Enhancing Flood Resilience Analysis Through Night Time Light: A Global Perspective

Flooding affects more people than any other hazard and is becoming increasingly severe. The concept of flood resilience, which focuses on the ability of people to anticipate, prepare for, respond to, and recover from flood events, is gaining increasing attention. However, due to data limitations, it is often challenging to quantify flood resilience, particularly during the post-flood recovery phase. This research investigates the potential of utilizing Night-Time Light (NTL) data to enhance flood resilience analysis on a global scale. By examining 24 significant flood events from 2013 to 2018, this study aims to establish a comprehensive system for assessing flood resilience through NTL data from both the event scale and the grid-scale.

The methodology integrates flood extent mapping using MODIS satellite products for flood detection and the generation of 36 months of cloud-free, seasonally adjusted NTL time series. The research summarizes the different behaviors of NTL before, during, and after floods, and analyzes the causes of these variations. Additionally, it introduces three NTL-based quantitative metrics for measuring flood impact, recovery duration, and after-flood transformation. These metrics were applied to the 24 studied events to evaluate their effectiveness, demonstrating the utility of NTL data in capturing the immediate effects of floods and monitoring long-term recovery. Furthermore, a case study of the August 2016 Louisiana floods in the USA involved a micro-scale grid analysis to examine the relationship between NTL changes and factors such as population, coastal proximity, and economy, with the results validated using multiple vulnerability indices.

The results showed significant variation in recovery periods among the studied flood areas, ranging from 5 to 12 months, and even floods occurring within the same country could have recovery durations differing by as much as 5 months. The grid-scale case study further indicated that NTL decreases at the micro-scale are related to population and economic conditions, with communities having better economic conditions showing a lower probability of NTL decrease, while those with higher populations showing a higher probability of NTL decrease.

This study concludes that NTL data, combined with adequate remote sensing and statistical methods, presents a valuable tool for global flood resilience analysis, addressing data gaps and improving disaster management strategies.