Skip to content

Module Overview⚓︎


Henderson, Vernon, Adam Storeygard, and David N. Weil. "A bright idea for measuring economic growth." American Economic Review 101, no. 3 (2011): 194-199.

Asher, Sam, Tobias Lunt, Ryu Matsuura, and Paul Novosad. "Development research at high geographic resolution: an analysis of night-lights, firms, and poverty in India using the shrug open data platform." The World Bank Economic Review 35, no. 4 (2021): 845-871.
  title = {{A Bright Idea for Mesuring Economic Growth}},
  author  = {Henderson, J. Vernon and Storeygard, Adam and Weil, David N.},
  journal = {American Economic Review},
  year = {2011}

  title={Development research at high geographic resolution: an analysis of night-lights, firms, and poverty in India using the shrug open data platform},
  author={Asher, Sam and Lunt, Tobias and Matsuura, Ryu and Novosad, Paul},
  journal={The World Bank Economic Review},
  publisher={Oxford University Press}
About Night lights are widely used as a proxy for some form of electrification or economic activity when time series data on economic activity is otherwise unavailable. Gridded night lights data are aggregated and calibrated for consistent time series estimation. Available annually from 1992–2021 across DMSP and VIIRS.
Geographic Coverage National
Aggregations Shrid, District, Subdistrict, Village/Town
Producer Colorado School of Mines, Earth Observatory Group
Source URL
  • Night lights from 1994 to 2013 are from the DMSP-OLS annual measures of night time luminosity, measured at 1/120 degree. dmsp_total light variable is the sum of the luminosity values (0–63) of all pixels in the unit. dmsp_total_light_cal variable is the same value, but calibrated for consistent estimation across time using the method of Elvidge et al. (2014).
  • Average pixel luminosity in a geographic unit is calculated by dividing one of the total light variables by dmsp_num_cells.
  • For each geographic aggregation level, we provide values from 2 different annualized VIIRS night lights rasters per geographic unit and year combination - average-masked and median-masked. The masked rasters zero out all background, biomass burning and aurora. The masked average takes the daily average pixel value and the masked median takes the daily median pixel value. It is upto the discretion of the user to choose which values suit their needs best based on their specific use-case. We suggest reading the original VIIRS paper for more specific details about how data were annualized and refer to the official website as well.

    Release Details⚓︎

    Release Number 2.0
    Release Name pakora
    Last updated June 27, 2023

    See anything amiss? Let us know!