This repository contains the datasets, analysis scripts, and supporting functions used in the manuscript "Separating Environmental Controls and Source Intermittency in Legacy-Well Methane Leakage: Insights from Dual CH₄–CO₂ Surface Plumes" by Pullen & Cahill. The repository supports a high-resolution field investigation of methane (CH₄) and carbon dioxide (CO₂) leakage from an integrity-compromised legacy gas well in northeastern British Columbia. It includes continuous surface flux measurements, repeated spatial flux surveys, and co-located soil and meteorological observations used to disentangle environmental modulation of surface emissions from intrinsic, intermittent subsurface leakage behaviour. Specifically, the repository provides: Processed CH₄ and CO₂ flux datasets from continuous hotspot monitoring and spatial surveys Soil and atmospheric datasets used in multivariate and machine-learning analyses Scripts and functions for flux calculation, quality control, temporal alignment, spatial interpolation, and footprint metric derivation Statistical and machine-learning workflows (PCA, multiple linear regression, and tree-based models) used to quantify environmental controls and identify source-driven emission events The materials are intended to support transparency, reproducibility, and reuse of the analysis framework for studies of legacy well leakage, fugitive gas emissions, and surface monitoring strategies relevant to abandoned wells and carbon capture and storage (CCS) applications.