We propose a novel approach for rapid segmentation of flooded buildings by fusing multiresolution, multisensor, and multitemporal satellite imagery in a convolutional neural network. Our model significantly expedites the generation of satellite imagery-based flood maps, crucial for first responders and local authorities in the early stages of flood events. By incorporating multitemporal satellite imagery, our model allows for rapid and accurate post-disaster damage assessment and can be used by governments to better coordinate medium- and long-term financial assistance programs for affected areas. The network consists of multiple streams of encoder-decoder architectures that extract spatiotemporal information from medium-resolution images and spatial information from high-resolution images before fusing the resulting representations into a single medium-resolution segmentation map of flooded buildings. We compare our model to state-of-the-art methods for building footprint segmentation as well as to alternative fusion approaches for the segmentation of flooded buildings and find that our model performs best on both tasks. We also demonstrate that our model produces highly accurate segmentation maps of flooded buildings using only publicly available medium-resolution data instead of significantly more detailed but sparsely available very high-resolution data. We release the first open-source dataset of fully preprocessed and labeled multiresolution, multispectral, and multitemporal satellite images of disaster sites along with our source code.

Single-pair differential synthetic aperture radar interferometry (DInSAR) as well as more advanced methods, such as persistent scatterer interferometry (PSI), allow vertical displacements to be detected at the sub-centimeter level. Since 2014 free SAR data—Sentinel-1—have been collected systematically under the COPERNICUS program at a high temporal resolution and with global coverage. Such an open-access policy greatly helps build a wide user-community and develop diverse SAR-based applications. In this study conventional single-pair DInSAR and the PSI techniques were employed to monitor the vertical displacements of the newly constructed D8 highway, more specifically the part passing through the České Středohoří Mountains, where, during highway construction, a massive landslide occurred in June 2013. For both DInSAR and PSI, free Sentinel-1 radar data were used; moreover, the conventional single-pair DInSAR workflow was processed using freely available SNAP software. Results from the radar interferometry were validated using in situ techniques, such as geodetic measurements, 3D inclinometers, and laser scanning. Both approaches proved their ability to achieve reliable results in detecting vertical displacement “hotspots”. Additionally, in terms of absolute values, the PSI interferometry corresponds very well with the in situ measurements. This study also shows that open-source solutions (free data and SW) provided under the COPERNICUS program bring a great potential for monitoring vertical displacements.