A harmonized inventory of ground-mounted photovoltaic installations in India, 2019–2022: processing pipeline

Abstract

This repository contains the processing pipeline used to build the dataset "A harmonized inventory of ground-mounted photovoltaic installations in India, 2019–2022." The pipeline is a staged, CPU-only workflow (ingest → standardise → conflate → reconcile → enrich → validate → export) that fuses three openly licensed source datasets into a single deduplicated, attribute-enriched inventory of ground-mounted PV installations, with per-record source provenance and a licence-verification step. To reproduce: create the environment from environment.yml, then run the pipeline end to end following the instructions in README.md. All input datasets, their versions, and their licences are listed in sources_manifest.yaml. The dataset produced by this pipeline is archived separately at https://doi.org/10.5281/zenodo.21867919