City nighttime light variations using ISS images

摘要

Earlier studies utilizing coarse resolution DMSP-OLS nighttime light imagery suggest a correlation between the amount of night light and poverty. The International Space Station (ISS) night light images offer higher resolutions enabling analysis at lower disaggregation levels - particularly of intra-urban variations. The aim of this study is to examine the capacity of ISS images for analyzing intra-urban night light differences compared to DMSP-OLS and whether these differences are correlated to variations in land use and socio-economic characteristics of residential areas. The two cities analyzed are the East African city Dar es Salaam, in Tanzania, and the Indian megacity Mumbai. We apply correlation analysis to the ISS night light images and classification results of VHR imagery, land use maps as well as built-up and population densities. The results show that intra-urban differences are related to the presence of main transport lines and industrial areas, which have the highest night light values, in contrast to deprived areas (e.g. slums), which are the relatively dark spots of the city. Furthermore, differences between Dar es Salaam and Mumbai are also clearly visible. Dar es Salaam's deprived areas in general have lower night light values than those in Mumbai. However, deprived areas in Mumbai show variations in night light values where large slum areas are often poorly illuminated. Our findings confirm that ISS data are a more relevant data source than DMSP-OLS data to generate base indicators for quality of life, environmental and deprivation studies, due to their higher resolution and public availability, but contextual knowledge of the specific city is needed for reliable interpretation.