Reproducibility data and code for Hyperspectral Patterns and Neural Networks for Exploring Diffuse Methane Emissions at a Landfill

Reproducibility materials for an exploratory landfill hyperspectral study from the 14 April 2026 campaign at Los Barrios, Spain. The study groups surface spectra before consulting external references and evaluates whether neural networks approximate an expanded spectral-reference rule from narrower inputs. TDLAS measurements provide methane context, not pixel labels. The package includes extracted samples, group centres, digitized reference curves with provenance, native-resolution masks, campaign records, image headers, fixed training/validation/test designs, numerical code, weights, predictions and independent verification records. It contains 112 main and secondary neural runs and 24 additional SWIR neural runs using a 3 m geographic exclusion on the same test zones, plus logistic and direct controls. Mean MLP AUC is 0.917 at 1 m and 0.915 at 3 m. Results measure spectral-rule learning, not validated methane detection. Compact inputs support the documented recalculations and neural retraining. Full-scene extraction and clustering require ten complete processed reflectance products (69,748,089,728 bytes), inventoried but not included; original detector recordings are not supplied. The repository also includes Technical Documentation with implementation settings, complete counts, paired results by zone and all scene maps. The README distinguishes these reproducibility levels and lists the verification commands. External digitized curves and mineral data retain their source terms. This package is prepared for manuscript review; no blanket public redistribution licence has been assigned.