Regulatory provenance and event-selection sensitivity in reported startup NOx emissions

Description

This dataset contains submission-stage reproducibility materials for a bounded 2023 Rhode Island analysis of regulatory provenance and startup-event screening. The study comprises five facilities, ten reporting units, 87,600 unit-hours, and 1,257 operational startup events. Submitted provenance fields are preserved losslessly for auditability, while Policy C uses a frozen NOXR-level classification that distinguishes resolvable non-substitution, verified regulatory substitution, and unresolved records. The archive contains derived event and policy outputs, source identifiers and checksums, Tables 1–4 and Figures 1–3 source data, portable production code, and quality-control materials. Original EPA source bytes are not redistributed. The archive characterizes reported-data handling and does not estimate true/corrected emissions or physical measurement error.

Steps to reproduce

Extract the archive and run `python code/run_all.py` to reproduce packaged numerical checks, table sources, and figures. Full raw-source Gate 2C reconstruction additionally requires separately obtained official EPA files identified in SOURCE_MANIFEST.csv; use the optional `--raw-source-root` workflow documented in README.md.

Institutions

• Kocaeli Üniversitesi

  Kocaeli, İzmit

Categories

Licence