Reconstructing soil acidity neutralization curves using Machine learning and chemical or spectral soil signatures
Why this work is in the frame
A frame that forgets how it found something cannot be audited. These are the routes that admitted this work.
Bibliographic record
Abstract
• Machine learning accurately reconstructs full soil titration curves using Ca(OH) 2 and standard soil inputs. • Models achieve R 2 > 0.92 and RMSE < 0.21 pH units, ensuring precision across soil types. • Dynamic lime prescriptions are derived from fitted titration curves and allow target-specific flexibility. • Spectral (SSM) and hybrid (HM) models offer rapid, reagent-free, and scalable alternatives to SMP. • A tiered deployment strategy enables labs to adopt models based on existing analytical capacities. Soil acidity management often relies on lime recommendation methods that are imprecise, time-consuming, or involve hazardous reagents like the SMP buffer solution. This study introduces an alternative approach by developing and evaluating machine learning (ML) models to predict the change in soil pH (ΔpH) following incremental applications of hydrated lime (Ca(OH) 2 ). A total of 418 soil samples from Eastern North America were analyzed for their chemical properties, mid-infrared (MIR) spectral signatures, and complete titration curves obtained through acid-base neutralization. Three ML models were tested: a Chemical Signature Model (CSM) based on routine soil analyses, a Spectral Signature Model (SSM) relying solely on MIR spectra, and a Hybrid Model (HM) combining both data sources. All models demonstrated high accuracy, achieving R 2 values above 92 % and RMSE values below 0.21 pH units. The HM achieved the highest performance (R 2 = 94 %, RMSE = 0.18), closely followed by the SSM, indicating the practical equivalence of the two approaches since converting ΔpH curves into absolute pH curves always requires the initial soil pH. SHapley Additive exPlanations ( SHAP) values were used to interpret variable importance in each model. In the CSM, lime dose and initial pH were dominant predictors, followed by organic matter, Mehlich-3 extractable Ca (Ca M3 ), and Al (Al M3 ). In the SSM and HM models, specific MIR spectral regions corresponding to hydroxyl, carboxylic, silicate, and organo-mineral functional groups were highly informative, confirming consistency with known soil chemistry principles. These findings enable the automated reconstruction of titration curves, paving the way for dynamic, accurate, and safe lime recommendation systems tailored to laboratory capabilities: CSM for immediate implementation and SSM or HM for laboratories adopting MIR spectroscopy. This approach aligns with precision agriculture principles, supporting sustainable and site-specific management of soil acidity.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
Full frame distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.000 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.000 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it