Satellite sensing night-time lights-A South African spatial econometric application

This study investigated how economic activity in one location may affect economic activities in adjacent locations. This study applies several statistical models to determine the economic activity effect of neighbouring locations within the South African context during the period 2012 to 2019, using data from satellite imagery. Decentralised gross domestic product (GDP) per location maps, i.e. to generate a gross domestic dataset at a very granular level was therefore developed. The level is a one km(2) representation per location. The gross domestic product for each location is estimated using the night-time light remote sensing data derived from the Suomi National Polar Partnership (SNPP) infrared imaging satellite radiometer suite (VIIRS). This study illustrates how spatial correlations can be determined, and how to measure them using explanatory spatial data analysis. The essence of the spatial analysis revealed that "space does indeed matter", as economic activity in one location is related to activity in neighbouring locations. Not only do spatial relationships exist, but economic activity seems to be fundamentally related to its own unique location.