Failure-Aware Explainable Decision Support for Circular Sulfate Adsorbent Design under Hydrochemical Domain Shift

This study develops a failure-aware, explainable environmental decision-support framework for circular sulfate adsorbent design and asks whether predictive transportability and engineering usefulness remain aligned under hydrochemical domain shift. A 58-sample Semarang groundwater dataset served as the development domain. Random forest achieved internal AUROC 0.710 ± 0.165, and a target-free matrix-complexity index (MCI) combined EC, Ca²⁺, Mg²⁺, Cl⁻, and HCO₃⁻ into four treatment regimes. Using the fixed Semarang model, discrimination was retained in the 25-sample Bayat basin (AUROC 0.794; balanced accuracy 0.762) but was lost in a larger cross-country Mekong challenge (n = 118; AUROC 0.491; 95% CI 0.367–0.617). In contrast, the fixed-reference MCI remained strongly associated with sulfate in Mekong (rs = 0.699; 95% CI 0.587–0.782), with 73 of 78 high-sulfate events in the complex-matrix domain. To assess consistency between the engineering interpretation and independent material evidence, open-access sulfate-adsorption studies were synthesized at study level. Zirconium modification increased reported pomelo-peel-biochar qmax from 1.02 to 35.21 mg/g (34.5-fold), while CaCl₂-modified kelp carbon retained substantial reported kinetic uptake in lake water (569.37 mg/g versus 624.39 mg/g in simulated water). The combined evidence supports a workflow in which external model validity governs predictive use, whereas transparent hydrochemical regimes and independent experimental evidence guide capacity, selectivity, stability, and functionalization priorities.

Authors

Publication Details

Journal
Applied and Computational Engineering
Published
2026-11-03
DOI
https://doi.org/10.54254/2755-2721/2027.ma38438
Primary Topic
Adsorption and biosorption for pollutant removal
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Failure-Aware Explainable Decision Support for Circular Sulfate Adsorbent Design under Hydrochemical Domain Shift

Sophia Chen
Applied and Computational Engineering
Adsorption and biosorption for pollutant removal
article

Failure-Aware Explainable Decision Support for Circular Sulfate Adsorbent Design under Hydrochemical Domain Shift

Sophia Chen
article en

Abstract

This study develops a failure-aware, explainable environmental decision-support framework for circular sulfate adsorbent design and asks whether predictive transportability and engineering usefulness remain aligned under hydrochemical domain shift. A 58-sample Semarang groundwater dataset served as the development domain. Random forest achieved internal AUROC 0.710 ± 0.165, and a target-free matrix-complexity index (MCI) combined EC, Ca²⁺, Mg²⁺, Cl⁻, and HCO₃⁻ into four treatment regimes. Using the fixed Semarang model, discrimination was retained in the 25-sample Bayat basin (AUROC 0.794; balanced accuracy 0.762) but was lost in a larger cross-country Mekong challenge (n = 118; AUROC 0.491; 95% CI 0.367–0.617). In contrast, the fixed-reference MCI remained strongly associated with sulfate in Mekong (rs = 0.699; 95% CI 0.587–0.782), with 73 of 78 high-sulfate events in the complex-matrix domain. To assess consistency between the engineering interpretation and independent material evidence, open-access sulfate-adsorption studies were synthesized at study level. Zirconium modification increased reported pomelo-peel-biochar qmax from 1.02 to 35.21 mg/g (34.5-fold), while CaCl₂-modified kelp carbon retained substantial reported kinetic uptake in lake water (569.37 mg/g versus 624.39 mg/g in simulated water). The combined evidence supports a workflow in which external model validity governs predictive use, whereas transparent hydrochemical regimes and independent experimental evidence guide capacity, selectivity, stability, and functionalization priorities.

Applied and Computational EngineeringVol. 270(1)
Openalex Percentile: Top 26%
Adsorption and biosorption for pollutant removal
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon