Data, Source Data, and Code Repository: Global collision of avian biodiversity and power transmission grids

Data, Source Data, and Code Repository: Global collision of avian biodiversity and power transmission grids Overview & Study ScopeThis repository contains the complete curated empirical datasets, tabular Figure Source Data, phylogenetic trees, machine-learning attribution matrices, 2050 Net-Zero supergrid routing layers, high-resolution figures, and canonical Python scripts accompanying the manuscript "Global collision of avian biodiversity and power transmission grids" submitted to Nature (anonymized in accordance with Nature's double-blind peer review policy). The empirical framework integrates1. Global Power Infrastructure: 3,678,243 mapped overhead power-line segments (13.52 million km total), disaggregated into 1,819,344 high-voltage transmission (>70 kV) and 1,858,899 medium-voltage distribution (10–70 kV) segments intersected against 78 seasonal and resident range geometries across N = 36 focal avian species (11 taxonomic orders).2. Curated Biodiversity Occurrences: n = 53,336 precision-filtered field occurrences curated from 616,495 raw decadal GBIF records (2016–2025) via the six-tier MI-STEP ([M]-[I]-[S]-[T]-[E]-[P]) spatial thinning and deduplication pipeline, paired with exact geodesic point-to-line distances.3. Evolutionary & Morphological Data: Biometric dual-hazard classifications (adult body mass ≥ 2.0 kg and wingspan ≥ 1.50 m) evaluated across 100 posterior time-calibrated Hackett phylogenetic trees (Pagel's lambda, Blomberg's K, and 9,999-permutation Mantel tests).4. Machine-Learning Attribution Matrix: Observer-bias-controlled XGBoost and exact TreeSHAP modelling dataset comprising n = 92,366 terrestrial cells (52,070 presences and 40,296 background points) across macro-climatic, topographic, and built-up surface predictors.5. 2050 Net-Zero Supergrid Routing: 113 simulated Least-Cost Path (LCP) ultra-high-voltage (UHV) transmission corridors connecting 50 high-irradiance solar supply hubs (GHI ≥ 4.25 kWh m⁻² d⁻¹) to 50 SSP2 metropolitan demand centres, alongside a 10-km global routing friction surface. Repository Architecture (6 Directories, 86 Files + Root Manifest with SHA-256 Checksums)- 00_Verification_Master/: Single-file, zero-hardcode automated forensic verification pipeline (master_verification_pipeline.py), audit results (master_audit_results.json), and complete audit dossier (MASTER_DATA_AUDIT_DOSSIER.md).- 01_Source_Data_Figures/: Complete tabular Source Data (.xlsx and .csv) underlying all 6 Main Figures (Figs. 1–6), 3 Extended Data Figures (Extended Data Figs. 1–3), and Extended Data Table 1.- 02_Curated_Empirical_Datasets/: Processed datasets organized by study objective: - Obj1_Assessment_Gap_and_Open_Ranges/: 36-species morphological trait roster, IUCN Red List Threat 4.2 cache, Track B HV/MV range-intersection summary table, and Track A open-access parametric range proxy geometries (.gpkg). - Obj2_Occurrences_and_Proximity/: 53,336 MI-STEP curated occurrence records (.csv and .gpkg), point-to-line distances, and 6-tier attrition summary. - Obj3_Phylogeny_and_Traits/: 100 posterior BirdTree Nexus phylogenies (.nex), AVONET trait database, and phylogenetic signal statistics. - Obj4_Environmental_ML_Matrix/: Final XGBoost/TreeSHAP modelling matrix (Objective4_Final_Dataset.csv) and presence/background covariate tables. - Obj5_NetZero_2050_Supergrid/: 2050 UHV corridor geometries (Future_NetZero_Grid_2050.geojson), 10-km friction surface (.tif), solar supply hubs, megacity demand sinks, and species-level future exposure risk table.- 03_Python_Scripts/: All 16 canonical production Python scripts organized into 6 workflow modules (01_Data_Curation_MISTEP through 06_Obj5_NetZero_2050_Supergrid) plus figure compilation utilities.- 04_Publication_Figures_300DPI/: Publication-ready figures (300 DPI PNG, vector PDF, and RGB JPG) for Main Figures 1–6 and Extended Data Figures 1–3.- 05_Reference_Bibliography/: Complete bibliography (reference.md, 41 citations with verified DOIs). Software Requirements & Computational ReproducibilityAll spatial, phylogenetic, machine-learning, and statistical workflows were executed in Python (3.11+) using GeoPandas (≥0.14), pyogrio (≥0.7), Shapely (2.0), rasterio (≥1.3), SciPy (≥1.11), Biopython (1.83), DendroPy (4.6), scikit-learn (≥1.4), XGBoost (≥2.0), SHAP (≥0.45), scikit-image (≥0.22), and Matplotlib (≥3.8). Reviewers can independently verify all statistical claims, sample sizes, and figure source tables by running:python 00_Verification_Master/master_verification_pipeline.py Data Provenance, Analyzed Outputs & Third-Party Copyright ComplianceAll custom code, analyzed datasets, and figure source data tables in this repository are released under the Creative Commons Attribution 4.0 International (CC-BY 4.0) license. For species distribution ranges evaluated from BirdLife International (Version 2025.1, https://datazone.birdlife.org), all analyzed spatial metrics, voltage-disaggregated intersection lengths, and species-level summary tables (Track B) are provided in full within this repository. In accordance with BirdLife International's intellectual property and non-redistribution terms, the raw third-party GIS polygon files are not re-hosted here and may be requested directly from BirdLife DataZone (http://datazone.birdlife.org/species/spcdownload). To guarantee immediate, out-of-the-box spatial code execution without proprietary restrictions, this repository also provides open-access parametric range geometries (Track A, .gpkg, CC-BY 4.0), which exhibit strong spatial concordance with the Track B baseline (log10 grid density Pearson r = 0.8090, P = 2.382 × 10⁻⁹; total intersecting length Spearman rho = 0.8569, P = 2.438 × 10⁻¹¹). Primary planetary datasets are publicly accessible from their permanent repositories: AVONET (https://doi.org/10.6084/m9.figshare.16586228), GridFinder/OSM (https://doi.org/10.5281/zenodo.3628142), CHELSA v2.1 (https://doi.org/10.16904/envidat.228.v2.1), ESA WorldCover v200 (https://doi.org/10.5281/zenodo.7254221), and JRC GHSL R2023A (https://ghsl.jrc.ec.europa.eu).