Using night lights satellite imagery to estimate spatial economic change in Johannesburg between 2011 and 2021

Night lights imagery is captured using a sensor on a satellite. The sensor is able to detect visible and infrared light to obtain quantitative measurements of light emissions at night. Light emissions data, in conjunction with supporting datasets, can be used to predict economic activity at fine spatial scales over time. There is a growing body of academic scholarship on how light emissions can be used to measure economic activity, and in some countries the approach has evolved to the point where economic metrics from night lights are now well-known indicators.

This work has been done as a proof of concept to show that night lights can be a valuable tool to help model spatial economic trends in the South African context, where there has historically been a dearth of reliable, regularly-updated and fine-grained spatial economic datasets. The work was done on request from, and in partnership with, the City of Johannesburg’s Group Strategy, Policy, Coordination and Relations Unit after officials there noted similar work being undertaken in the eThekwini Metropolitan Municipality. An earlier version of this paper was delivered to the City in January 2023.

Using Johannesburg as a case study, this project employs open data, predominantly in the form of free-to-use satellite imagery on night lights, to investigate the spatial spread of economic activity across the City of Johannesburg, and to examine changes in that activity over time. Night lights data from the Visible Infrared Imaging Radiometer Suite (VIIRS) sensor, collected at a spatial resolution of 500m, and for the period 2011 to 2021, is melded with Landscan population data and Gross Value Added (GVA) estimates in CSIR mesozones to develop and calibrate a model that can accurately predict GVA from changing light intensity.

The maps and trends analysis produced from this work model spatial economic activity to a scale of 1km². The analysis of pixels representing 1km² is then scaled up to Johannesburg’s City Planning Regions and suburbs. Averages for various points in time over a 10-year period (2011–2021) have been obtained, enabling analysis of change.

The analysis suggests that the COVID-19 pandemic and its associated lockdowns, as well as increasing loadshedding – South Africa’s strategy of limiting overdemand for electricity on a fragile and constrained power grid in danger of collapse if demand exceeds supply – has had a severe impact on economic activity in Johannesburg. Estimated GVA values show that while there has been slow but steady economic growth between 2011 and 2019, the following two years (2020–2021) saw significant decline that effectively eroded the eight years of growth before that. Two of Johannesburg’s City Planning Regions (Regions B and C) are now in a worse position than they were in 2011. However, within this broad picture, the more detailed analysis at the 1km2 and suburb level scales show significant variegation, with some parts of regions growing, others declining, and some remaining largely static.

The modelling and analysis that has been attempted in this project could help policy-makers better fill gaps in understanding where no economic data is available. By visualising what changes are taking place, where and at what pace – and in turn seeking to explain the ‘why’ behind these changes – we hope that this work will help to inform better economic development policies in data poor contexts.

Suggested citation: Naidoo Y., Maree, G., Naidoo, L. and Götz, G. (2024). Using night lights satellite imagery to estimate spatial economic change in Johannesburg between 2011 and 2021. GCRO Rapid Research Paper. Gauteng City-Region Observatory (GCRO). Johannesburg. https://doi.org/10.36634/JBDZ2536

Also see this Map of the Month, where the method developed in this Rapid Research Paper has been used to analyse spatial economic change across Gauteng.