Long-term normalized difference urban index (NDUI) data time series for urban studies

arxiv(2023)

Cited 0|Views19
No score
Abstract
Keeping continuous, long-term data to examine changes in urban surroundings is crucial as cities expand and develop. The DMSP OLS nighttime lights data and the Landsat NDVI were used to create the Normalized Difference Urbanization Index (NDUI), which has proven to be an invaluable resource for studying urban areas. However, DMSP's reach and usefulness are constrained by the fact that data collecting ended in 2014 while VIIRS has continued to collect the nighttime lights data since 2012. The unavailability of DMSP translates to a challenge in performing urban studies using the NDUI. In this work, we address this difficulty and suggest a novel approach to bringing the NDUI time series up to date. We first map the VIIRS to DMSP using 2012 as a calibration year and then construct an updated NDUI time series. ClimateDownscaleSuite is used and Swin Transformer is selected as the best model for the mapping. The Swin Transformer model and the sophisticated machine learning capabilities it offers are used in conjunction with the VIIRS evening lighting data collected after 2012. By using this strategy, not only is the NDUI time series extended, but the potential of AI in filling in data gaps and boosting urban studies is also highlighted.
More
Translated text
AI Read Science
Must-Reading Tree
Example
Generate MRT to find the research sequence of this paper
Chat Paper
Summary is being generated by the instructions you defined