(9/9) Overall Discussion and Conclusion

We recommend starting with the intro in Part 1 of this series.

EC can be a serious MRV proxy — if you treat it like a calibrated instrument, not a shortcut

Our results underline a simple point: Total alkalinity (TA) is central for quantifying CO₂ removal in enhanced weathering via leachate, but measuring TA frequently enough is hard.

At the dataset-wide level, we found strong correlations between TA and:

• EC

• Ca²⁺

• Mg²⁺

This supports the idea that simpler measurements can carry the alkalinity signal — at least under stable conditions — and can make MRV more scalable. But the same dataset exposes the core caveat: Proxy relationships that are valid “in general” can fail in specific soil–feedstock combinations or at certain times/events.

When stratifying the data, the EC–TA relationship often deviated from the overall trend — and sometimes disappeared entirely. Even Ca²⁺ and Mg²⁺, which are fundamentally tied to alkalinity formation, were not perfectly coherent with TA in every treatment at micro-scale.

This is likely driven by the complexity of soil systems: soils can store, exchange, and transform ions in ways that change what ends up in leachate.

What we can say confidently (based on this dataset)

EC is a strong proxy candidate at macro-scale

After excluding disturbance phases, EC correlates with TA at r ~0.95. That is encouraging because EC can be measured cheaply and continuously.

pH is not a quantitative MRV proxy for CDR

pH is important chemistry, but it does not track alkalinity export reliably enough for quantification.

Disturbances can completely decouple EC from TA

Initial soil disturbance (setup) and fertilizer pulses can generate large EC shifts without corresponding alkalinity changes. EC sensors are therefore valuable as anomaly detectors, but calibration must be revisited after such events.

Site specificity dominates micro-scale proxy performance

Some soil–feedstock combinations show robust EC–TA relationships. Others do not. That variability means: If EC is used as a proxy, it must be calibrated and validated for each site and over time, with continued cross-checks against TA measurements.

The practical MRV strategy that falls out of this

In a field deployment (where you cannot titrate everything all the time), the most defensible approach is hybrid:

1. Use low-cost EC sensors to collect high-frequency or even continuous data in situ.

2. Use periodic TA measurements to anchor and recalibrate the EC–TA relationship (now you can build a continuous TA data set from the continuous EC data).

3. Optionally include periodic Ca²⁺/Mg²⁺ to cross-check geochemical consistency and help diagnose anomalies.

4. Treat EC spikes as a prompt to investigate: fertilization, rain flush, irrigation changes, sensor drift, or other interventions.

EC fills the temporal gaps; TA is the quality control that keeps the method honest.

Limits of what we can claim from this study

We explicitly ran a greenhouse mesocosm experiment, the following aspects of course affect our results which might make it harder to compare the results with outdoor field data/results:

• no prolonged dry/wet cycles

• controlled temperature (>19°C year-round)

• regular watering

• forced leachate collection at a defined outlet depth

• German soils only

• limited feedstock set (mostly (ultra)mafic silicates plus selected industrial materials)

Field deployments add hydrology complexity, crop-cycles, management variability, and climatic extremes. So the macro-scale correlations we observed provide a baseline expectation, not a universal rule.

We also could not answer two logical next questions from this dataset alone:

• Which soil or feedstock parameters control CDR performance and/or EC–TA stability?

• Which soil processes govern the fate of weathering products (retention, secondary phases, biomass pathways)?

Answering those requires a broader dataset and additional soil analyses (extractions, cation retention pools, secondary carbonate formation, biomass, etc.).

The bottom line

Despite these challenges, integrating low-cost sensors into EW monitoring can significantly improve data resolution and reduce MRV costs.

The strongest conclusion we can support from this work is:

Calibrated EC sensors, combined with intermittent TA measurements, can enable cost-effective, high-frequency tracking of EW-induced alkalinity export — but only when the EC–TA relationship is stable for that specific site, and only with ongoing recalibration and event handling.

Read more…

MRV Proxies for EW? A Guided Tour Through Our Data From Our Two-Year Greenhouse Experiment

• Part 1: Intro

• Part 2: Carbon Removal via Weathering and Treatment Variability

• Part 3: Transient Disturbances: Initial Flush and Fertilizer Event

• Part 4: Proxy Correlations on Aggregate Data (Macro-Scale)

• Part 5: Leachate pH (Macro- and Micro-Scale)

• Part 6: Leachate Electrical Conductivity (Macro-Scale)

• Part 7: EC’s Proxy Performance for Individual Treatments (Micro-Scale)

• Part 8: Substituting TA measurements with EC measurements

• Part 9: Overall Discussion and Conclusion

Download our scientific data report as PDF (Pre-Print)

Download the full PDF companion report (PDF, 4 MB, DOI https://doi.org/10.13140/RG.2.2.23232.39688) which is the reference backbone for the series. The data is available on Github https://github.com/dirkpaessler/carbdown_greenhouse_2023_2024 and via DOI https://doi.org/10.5281/zenodo.18360183.