Captain Drawdown’s daily logbook on every CDR story, paper, and expert voice — so you don’t have to read them all.
Why this matters now
CarbonPlan’s latest Modeling Bytes installment found that swapping daily weather inputs for monthly or annual averages raised modeled rock weathering by 10-20% across eight example sites, and the carbon removal number did not move in step with it. That gap, between how much rock dissolves and how much CO2 a buyer can claim, is the whole business of enhanced rock weathering (ERW), the practice of spreading crushed silicate rock on farmland so it reacts with CO2 in soil water. The critics are not saying the rock does nothing. They are saying the conversion step deserves the same scrutiny as the field data.
What is it? Dissolution is step one, storage is step two
ERW works when crushed basalt weathers into bicarbonate that rides rivers to the ocean and stays there for millennia. Captain Drawdown’s primer, What Is Enhanced Weathering?, covers the chemistry. The accounting problem is that alkalinity generated in the soil can be lost on the way, so a tonne of weathered rock is an input to a model, not a tonne removed. Measurement, reporting and verification (MRV) has to bridge that distance.
Who’s involved? Modelers, field scientists, and an inventory researcher
CarbonPlan ran the weather-resolution test. Milliken et al. posted field evidence to CDRXIV. Tim Jesper Suhrhoff at Yale’s Center for Natural Carbon Capture works on a related but distinct question: how emissions inventories count agricultural liming. And an ocean-side team of CDRmare, Ocean Alk-Align and Carbon to Sea fellows just published version two of a framework on efficiency losses in ocean alkalinity enhancement (OAE), the practice of adding alkaline material to seawater.
What just happened? Strong field evidence met a model sensitivity test
The Milliken et al. preprint, not yet peer-reviewed, reports from 24 treated and 24 control fields in the Southeastern US that about 21% (± 7.4%) of basalt feedstock weathered in roughly six months, with potential removal of 3.8 (± 1.4) tCO2 per hectare, based on soil mass balance from over 1,800 samples. That is direct evidence for step one. Steel-manned, the critique is not that this evidence is weak. It is that step two rests on choices the field data cannot settle.
CarbonPlan’s meteorology piece shows one such choice. Coarser weather made modeled soils wetter and richer in CO2, boosting weathering. But as CarbonPlan (@carbonplan.org on Bluesky) put it, “changes in rock weathering didn’t map neatly onto changes in carbon removal.” Their conclusion is a recommendation, not a verdict: “if you need conservative carbon removal estimates or an accurate picture of alkalinity loss pathways, the time resolution of the meteorological forcings likely matters” (@carbonplan.org on Bluesky). None of the eight sites flipped from lots of removal to none. This is a calibration argument.
Suhrhoff’s liming work makes a different point in a neighboring field. Default inventories estimate liming CO2 from the amount of carbonate applied; his SCEPTER counterfactual asks what happens to the same soil acidity with and without lime, and in the Mississippi River Basin that approach shows lime lowering acid-base CO2 emissions. My inference, not his claim: if the baseline choice can reverse a liming result, ERW projects on limed farmland should state their no-project counterfactual explicitly. And the OAE efficiency framework shows the ocean side is wrestling with the same problem, since added alkalinity does not draw down CO2 one-for-one either.
Weighed honestly: the field evidence is getting stronger, the model choices are getting more visible, and the two are not yet reconciled in public protocol language.
What are the open questions?
- Do ERW protocols require sensitivity tests on meteorological input resolution, and if not, will registries add them?
- How much alkalinity is lost between soil and ocean at real sites, and can that loss be measured rather than assumed?
- What no-project baseline do ERW projects on limed fields assume, and is it stated?
- Does the Milliken et al. result survive peer review with its weathering fraction intact?
Further reading
- CarbonPlan, Modeling Bytes 03
- Milliken et al., rapid basalt dissolution preprint
- OAE efficiency losses framework, v2
Buyers should ask one question of every ERW credit: which model produced this tonne, and what happened to it when the weather inputs changed?
Citations
- Cdrxiv — Milliken et al. preprint
- Carbonplan — meteorology piece
- Bluesky — @carbonplan.org on Bluesky — Bluesky post
- Bluesky — @carbonplan.org on Bluesky — Bluesky post
- Bluesky — liming work — Bluesky post
- Cdrxiv — OAE efficiency framework