Quebec Contaminated-Site Screening: Groundwater-Well Proximity, Site Geometry, and Decision Uncertainty — Data and Reproducibility Code

Description This open research dataset and reproducibility archive supports the manuscript “Screening Incomplete Contaminated-Site Inventories: Groundwater-Receptor Proximity, Spatial Representation, and Decision Uncertainty in Quebec, Canada,” prepared for the Journal of Environmental Management. The study investigates how spatial representation influences preliminary contaminated-site screening against registered groundwater locations. It compares nearest-well distances calculated from published representative site coordinates with those calculated from mapped site geometries, including polygons, lines and multipoint features. The archive supports reproducible assessment of screening classification disagreements and spatial-data quality limitations. Quebec contaminated-site analysis The archived government datasets, downloaded on 26 October 2026, include 11,502 unique contaminated-site records and 263,735 registered groundwater wells. The analysis identifies 989 sites represented by non-point geometries and uses a defined single-region subset for regional comparisons. Comparing representative points with full registered geometries produces 105 additional screening flags at 100 m, 61 at 500 m and 33 at 1,000 m among eligible single-region records. At 1,000 m, the analysis identifies 33 geometry-only classifications, highlighting the potential for representative coordinates to omit proximity relationships captured by the complete registered geometry. The archive also includes a conservative geometric-distance bound analysis for identifying records whose screening classification can be established without an exact geometry-to-well distance calculation. This procedure uses established geometric principles and is not presented as a newly invented distance algorithm. United States external illustration An additional application examines 1,292 paired U.S. Environmental Protection Agency National Priorities List sites using EPA point locations and mapped site boundaries, together with U.S. Geological Survey groundwater-location records. At a 1,000 m screening radius, 153 sites are newly flagged using full boundaries, while one point-based flag is removed. The U.S. analysis provides an external illustration of representation-sensitive screening. Differences between the Quebec and U.S. inventory structures, point definitions and groundwater-location data mean that these results should not be interpreted as equivalent environmental-risk estimates. Archive contents The deposited ZIP includes: Archived Quebec contaminated-site and groundwater-well GeoPackages. U.S. EPA National Priorities List coordinates and boundary geometries. USGS groundwater-location data. Python scripts for spatial screening, distance validation, geometric-bound calculations and figure generation. Derived screening classifications, tabulated outputs and figures. Data dictionaries, provenance records, software requirements, checksums and reproducibility instructions. Reproducibility and limitations The Quebec analysis uses EPSG:32198 and the U.S. analysis uses EPSG:5070, with distances reported in metres. The archived source snapshots should be used to reproduce the reported findings because government inventories may change over time. Screening flags indicate geographic proximity only. They do not establish contaminant transport, groundwater contamination, confirmed drinking-water exposure, regulatory exceedances or remediation priority. The screening distances are exploratory rather than regulatory thresholds. Citation and attribution Version: 1.0 Creators: Usman Haider, Asif Ali, Muhammad Humayon, Wesley J. Johnston and Hina Iqbal. DOI: https://doi.org/10.5281/zenodo.23751731 Data sources: Données Québec, U.S. Environmental Protection Agency and U.S. Geological Survey. The archive distinguishes MIT-licensed analysis code from source datasets and other materials governed by their respective licensing and attribution conditions.