Authors: Yang Ju*, University of California - Berkeley, Iryna Dronova, University of California - Berkeley
Topics: Remote Sensing, Urban and Regional Planning, Europe
Keywords: Nighttime light, time-series, urbanization, classification
Session Type: Paper
Start / End Time: 10:00 AM / 11:40 AM
Room: Lafayette, Marriott, River Tower Elevators, 41st Floor
Presentation File: No File Uploaded
Time-series stack of nighttime light (NTL) images have been increasingly used to analyze urbanization processes. However, few studies have focused on places under drastic socioeconomic shifts. In this study, we identified urbanization trends from pixel-based DMSP-OLS v4 NTL time-series trajectories in Ukraine between 1992 and 2013, during which the country experienced major changes characterized by economic decline before 2000 and fluctuations after. We extracted NTL time-series trajectories from 1 km by 1 km image pixels, and classified the trajectories based on their statistical parameters to identify main urbanization trends. Particularly, the analysis was only conducted in areas with urban and built-up landcover to reduce the scattering effects of DMSP-OLS v4 NTL images, and to capture built-up areas with limited electrification. We found six trends between 1992 and 2013, with four showing some level of overall decline in NTL intensity and one with mostly no light. Many time-series trajectories also tended to decline before 2000 and then increase slightly, indicating 2000 being a key change point of Ukraine’s urbanization process. We will carry out similar classification procedure for the period of 1992-2000 and 2000-2013 separately to contrast urbanization trends before and after 2000. Through this study, we will produce a spatially explicit dataset of Ukraine's urbanization process that can support other urban applications. The analysis framework can also be extended to other regions under major socioeconomic changes.