This material was jointly developed by the DECDI ieConnect team and the World Bank Data Lab as an introduction to using nighttime lights for economic analysis.
Nighttime lights have become a widely used data sources, including in the social sciencies literature. Henderson, Storeygard, and Weil's seminal 2012 paper illustrated the use of leveraging nighttime lights to measure economic growth. Their paper helped launch the use of nighttime lights in a variety of applications; a Google scholar search of "nighttime lights economics" brings over 40,000 responses. In addition to leverage nighttime lights as a proxy for economic activity, nighttime lights has been used for various applications such as tracking urbanization and examining impacts of natural disasters, conflict, and infrastructure improvements.
This course provides an overview of using nighttime lights data, with a focus for economic applications. It covers the different sources of nighttime lights, how to query and aggregate data, and addressing data quality with nighttime lights (eg, cloud cover). The course focuses on NASA Black Marble data, using the BlackMarbleR (for R) and BlackMarblePy (for Python) packages for querying data.
The course assumes familiarity with R or Python. For an introduction to these programming languages, see the DIME Analytics R training and the DIME Analytics and DECID Python training.
- Introduction to Spatial Analysis [R | Python coming later!]: Overview of working with vector and raster spatial data in R.
- Nighttime Lights for Economic Analysis [PDF]: Overview of nighttime light datasets and use of nighttime lights for economic and social science analysis.
- Nighttime Lights Analysis in R [R | Python coming later!]: Provides of overview of querying and analyzing nighttime lights data in R.
In addition to providing training content, this repository contains code to quickly (1) produce nighttime lights data for any country (at the ADM0 - ADM3 level, and at the city level) and (2) produces analysis of nighttime lights (e.g., trends and maps). This code file is intended as a start to nighttime lights analysis for a country; the code can then be adapt for further analysis.
For more information, see here.
Spatial analysis in R
- Spatial Data Science with R and "terra"
- Spatial Statistics for Data Science: Theory and Practice with R
Nighttime lights
- World Bank Open Nighttime Lights tutorial
- Spatial Edge: Downloading and processing Black Marble nightlights data in R
- Blog about BlackMarbleR/Py
World Bank Data Partnership examples using nighttime lights