India District Nighttime Lights (VIIRS)
District-level nighttime lights panel for India (641 districts, 2012–2024) from VIIRS satellite data, with annual radiance statistics and an automated pipeline
github.comThis repository provides an automated Python pipeline and ready-to-use panel dataset tracking satellite-observed nighttime luminosity across all 641 Indian districts (Census 2011 boundaries) from 2012 to 2024.
Built using NOAA / NASA Suomi NPP VIIRS Day/Night Band (DNB) annual median radiance composites via Google Earth Engine, the dataset includes:
- Zonal Statistics: Mean, median, sum, standard deviation, minimum, and maximum nighttime radiance (in nW/cm²/sr) per district-year, alongside valid pixel counts and log-transformed values (
log1p_mean,log1p_median). - Data Formats: An 8,333-row flat CSV panel (
output/csv/nightlights_district_panel.csv) for immediate econometric analysis in Stata, R, Python, or DuckDB, plus annual GeoJSON layers for spatial mapping in QGIS or Kepler.gl. - Reproducible Pipeline: An open-source workflow enabling researchers to authenticate with Earth Engine and customize boundary inputs, year ranges, or composite aggregations.
Nighttime luminosity serves as a widely utilized empirical proxy for subnational economic growth, electrification progress, and local urban density where high-frequency district GDP figures are unavailable.
Details
- Kind
- Data source
- Run by
- Individual
- Topics
- Where
- India
- Licence
- Open licence
- Cost
- Free
- Licence terms
- MIT