Within the realm of deep learning techniques, numerous remote sensing applications can be effectively addressed using deep learning algorithms. However, there is a scarcity of studies in synthetic Aperture radar (SAR)-based urban flood mapping involving deep learning techniques, primarily due to two reasons. First, SAR-based urban flood mapping is inherently rooted in change detection, resulting in a complex multi-modality problem within the imbalance data. This complexity arises from the integration of SAR intensity, InSAR coherence, and even SAR phase information acquired from different polarizations (i.e., VV and VH polarization in Sentinel-1 data) both before and after the event. The second challenge is the absence of a benchmark dataset specifically designed for SAR-based urban flood mapping. In an effort to fill this gap, a benchmark dataset for large-scale flood mapping using Sentinel-1 data, which includes not only SAR intensity but also InSAR coherence, should be created. The SAR pre-processing should be carefully checked at the very beginning. With this aim, we tested the specking filter and kernel size selection for the SAR preprocessing for deep learning models. Through this initiative, we found that despeckling SAR intensity and selecting the kernel size in InSAR coherence calculation do not significantly affect the accuracy in deep learning-based urban flood mapping using Sentinel-1 data. The curated benchmark dataset will be presented in the final paper.

The issue of sun glint poses a significant challenge for ocean remote sensing with high-resolution ocean drone imagery, as it contaminates images and obstructs crucial features in shallow-waters, leading to inaccurate benthic substrates identification. While various physics-based statistical solutions have been proposed to address this optical issue in remote sensing, there is a lack of sun glint detection and removal methods specifically designed for high-resolution consumer-grade drone RGB imagery. In this paper, we present a pioneering pipeline for sun glint detection and removal in high-resolution drone RGB images, aiming to restore the real features that are hindered by sun glint. Our approach involves the development of a Foreground Attention-based Semantic Segmentation Network (FANet) for accurate and precise sun glint detection, while effective sun glint removal is achieved through pixel propagation using an optical flow field. Experimental results demonstrate the effectiveness of our FANet in identifying sun glint, achieving IoU accuracy of 81.34% for sun glint pixels and 99.52% for non-sun glint background pixels. Furthermore, the quantitative evaluation of sun glint removal using two well-known metrics show that our method outperforms the GAN-based image restoration method (DeepFillv2) and the conventional image interpolation method (Fast Marching Method, hereafter referred to as FMM). Thus, our pipeline lays the foundation for accurate and precise marine coastal ecological monitoring and seafloor topographic mapping using consumer-grade drone at a low cost.

Due to the complexity of urban environments, the synthetic aperture radar (SAR) based mapping of floodwater is impacted by different factors such as water depth, building orientation and the density of built-up areas. Several studies have proven that both SAR multitemporal intensity and interferometric SAR (InSAR) coherence data acquired in VV and VH polarizations support the urban flood mapping. We propose a deep learning (DL) based method using dual-polarization Sentinel-1 multitemporal intensity and coherence data combined with prior information to map floodwater in urbanized areas. The proposed method aims at mapping flooded areas in urbanized regions and bare soils/sparsely vegetated areas within the entire frame of a Sentinel-1 image. In this paper, our method is evaluated for the Houston (US) urban flood event in 2017 via a qualitative and quantitative comparison with two established DL models. The proposed method has the lowest number of false alarms in flooded urban areas, indicating that the prior information from the probabilistic urban mask is valuable.

Due to the complexity of backscattering mechanisms in built-up areas, the synthetic aperture radar (SAR)-based mapping of floodwater in urban areas remains challenging. Open areas affected by flooding have low backscatter due to the specular reflection of calm water surfaces. Floodwater within built-up areas leads to double-bounce effects, the complexity of which depends on the configuration of floodwater concerning the facades of the surrounding buildings. Hence, it has been shown that the analysis of interferometric SAR coherence reduces the underdetection of floods in urbanized areas. Moreover, the high potential of deep convolutional neural networks for advancing SAR-based flood mapping is widely acknowledged. Therefore, we introduce an urban-aware U-Net model using dual-polarization Sentinel-1 multitemporal intensity and coherence data to map the extent of flooding in urban environments. It usesa prioriinformation (i.e., an SAR-derived probabilistic urban mask) in the proposed urban-aware module, consisting of channel-wise attention and urban-aware normalization submodules to calibrate features and improve the final predictions. In this study, Sentinel-1 single-look complex data acquired over four study sites from three continents have been considered. The qualitative evaluation and quantitative analysis have been carried out using six urban flood cases. A comparison with previous methods reveals a significant enhancement in the accuracy of urban flood mapping: the F1 score of flooded urban increased from 0.3 to 0.6 with few false alarms in urban area using our method. Experimental results indicate that the proposed model trained with limited datasets has strong potential for near-real-time urban flood mapping.

Operational services in the fields of flood monitoring and prevention are benefitting from the large scale and systematic availability of synthetic aperture radar (SAR) data. The main advantages of SAR data are that they provide synoptic views over wide areas, day and night and all-weather condition acquisitions and a reliable data acquisition schedule. Satellite SAR data availability has increased over the past few years due to renewed efforts of several space agencies to put in place new satellite constellations. The latter enable the reduction of the satellite time access to areas of interest and provide enriched information with increased spatial resolution as well as variable polarizations and frequencies. The current situation tells us that there are regions in the world and land cover classes where SAR-derived flood maps are very reliable and accurate, but others where uncertainty is still very high, or where SAR is even unable to provide flood extent information. Therefore, the aim of this paper is to provide an overall picture of SAR-based floodwater mapping algorithms and their suitability for operational applications.

Due to the similarity of the radar backscatter in flooded and unflooded conditions over particular areas, it is not possible to carry out a comprehensive SAR-based flood mapping at large scale. In this paper, an additional information layer derived from Sentinel-l time series data, called Exclusion map (EX-map), is introduced. Its aim is to enhance and complement the results of automatic change detection-based flood mapping methods. The EX-map aims at delineating areas where observed variations of SAR backscatter do not allow detecting the appearance of floodwater. The EX-map is mainly composed of the following land cover classes: topographic shadow/layover, double bounce and smooth tarmac in urban areas, arid areas, dense vegetation and permanent water bodies. The method is evaluated over six study sites across the globe and tested for different flood events. The EX-map not only increases the classification accuracy of change detection-based flood maps derived from Sentinel-l data from 95.92% to 97.02%, but also enables a better interpretation of any SAR-based floodwater map.

The main objective of this study is to introduce and evaluate a SAR-based flood mapping algorithm enabling the automatic generation of a large-scale flood record from the ENVISAT ASAR data archive. The flood mapping algorithm is based on a change detection approach and requires an automatic selection of optimal reference images. The flood mapping algorithm is applied to selected pairs of images to sequentially generate a record of flood extent maps. False alarms caused by water-like areas are reduced using auxiliary data sources such as the Height Above Nearest Drainage (HAND) index derived from topography data. The proposed method is applied to several ENVISAT WS ASAR datasets acquired over the UK and results are validated with a flood extent map derived from aerial photography. Results presented in this paper demonstrate the effectiveness of the methodology.