon in a particular soil fraction than control plots, at one site. The study does not show that adding rock flour builds stable soil carbon by a specific route.

A caution about the direction of downstream effects

The marine half of the calculation deserves more scrutiny than it usually gets, because glacial material reaching coastal waters does not automatically mean more carbon drawn down. A review of Arctic glacier fjords concluded that whether freshwater and sediment discharge increases or decreases marine primary production depends on the glacier type — marine- or land-terminating — and on which resource limits phytoplankton growth in that particular place and season: light, macronutrients or micronutrients 3. Different fjords show different discharge-productivity relationships, and the review argues multiple case studies are needed to get at net effects 3.

So the assumption that weathered material exported from a field ends up as a net carbon sink in the ocean is a simplification. It may hold in aggregate. It is not settled fact, and the 6.39 figure inherits that uncertainty.

Practical reading for growers

If someone is selling rock dust on the strength of this paper, the useful questions are: which number are they quoting, the in-field range or the one with the marine assumption? What is their evidence that carbon moved into a persistent fraction in their soil rather than someone else's? And what is the mass balance on haulage, given that 50 tonnes per hectare is 5 kilograms per square metre?

The other thing to watch is what accumulates. The flour carries iron, manganese and other trace metals 5. The abstract of this trial does not report trace metal monitoring in the soil after five years 7, and one application is not the same as repeated ones.

Where the evidence runs out

Replication. One trial, one temperate site, one sandy soil, one rock flour source, five years. Nothing here establishes how the numbers change on a clay soil, in a wetter or drier climate, or with a different rock dust. Glacier-derived flour from Greenland has a specific mineralogy. Basalt dust, the more common agricultural product, does not behave identically, and the weathering style within a single glacier catchment already varies between outflows 6.

Duration. Five years is long by the standards of this literature and short by the standards of soil carbon. The 44 percent of applied flour that had not dissolved by year five has an unknown trajectory. The organic carbon gain could continue, plateau or reverse. The abstract offers no projection either way.

Mechanism. The cause of the organic carbon increase is not identified. That is the single largest gap in the paper's most striking result.

Soil chemistry and trace metals. Not reported in the abstract 7. Repeated applications of a material with high trace metal content 5 raise questions the trial does not address.

The marine step. The 6.39 figure depends on downstream re-equilibration and on assumptions about marine productivity responses that vary by system 3.

Permanence. The mineral-associated fraction is the better bet for persistence, but five years of stock change at one site is not proof that the carbon stays put.

None of this makes the result small. A 56 percent dissolution rate measured in the field over five years is a useful constraint, and the organic carbon result is worth following up precisely because nobody yet knows why it happened. But the honest summary is that this is one good experiment, not a settled carbon removal rate — and the number it will most often be quoted for is the one that depends on the ocean.

References

  1. [1]The Iron Biogeochemical Cycle Past and PresentRobert Raiswell, Donald E. Canfield · 2012 · Geochemical Perspectivesdoi:10.7185/geochempersp.1.1
  2. [2]Fjords as Aquatic Critical Zones (ACZs)Thomas S. Bianchi, Sandra Arndt, William E.N. Austin, Douglas I. Benn, Sebastien Bertrand, Xingqian Cui · 2020 · Earth-Science Reviewsdoi:10.1016/j.earscirev.2020.103145
  3. [3]Review article: How does glacier discharge affect marine biogeochemistry and primary production in the Arctic?Mark J. Hopwood, Dustin Carroll, Thorben Dunse, Andy Hodson, Johnna M. Holding, José L. Iriarte · 2020 · The Cryospheredoi:10.5194/tc-14-1347-2020
  4. [4]Principal role of fungi in soil carbon stabilization during early pedogenesis in the high ArcticJuan Carlos Trejos-Espeleta, Juan P. Marin-Jaramillo, Steven K. Schmidt, Pacifica Sommers, James A. Bradley, William D. Orsi · 2024 · Proceedings of the National Academy of Sciencesdoi:10.1073/pnas.2402689121
  5. [5]Trace metal distributions in the transition zone from the Greenland Ice Sheet to the surface water in Kangerlussuaq fjord (67° N)Clara R. Vives, Jørgen Bendtsen, Rasmus D. Dahms, Niels Daugbjerg, Kristina V. Larsen, Minik T. Rosing · 2025 · The Cryospheredoi:10.5194/tc-19-3107-2025
  6. [6]Controls of sediment-bound and dissolved nutrient transport from a glacierised metasedimentary catchment in the high ArcticLukasz Stachnik, Jon Hawkings, Andrea Spolaor, Katarzyna Stachniak, Dariusz Ignatiuk, Sławomir Sitek · 2025 · Chemical Geologydoi:10.1016/j.chemgeo.2025.122940
  7. [7]Enhanced weathering of glacial rock flour drives coupled inorganic and organic carbon sequestration in a five-year field experimentChristiana Dietzen, Malgorzata Rizzi, Tim Jesper Suhrhoff, Minik Thorleif Rosing · 2026 · Biogeosciencesdoi:10.5194/bg-23-7727-2026
  8. [8]Groundwater quality dynamics and hydrochemical indicators of water insecurity in the semi-arid Sokoto Basin, Northwestern NigeriaSaadu Umar Wali, Noraliani Binti Alias, Ismail Usman Kaoje, Sa’ad Ibrahim, Abdullahi Umar · 2026 · Scientific Reportsdoi:10.1038/s41598-026-76975-y
  9. [9]Co-application of biochar and poultry manure improves plant water availability and drought tolerance of Ananas comosus in a clayey FerralsolGustavo Marques Vianna Querino, Rannah Raquel Barreto Silva, Raimundo Barros da Trindade Neto, Francisco Alisson da Silva Xavier, Laercio Duarte Souza, Quirijn de Jong van Lier · 2026 · Discover Soildoi:10.1007/s44378-026-00367-x
  10. [10]Linking plantain derived metabolites in sheep urine with nitrification inhibition in soilMichelle Peterson, Nigel Joyce, John van Klink, Glenn Judson, Trish Fraser, Craig Anderson · 2026 · Nutrient Cycling in Agroecosystemsdoi:10.1007/s10705-026-10542-9
  11. [11]Multi-hazard assessment of the eastern slope of the Sistema de Famatina, La Rioja province (Argentina)Maria Cecilia Corbat, Oscar Juárez, Enrique Fucks, José Augusto Casas · 2026 · Natural Hazardsdoi:10.1007/s11069-026-08492-0
  12. [12]Precise stratigraphic framework of the Bashijiqike Formation of Cretaceous in the northern Kuqa Depression, Tarim Basin and its geological significanceZhu Songbai, Zhang Xianjun, Tang Yongliang, Lin Na, Nie Yanbo, Zhang Zhentao · 2026 · Frontiers in Earth Sciencedoi:10.3389/feart.2026.1920263
  13. [13]From burial to barrier: how burial history controls the hydraulic conductivity in argillaceous formationsRaphael Burchartz, Brian Mutuma Mbui, Peter Achtziger-Zupančič, Garri Gaus, Timo Seemann, Lisa Winhausen · 2026 · Solid Earthdoi:10.5194/se-17-1239-2026
  14. [14]Atmospheric radionuclides accumulation in supraglacial sediments and mosses on two Arctic glaciers in SvalbardKatarzyna Kołtonik, Krzysztof Zawierucha, Jakub Buda, Dariusz Sala, Kamil Wojciechowski, Krzysztof Samolej · 2026 · Polish Polar Researchdoi:10.24425/ppr.2026.8077

All 14 references resolved against Crossref on Nov 6, 2026. Full verification record.

Get the findings, not the noise

A short email when something worth knowing clears the verification gate. No takes. Unsubscribe in one click.