Truace
Subscribe

Explainable machine learning-driven targeted modification of biochar and its remediation performance for saline-alkaline soil

Biochar is a low-cost, eco-friendly candidate for saline-alkali soil remediation. Its practical efficacy, however, is constrained by high alkalinity, an underdeveloped pore structure, and limited surface functional sites, all of which conventional modification methods have yet to adequately address. To bridge this gap, this study developed an explainable machine learning prediction system that integrates multi-source data spanning biochar preparation parameters, physicochemical properties, and remediation outcom…

By the Truvace desk2 min read1 sourcePeer-reviewedTRV-2026-1507· cite & verify
Explainable machine learning-driven targeted modification of biochar and its remediation performance for saline-alkaline soil
Long-term biodegradation of aged saline-alkali oily sludge with the addition of bulking agents and microbial agents by Wang, Shijie; Wang, Xiang. CC BY 4.0 · https://creativecommons.org/licenses/by/4.0

In brief

Researchers built an explainable machine learning system that links how biochar is prepared, its physical and chemical properties, and how well it remediates saline-alkali soil. SHAP analysis was used to derive recommended parameter ranges, which guided a phosphoric acid-chitosan composite modification.

The result matters because biochar is a low-cost option for degraded soils but is limited by high alkalinity and weak pore structure. The model-guided material lowered soil pH over 14 days, though the evidence reported covers one modified biochar and a short test period.

Main points

  1. The model integrated biochar preparation parameters, physicochemical properties, and remediation outcomes to enable reverse design of modification processes.
  2. Feature importance ranking put electrical conductivity first, followed by pH, specific surface area, and cation exchange capacity.
  3. Recommended ranges included pyrolysis temperature of 600-800 °C and an addition ratio of 2%-10%.
  4. The modified biochar had SSA of 21 m2/g, O/C of 0.31, and EC of 625 μS/cm.

The gain

Machine-learning-guided biochar modification cut saline-alkali soil pH from 10.26 to 9.28 in 14 days, while unmodified biochar left it near 10.2.

The rundown

The study starts from a known limitation: biochar is a low-cost, eco-friendly candidate for saline-alkali remediation, but high alkalinity, underdeveloped pores, and limited surface functional sites constrain its efficacy, and conventional modification methods have not adequately addressed these issues.

From SHAP importance and dependence results, the authors derived recommended ranges for several parameters, including particle size of 10-70 mesh, CEC of 80-273 cmol/kg, and O/C ratio of 0.36-0.56. Electrical conductivity was the most important feature in the ranking.

Based on these results, they proposed a phosphoric acid-chitosan composite modification. The product had an enriched pore structure and abundant oxygen-containing functional groups. Its SSA and O/C ratio fell outside the recommended ranges, but its EC was within the ideal range, which the authors take as confirming EC as a core indicator.

Sources

  1. Peer-reviewedEnvironmental Research2026-09-28

The debate