A Methane-constrained Virtual Sensing Framework for Field Diagnosis of Sewer Sediment Organic Loading

This record contains the computational reproducibility materials for the study “A Methane-constrained Virtual Sensing Framework for Field Diagnosis of Sewer Sediment Organic Loading”. The package includes: (1) processed and reconstructed modelling input tables and field-application reference measurements; (2) Python code for H-CatBoost and CatBoost modelling, bootstrap prediction intervals, and Desilting event-level evaluation; and (3) numerical outputs comprising pointwise predictions, prediction intervals, and system-level and event-level performance metrics. The README.md file describes the package structure, required software environment, input variables, execution sequence, and expected outputs. The deposited tables are intended to reproduce the reported computational workflow and numerical outputs.