iiinova — where your digital self keeps evolving

Dirk Paessler @dpaessler.bsky.social · 07/10/2026

Remy 0.9 is out 🐀 — my open-source plugin that lets Claude work in shared Google Docs now writes native suggestions: real Accept/Reject buttons, even on comment-only links. Not one character changes without a human click. AI drafts, you decide. github.com/dirkpaessler/remy

110

Dirk Paessler @dpaessler.bsky.social · 06/10/2026

Ich schlage vor das Paper zu lesen, bevor ein „spekulativ“ Urteil gefällt wird.

110

Dirk Paessler @dpaessler.bsky.social · 28/08/2026

It's 𝗢𝗰𝗲𝗮𝗻 𝗣𝗵𝗼𝘁𝗼 𝗙𝗿𝗶𝗱𝗮𝘆 where I invite you to a 15 second ocean meditation with one of my photos or videos. Let's protect the beauty and abundance of our oceans!

030

Dirk Paessler @dpaessler.bsky.social · 27/08/2026

None of this argues against in-situ MRV. It argues for engineering it properly: anchor your gauges, over-provision your sensors, report coverage honestly.

100

Dirk Paessler @dpaessler.bsky.social · 27/08/2026

Sensor survival was the bigger problem. Our soil-CO₂ sensors fell from twenty working channels to a single survivor by early 2025. We checked whether dying sensors could fake our correlations over time. They didn't.

100

Dirk Paessler @dpaessler.bsky.social · 27/08/2026

This last part of our XXL Lysimeter Experiment is about plumbing, and it's not glamorous: keeping measurements trustworthy over years is harder than the chemistry itself. Our on-site rain gauge under-caught by nearly half against the official station. Anchor to an official reference, always.

131

Dirk Paessler @dpaessler.bsky.social · 26/08/2026

And the basalt dose barely moves any of it. Every treatment overlaps control, the same verdict every other instrument in this series has returned. Soil-CO₂ sensors won't count your carbon. But they're a genuine window into the soil as a living system, and that's worth building.

100

Dirk Paessler @dpaessler.bsky.social · 26/08/2026

How much carbon actually exports depends on how much rock dissolves and how much water drains through, and a gas sensor is blind to the water.

100

Dirk Paessler @dpaessler.bsky.social · 26/08/2026

Here's the catch. Soil CO₂ strongly predicts leachate pH, but only weakly tracks the alkalinity that actually leaves the pot. The gas is the cause at the top of the chain.

100

Dirk Paessler @dpaessler.bsky.social · 26/08/2026

Every time a dry spell breaks, the soil wakes up and exhales a burst of CO₂, the Birch effect. We caught it happening five separate times across four years, and it only fires hard when rain and warmth arrive together.

100

Dirk Paessler @dpaessler.bsky.social · 26/08/2026

We logged soil CO₂ daily in all 20 pots for four years. Soil CO₂ ran 5 to 20 times higher than outside air, breathing hard in summer and collapsing in winter. Two things throttle that breath: warmth drives it up, but past about 18°C the soil dries out and the breath falls despite the heat.

100

Dirk Paessler @dpaessler.bsky.social · 26/08/2026

A soil-CO₂ sensor sounds like the ultimate shortcut. Weathering consumes CO₂, so bury a probe and read carbon removal straight out of the ground. Part 5 of our XXL Lysimeter Experiment shows why that's not how it works, and what these sensors are actually good for instead.

111

Dirk Paessler @dpaessler.bsky.social · 21/08/2026

010

Dirk Paessler @dpaessler.bsky.social · 20/08/2026

Next: we test whether soil CO₂ itself carries a weathering fingerprint.

100

Dirk Paessler @dpaessler.bsky.social · 20/08/2026

The practical recipe: bury a 60 cm EC probe for continuous data, keep periodic lab titrations to calibrate it, and treat a sudden EC spike as a reason to investigate, not a reason to claim credits.

100

Dirk Paessler @dpaessler.bsky.social · 20/08/2026

One caveat we're stating plainly: this calibration is specific to our soil and climate. It won't transfer cleanly to another site without re-anchoring against real leachate samples there.

100

Dirk Paessler @dpaessler.bsky.social · 20/08/2026

Depth is not a detail. The 30 cm probe barely works, at r = 0.39. Shallow soil dries and channels water unevenly. Sixty centimeters sits close enough to the draining water to matter.

100

Dirk Paessler @dpaessler.bsky.social · 20/08/2026

Since leachate EC is already a strong proxy for lab alkalinity, the buried probe reaches alkalinity through that chain.

100

Dirk Paessler @dpaessler.bsky.social · 20/08/2026

Qualified yes, at 60 cm depth. We buried 120 low-cost EC and pH probes across our 20 pots. The one at 60 cm tracks leachate EC at r = 0.70 across 355 paired readings, and it holds up in every single dose, not just the high ones.

100

Dirk Paessler @dpaessler.bsky.social · 20/08/2026

Alkalinity is the number we actually care about, and it's slow and expensive: pump the leachate, ship it to a lab, titrate. Part 4 of our XXL Lysimeter Experiment asks a more practical question: can a sensor buried in the soil, logging continuously, stand in for that lab work?

100

Dirk Paessler @dpaessler.bsky.social · 19/08/2026

Takeaway: cheap, continuous sensors carry real signals about weathering flux between expensive lab tests.

100

Dirk Paessler @dpaessler.bsky.social · 19/08/2026

CO₂ peaks in summer, pH dips in anti-phase. That's a warning for anyone doing occasional grab sampling. A July reading and a January reading aren't the same distribution. They're two points on a wave.

100

Dirk Paessler @dpaessler.bsky.social · 19/08/2026

It holds in every pot, control included. This is the soil's seasons at work, not the basalt.

100

Dirk Paessler @dpaessler.bsky.social · 19/08/2026

Twenty sensors, logging daily for four years, trace a clean chain: warmer soil respires more, more respiration means more soil CO₂, more CO₂ makes more carbonic acid, and that acidity is what dissolves minerals and releases the alkalinity we measure.

100

Dirk Paessler @dpaessler.bsky.social · 19/08/2026

Part 2 of our XXL Lysimeter Experiment left the leachate data flat. Part 3 turns to the buried sensors, and the picture changes.

110

Dirk Paessler @dpaessler.bsky.social · 14/08/2026

040

Dirk Paessler @dpaessler.bsky.social · 13/08/2026

This isn't an argument against enhanced weathering. It's an argument against assuming dose-response instead of measuring it. Any MRV approach that credits "more rock, more CDR" without a real control on the real soil is crediting a guess.

100

Dirk Paessler @dpaessler.bsky.social · 13/08/2026

Some weathering never reaches the drainage we sample. Base cations can lock into the soil itself before we ever see them.

100

Dirk Paessler @dpaessler.bsky.social · 13/08/2026

The baseline is large, and part of it isn't weathering at all. It includes a disturbance flush from when the soil was first built and soaked. Four replicate pots isn't much statistical power. A modest but real effect could easily hide inside these error bars.

100

Dirk Paessler @dpaessler.bsky.social · 13/08/2026

We tested this properly at four separate timepoints. No dose was significantly different from control at any of them. Why the signal hides:

100

Dirk Paessler @dpaessler.bsky.social · 13/08/2026

Every treatment's error bars overlap the control. Whatever the basalt is doing here, it's smaller than the pot-to-pot noise of the soil itself.

100

Dirk Paessler @dpaessler.bsky.social · 13/08/2026

At day 1,011, the last point with all four replicate pots intact, our unamended control sat at 1.58 tCO₂e/ha. That's in the middle of the pack, level with 200 t/ha, above both 100 and 400 t/ha.

100

Dirk Paessler @dpaessler.bsky.social · 13/08/2026

Does more basalt mean more CO₂ removal? Our XXL Lysimeter Experiment says: not so fast. The intuition is a straight line. Spread twice as much rock, remove twice as much carbon. Over 1,400 days of our XXL Lysimeter Experiment, that line does not hold.

130

Dirk Paessler @dpaessler.bsky.social · 12/08/2026

Turns out we picked the wrong soil-feedstock pairing for this site. A parallel greenhouse trial confirmed it. The upside: our buried sensors found something better. Daily CO₂ data showing how the soil "breathes" a seasonal cycle that drives the whole weathering process.

100

Dirk Paessler @dpaessler.bsky.social · 12/08/2026

At the macro scale, the data is beautiful and reproducible. At the scale that matters, proving more rock removes more CO₂, the signal disappears.

100

Dirk Paessler @dpaessler.bsky.social · 12/08/2026

We ran one of the longest, most detailed enhanced-weathering field experiments out there: daily sensor logs, monthly water chemistry, four replicates per treatment.

100

Dirk Paessler @dpaessler.bsky.social · 12/08/2026

1,400 days. Four winters. Twenty pots of basalt-amended soil. That's our XXL Lysimeter Experiment, and the treated pots removed no more CO₂ than the untreated controls.

212

Dirk Paessler @dpaessler.bsky.social · 07/08/2026

020

Dirk Paessler @dpaessler.bsky.social · 06/08/2026

The result is ERW projects that are cheaper to start, cheaper to verify, and easier to finance. AEROC doesn't compete with project developers. It makes them more effective. That's the kind of infrastructure the ERW industry needs right now.

100

Dirk Paessler @dpaessler.bsky.social · 06/08/2026

AEROC fixes both. Using hyperspectral satellite imagery and machine learning, they help ERW project developers identify the best sites worldwide at up to 95% lower cost, and reduce the soil samples needed for MRV by up to 60%.

110

Dirk Paessler @dpaessler.bsky.social · 06/08/2026

Enhanced rock weathering has a problem that doesn't get talked about enough. It's not the science. The science works. It's that finding the right fields is slow and expensive, and verifying that the rock did its job can eat up nearly half of total project costs.

100