Spatio-Temporal Variability Estimations of Street Lighting Places Using Night-Time Lights

Quantitative assessments of spatial and temporal distributions of street lighting across larger areas are difficult to make and express in such a form that would be appropriate as model inputs. This paper examines the density of street lighting places based on Earth Observing (EO) data in four urban areas of the Czech Republic. Previously processed GIS data of the night-time lights and planar point patterns of street lighting places were utilised. First, spatial measures of central tendency and dispersion were used to investigate the centrality and directionality of the point patterns. Second, density-based methods were conducted to assess the local and global densities of the patterns and measure the tendency to depart from Complete Spatial Randomness (CSR). Finally, the dependence of the point densities on the night-time light radiance was investigated by quadrat counting. The results of the spatial measures confirm strong directional patterns in smaller areas with sparse street lighting networks. Conditional Monte Carlo tests confirm that the point patterns deviate from CRS. Lastly, the density estimations based on night-time light radiance revealed that the higher density of points is accompanied by higher values of the radiance.