Nighttime-light estimated electricity access in Eswatini

Published: 19 March 2021| Version 2 | DOI: 10.17632/g4t9hrzt2w.2
Wisdom Dlamini


The dataset comprises a set of 450m resolution raster datasets (and accompanying TFW files) of predicted probability of electricity access in Eswatini. These are drived from random forest modelling of Visible Infrared Imaging Radiometer Suite (VIIRS) Day/Night Band (DNB) nighttime light and other anciallry data for the period 2013 to 2019. The value of each pixel idnicates the probability of a household occupying that pixel having access to electricity. The files ending with Q_LOW and Q_HIGH indicate the lower and upper bounds of the prediction interval (95%).