Urban Flood Detection with Sentinel-1 Multi-Temporal Synthetic Aperture Radar (SAR) Observations in a Bayesian Framework: A Case Study for Hurricane Matthew

Lin Lin,Bhardwaj Bhardwaj,Hill Hill,Yun Yun

doi:10.3390/rs11151778

Abstract

In this study we explored the application of synthetic aperture radar (SAR) intensity time series for urban flood detection. Our test case was the flood in Lumberton, North Carolina, USA, caused by the landfall of Hurricane Matthew on 8 October 2016, for which airborne imagery—taken on the same day as the SAR overpass—is available for validation of our technique. To map the flood, we first carried out normalization of the SAR intensity observations, based on the statistics from the time series, and then construct a Bayesian probability function for intensity decrease (due to specular reflection of the signal) and intensity increase (due to double bounce) cases separately. We then formed a flood probability map, which we used to create our preferred flood extent map using a global cutoff probability of 0.5. Our flood map in the urban area showed a complicated mosaicking pattern of pixels showing SAR intensity decrease, pixels showing intensity increase, and pixels without significant intensity changes. Our approach shows improved performance when compared with global thresholding on log intensity ratios, as the time series-based normalization has accounted for a certain level of spatial variation by considering the different history for each pixel. This resulted in improved performance for urban and vegetated regions. We identified smooth surfaces, like asphalt roads, and SAR shadows as the major sources of underprediction, and aquatic plants and soil moisture changes were the major sources of overprediction.

Highlights

Flood extent maps based on synthetic aperture radar (SAR) have increasingly been used in recent emergency response operations
Since we only had 1 flooded epoch in the whole time series, we proposed to use the statistics from the spatial domain, which we can estimate from histogram fitting of the normalized during-event backscattering (σ0k) using the Levenberg–Marquardt algorithm [30] as suggested in [29,31]
In the urban area (Figure 6b), we saw that a large proportion of the urban floods were mapped by the criterion of pU ≥ 0.5, indicating that these pixels experienced an intensity increase during the event epoch as compared with other epochs in their own history, and the increase was significant enough such that the probability of being flooded was over 50%

Summary

Introduction

Flood extent maps based on synthetic aperture radar (SAR) have increasingly been used in recent emergency response operations. Jaxa.jp), where national space agencies, research institutions, and end-user organizations work together on emergency observation requests in the Asia-Pacific region, the responses to 20 of 23 activated flood-related events in the year of 2017 used ALOS-2 SAR flood-mapping results [1]. These statistics demonstrate the need for radar’s all-weather, day-and-night sensing capability, where in most cases cloud cover and rains persist for the duration of a flood. The low intensity within the shadow may lead to false flood detection if one uses a during-flood SAR image alone

Methods

Results

Discussion

Conclusion