Abstract
Near realtime flood mapping in densely populated urban areas is critical for emergency response. The strong heterogeneity of urban areas poses a big challenge for accurate near realtime flood mapping. However, previous studies on automatic methods for urban flood mapping perform infeasible in near realtime or fail to generalize well to other floods, for several reasons. First, multitemporal pixel-wise flood mapping requires accurate image registration, hindering the efficiency of large-scale processing. Although automatic image registration has been investigated, precisely coregistered multitemporal image sequence requires time-consuming fine tuning. Additionally, the floods may lead to the loss of many corresponding image points across multitemporal images for accurate coregistration. Second, existing unsupervised methods generally rely on hand-crafted features for floodwater detection. Such features may not well represent the patterns of floodwaters in different areas due to inconsistent weather conditions, illumination, and floodwater spectra. This article proposes a self-supervised learning framework for patch-wise urban flood mapping using bitemporal multispectral satellite imagery. Patch-wise change vector analysis is used with patch features learned through a self-supervised autoencoder to produce patch-wise change maps showing potentially flood-affected areas. Postprocessing including spectral and spatial filtering is applied to these patch-wise change maps to remove nonflood related changes. Final flood maps and parameter sensitivities were evaluated using several performance metrics. Two flood events from areas with differing degrees of urbanization were considered: Hurricane Harvey flood (2017) in Houston, Texas, and Hurricane Florence flood (2018) in Lumberton, North Carolina. The proposed method shows strong performance for self-supervised urban flood mapping.
Highlights
T HROUGHOUT the history of human civilization, floods have brought catastrophe to human settlements, including huge losses of life and property
After removing building pixels with building footprints, some of initial changed patches were further classified as NF patches, in which no floodwater pixels were found based on the results of the twoclass K-Means clustering
We propose a fully automated patch-wise urban flood extent mapping method via a self-supervised learning framework
Summary
T HROUGHOUT the history of human civilization, floods have brought catastrophe to human settlements, including huge losses of life and property. The United Nations (UN) has set the goal to rapidly and accurately respond to upcoming floods for protecting vulnerable people and mitigating economic losses, as stated in the UN Sustainable Development Goal 11 (2015– 2030) [3] To help meet this goal, improved methods for realtime flood extent mapping over dense urban regions to support flood response efforts are needed. Du et al [10] and Tong et al [11] proposed improved particle swarm optimization methods for endmember extraction, which have great potential for subpixel flood mapping. These aforementioned flood mapping studies, have focused on rural areas with relatively homogeneous image backgrounds. Flood extent mapping is insufficiently investigated in urban areas due to heterogeneous land cover and land use, low spatial resolution of MS imagery, and lack of flood extent ground truth datasets [12], [13]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
