Urban Flood Mapping Research Articles

A Systematic Review of Urban Flood Susceptibility Mapping: Remote Sensing, Machine Learning, and Other Modeling Approaches

Climate change has led to an increase in global temperature and frequent intense precipitation, resulting in a rise in severe and intense urban flooding worldwide. This growing threat is exacerbated by rapid urbanization, impervious surface expansion, and overwhelmed drainage systems, particularly in urban regions. As urban flooding becomes more catastrophic and causes significant environmental and property damage, there is an urgent need to understand and address urban flood susceptibility to mitigate future damage. This review aims to evaluate remote sensing datasets and key parameters influencing urban flood susceptibility and provide a comprehensive overview of the flood causative factors utilized in urban flood susceptibility mapping. This review also highlights the evolution of traditional, data-driven, big data, GISs (geographic information systems), and machine learning approaches and discusses the advantages and limitations of different urban flood mapping approaches. By evaluating the challenges associated with current flood mapping practices, this paper offers insights into future directions for improving urban flood management strategies. Understanding urban flood mapping approaches and identifying a foundation for developing more effective and resilient urban flood management practices will be beneficial for mitigating future urban flood damage.

Enhancing generalizability of data-driven urban flood models by incorporating contextual information

Abstract. Fast urban pluvial flood models are necessary for a range of applications, such as near real-time flood nowcasting or processing large rainfall ensembles for uncertainty analysis. Data-driven models can help overcome the long computational time of traditional flood simulation models, and the state-of-the-art models have shown promising accuracy. Yet the lack of generalizability of data-driven urban pluvial flood models to both unseen rainfall and distinctively different terrain, at the fine resolution required for urban flood mapping, still limits their application. These models usually adopt a patch-based framework to overcome multiple bottlenecks, such as data availability and computational and memory constraints. However, this approach does not incorporate contextual information of the terrain surrounding the small image patch (typically 256 m×256 m). We propose a new deep-learning model that maintains the high-resolution information of the local patch and incorporates a larger surrounding area to increase the visual field of the model with the aim of enhancing the generalizability of data-driven urban pluvial flood models. We trained and tested the model in the city of Zurich (Switzerland), at a spatial resolution of 1 m, for 1 h rainfall events at 5 min temporal resolution. We demonstrate that our model can faithfully represent flood depths for a wide range of rainfall events, with peak rainfall intensities ranging from 42.5 to 161.4 mm h−1. Then, we assessed the model's terrain generalizability in distinct urban settings, namely, Lucerne (Switzerland) and Singapore. The model accurately identifies locations of water accumulation, which constitutes an improvement compared to other deep-learning models. Using transfer learning, the model was successfully retrained in the new cities, requiring only a single rainfall event to adapt the model to new terrains while preserving adaptability across diverse rainfall conditions. Our results indicate that by incorporating contextual terrain information into the local patches, our proposed model effectively simulates high-resolution urban pluvial flood maps, demonstrating applicability across varied terrains and rainfall events.

Urban Flood Risk Assessment and Mapping Using GIS-DEMATEL Method: Case of the Serafa River Watershed, Poland

This paper develops a method integrating Geographic Information Systems (GIS) and the Decision-Making Trials and Evaluation Laboratory (DEMATEL) for the analysis of factors influencing urban flood risk and the identification of flood-prone areas. The method is based on nine selected factors: land use/land cover (LULC: the ratio of built-up areas, the ratio of greenery areas), elevation, slope, population density, distance from the river, soil, Topographic Wetness Index (TWI), and Normalized Difference Vegetation Index (NDVI). The DEMATEL method is used to determine the cause–effect relationship between selected factors, allowing for key criteria and their weights to be determined. LULC and population density were identified as the most important risk factors for urban floods. The method was applied to a case study—the Serafa River watershed (Poland), an urbanized catchment covering housing estates of cities of Kraków and Wieliczka frequently affected by flooding. GIS analysis based on publicly available data using QGIS with weights obtained from DEMATEL identified the vulnerable areas. 45% of the total catchment area was classified as areas with a very high or high level of flood risk. The results match the actual data on inundation incidents that occurred in recent years in this area. The study shows the potential and possibility of using the DEMATEL-GIS method to determine the significance of factors and to designate flood-prone areas.

Classification of compressed full-waveform airborne lidar data

ABSTRACT Airborne laser scanners produce 3D data that can be used for a range of applications, such as urban planning, facility monitoring, flood mapping, and forest management. Additional information on the surveyed area can be obtained from the backscattered waveforms recorded by modern light detection and ranging (lidar) sensors. However, the high-dimensional representation of full-waveform data has hindered progress in its use due to difficulties in processing and storage. This paper develops a quantized convolutional autoencoder network to compress lidar waveform data into a condensed feature representation, resulting in a compression rate of up to 20:1. This, together with height information, is fed into a U-net convolutional neural network that achieves an accuracy of 93.7% on six classes.

Urban flood susceptibility mapping using remote sensing, social sensing and an ensemble machine learning model

Flood susceptibility mapping is crucial for urban disaster management. However, the heterogeneity of urban land use and the complexity of terrain pose challenges to the accuracy and generalizability of flood models. In this paper, we propose a framework named the EMF model for urban flood mapping. Specifically, the social sensing and remote sensing were used for flooding information collection. The XGBoost, Support Vector Classifier (SVC), Multilayer Perceptron (MLP), and Multimodal Deep Learning (MDL), were used as the predictive models, and their results were ensembled using a Random Forest model to produce the final outcome. Results show that the EMF model outperforms the standalone models, with an accuracy of 0.942 on the training set and 0.940 on the testing set. The accuracy of the five models is ranked as EMF > MDL > XGBoost > SVC > MLP. The flood map indicates that flooding has a greater impact on the outskirts and suburbs of the city compared to its central urban areas. Farmland is the most affected land type, making up 54.8 % of the flooded area. Overall, the proposed framework enables rapid and accurate identification of flood-prone areas, providing technical support for managers in formulating effective flood prevention strategies.

An efficient urban flood mapping framework towards disaster response driven by weakly supervised semantic segmentation with decoupled training samples

Despite the proven effectiveness of data-driven deep learning techniques in urban flood mapping, the availability of annotation data remains a critical factor impeding their timeliness in real applications. Recent progress in weakly supervised semantic segmentation (WSSS) presents promising solutions for addressing this limitation. To accomplish prompt and accurate flood mapping in complex urban areas from high-resolution remote sensing images to support disaster management and response, this study makes contributions in three key aspects: weak training data generation, the improvement of WSSS algorithm, and the construction of benchmark datasets. Firstly, we present a novel yet efficient weak training data generation strategy by decoupling the acquisition of positive and negative samples. This strategy enables the rapid generation of block-level weak annotations assisted by pre-flood river data or the segment anything model (SAM) for zero-shot segmentation, thereby alleviating the burden of weak data labeling. Secondly, to enhance the flood detection results in complex urban landscapes based on low-cost weak labels, we design an end-to-end WSSS framework incorporating tree filtering-based structure constraints and a perturbed dual-branch cross self-distillation mechanism. Lastly, to evaluate the performance of the proposed approach, we constructed two large aerial imagery datasets, namely Calgary-Flood and Huston-Flood. These datasets encompass diverse urban land covers and include challenging scenarios with extensive shadows, providing a robust benchmark for evaluating our method against various urban environments. Experimental results demonstrate that our weak data annotation strategy substantially enhances efficiency. Additionally, the proposed WSSS framework exhibits superior performance in comparison to existing state-of-the-art methods, particularly in terms of edge delineation and algorithmic stability. The advancements in weak data annotation strategy, end-to-end model architecture, and benchmark dataset development in this study collectively exploit the potential of the weakly supervised paradigm for rapid flood mapping in urgent situations. The datasets and code will be publicly available at (https://github.com/YJ-He/Flood_Mapping_WSSS).

Urban Flood Mapping Using Satellite Synthetic Aperture Radar Data: A review of characteristics, approaches, and datasets

Urban Flood Mapping Using Satellite Synthetic Aperture Radar Data: A review of characteristics, approaches, and datasets

A rapid high-resolution multi-sensory urban flood mapping framework via DEM upscaling

Urban floods can cause severe loss of economic and social assets, and remote sensing has been an effective tool for flood mapping during disaster response. Due to the complexity of high-density urban structures, high-resolution (HR) optical images can only extract visible floods in open spaces, and floods in shadows and under the canopy are challenging to map. Accurate digital elevation models (DEMs) are essential for inundation estimation towards urban flood mapping, but HR DEMs are often unavailable due to the high acquisition costs. Through DEM upscaling, HR DEMs could be obtained from existing low-resolution (LR) DEMs using deep learning. To this end, a novel multi-sensory HR urban flood mapping framework is proposed in this research. The framework consists of three components: 1) a new DEM upscaling network to infer HR DEMs from existing LR DEMs with a fusion approach, 2) a rapid flood segmentation network to extract visible flood from very-high-resolution (VHR) optical images with limited human labelling, and 3) an accurate Geographical Information System (GIS)-based tool for floodwater extent and depth estimation from the visible flood information along with HR DEMs. The proposed framework was validated on a fluvial flood that occurred in Calgary, Canada, in 2013, where the proposed DEM upscaling network produced an upscaled HR DEM at 2 m resolution from an existing LR DEM at 18 m resolution. In addition, the proposed flood segmentation network has shown accurate visible flood extraction from VHR RGB aerial imagery with over 80% intersection-over-union (IoU) using 10% of human labelling as training samples. Finally, the floodwater extent and floodwater depth estimation using the proposed GIS tool showed significant improvement over conventional flood mapping methods.

A Framework on Fast Mapping of Urban Flood Based on a Multi-Objective Random Forest Model

Fast and accurate prediction of urban flood is of considerable practical importance to mitigate the effects of frequent flood disasters in advance. To improve urban flood prediction efficiency and accuracy, we proposed a framework for fast mapping of urban flood: a coupled model based on physical mechanisms was first constructed, a rainfall-inundation database was generated, and a hybrid flood mapping model was finally proposed using the multi-objective random forest (MORF) method. The results show that the coupled model had good reliability in modelling urban flood, and 48 rainfall-inundation scenarios were then specified. The proposed hybrid MORF model in the framework also demonstrated good performance in predicting inundated depth under the observed and scenario rainfall events. The spatial inundated depths predicted by the MORF model were close to those of the coupled model, with differences typically less than 0.1 m and an average correlation coefficient reaching 0.951. The MORF model, however, achieved a computational speed of 200 times faster than the coupled model. The overall prediction performance of the MORF model was also better than that of the k-nearest neighbor model. Our research provides a novel approach to rapid urban flood mapping and flood early warning.

Combining SAR images with land cover products for rapid urban flood mapping

Synthetic Aperture Radar (SAR) is an indispensable source of data for mapping and monitoring flood hazards, thanks to its ability to image the Earth’s surface in all weather conditions and at all times. Through cloud computing platforms such as Google Earth Engine (GEE), SAR imagery can be used in near-real time for rapid flood mapping. This has facilitated the disaster response community to make informed decisions in flood hazard interventions and management plans. However, rapid urban flood mapping using SAR is challenging, due to the complex land cover configuration in urban environments, coupled with complicated backscattering mechanisms. Here, we propose a novel method to utilise SAR imagery and land use-land cover (LULC) products for rapid urban flood mapping. Our approach uses a Land Cover Product to segment the study area into LULC types and differentiate each type with respect to whether double bounce is expected to occur during the flooding events. The normalised difference index was derived using a multi-temporal SAR image stack, and the threshold segmentation method was adopted for flood mapping. In addition, DEM and Surface Water datasets were employed to refine the flood extraction results using a morphological correction approach. We assessed the method quantitatively using two use cases: the 2017 Houston and 2022 Coraki flood events. Based on fine spatial resolution optical imagery, the proposed method achieved an accuracy of 92.7% for the August 2017 Houston flood mapping task and 89% for the March 2022 Coraki flood mapping task, which not only represents at least 13% in accuracy compared to non-LCP based flood extraction method, but also provides strong capability for rapid flood mapping in urban settings.

An exploratory study of Sentinel-1 SAR for rapid urban flood mapping on Google Earth Engine

Real-time, near-real-time, and accurate flood extent information is critical for emergency response during disaster events such as floods. Accurate extents are critical for disaster management and relief efforts. Despite multiple efforts, there are still many challenges in automated processing of Sentinel-1 SAR to generate reliable inundation maps. The major advantage of SAR compared to optical imagery is its data collection capability despite any weather conditions even thick cloud situation. Currently, there is a knowledge gap of employing different polarization combinations of SAR imagery for flooding research. First, ten different combinations of the two original VH and VV polarizations are designed for rapid and accurate urban flood mapping. To examine the significant potentials of the polarization combinations for flood mapping, four flood mapping methods namely threshold, change detection, unsupervised and supervised classification, in combination with a zero-depth flood method, are designed and used to map flood extents. Among different polarization combinations, the multiplication, squared multiplication, addition, and squared addition combinations have resulted in good results for flood extent mapping. In addition, a flood depth estimation approach has been used to address the overestimation of urban flooded areas. In all four methods, the deduction of overestimated flooded areas using the threshold of zero flood depth has improved the overall accuracy on average 7 % for all methods. The results show that all four methods implemented on Google Earth Engine are good using different combinations to identify flooded areas but change detection method requires little user involvement, and this can be applied to new study areas without estimating flooding depth for the affected areas. Whereas the supervised classification will need more user’s involvement to collect sample points. Among all the combinations, squared addition of polarizations has been consistently performed well for all methods. All the analysis has been done on the Google Earth Engine platform, and this strategy can be used to map flood in any urban environment. The finding of this study will enhance local governments and federal agencies rapid assessment of flooding disasters and making accurate decisions.

Evaluating a new method of remote sensing for flood mapping in the urban and peri-urban areas: Applied to Addis Ababa and the Akaki catchment in Ethiopia

The Sentinel-1 SAR dataset provides the opportunity to monitor floods at unprecedentedly high spatial and temporal resolutions. However, the accuracy of the flood maps can be affected by the image polarization, the flood detection method used, and the reference data. This research compared change detection and histogram thresholding methods using co-polarization (VV) and cross-polarization (VH) images for flood mapping in the Akaki catchment, Ethiopia, where Addis Ababa city is located. Reference data for the accuracy assessment were collected on the satellite overpass date. A new method, Root of Normalized Image Difference (RNID), has been developed for change detection. Multi-temporal flood maps using the best performing method and image polarization were generated from April to November of 2017–2020. Better accuracy was observed when using the RNID method on the VH polarization image with an overall accuracy of 95% and a kappa coefficient of 0.86. Results showed that flooding in the Akaki commonly begins in May and recedes in November, but flooding was most frequent and widespread from June to September. Irrigated land and built-up area accounted for 1057 ​ha and 544 ​ha of the inundated area, respectively. Several major roads in the study area were also affected by the floods during this period. Our findings indicate that the S-1 images were very useful for flood inundation mapping, the new change detection method (RNID) performed better in urban and peri-urban flood mapping, but the accuracy of the flood map significantly varied with the flood detection method and the image polarization.

Urban-Aware U-Net for Large-Scale Urban Flood Mapping Using Multitemporal Sentinel-1 Intensity and Interferometric Coherence

Urban-Aware U-Net for Large-Scale Urban Flood Mapping Using Multitemporal Sentinel-1 Intensity and Interferometric Coherence

Improving Urban Flood Mapping by Merging Synthetic Aperture Radar-Derived Flood Footprints with Flood Hazard Maps

Remotely sensed flood extents obtained in near real-time can be used for emergency flood incident management and as observations for assimilation into flood forecasting models. High-resolution synthetic aperture radar (SAR) sensors have the potential to detect flood extents in urban areas through clouds during both day- and night-time. This paper considers a method for detecting flooding in urban areas by merging near real-time SAR flood extents with model-derived flood hazard maps. This allows a two-way symbiosis, whereby currently available SAR urban flood extent improves future model flood predictions, while flood hazard maps obtained after the SAR overpasses improve the SAR estimate of urban flood extents. The method estimates urban flooding using SAR backscatter only in rural areas adjacent to urban ones. It was compared to an existing method using SAR returns in both rural and urban areas. The method using SAR solely in rural areas gave an average flood detection accuracy of 94% and a false positive rate of 9% in the urban areas and was more accurate than the existing method.

City scale urban flood modelling and mapping

Inadequate drainage capacity often fails to offer the desired response in urban flooding. This aim of this study was to develop a flood hazard map through a field survey and numerical modelling. The field survey comprises cross-sectional data for tertiary drains; these are then transferred for setting up the numerical model. To feature flooding conditions, both hydrological and hydraulic model studies were carried out using the Hydrologic Engineering Center–hydrologic modelling system (HEC-HMS) developed by the US Army Corps of Engineers, and the personal computer storm water management model (PCSWMM). The use of HEC-HMS generated the runoff; the PCSWMM model simulation was carried out for 760 conduits and 693 junctions in the city that was studied. Three major outlets were considered for model calibration for 2014, the model simulated runoff showed a reasonable match with field records – that is, R2 = 0.782, R2 = 0.719 and R2 = 0.768 for the reaches Chaktai Khal, Rajakhali Khal and Mahesh Khal, respectively. Thus, the validated model was used for flood hazard mapping. It is envisaged that development authority and city planners would benefit from this coupled model.

Spatio-temporal-spectral observation model for urban remote sensing

ABSTRACT Taking cities as objects being observed, urban remote sensing is an important branch of remote sensing. Given the complexity of the urban scenes, urban remote sensing observation requires data with a high temporal resolution, high spatial resolution, and high spectral resolution. To the best of our knowledge, however, no satellite owns all the above characteristics. Thus, it is necessary to coordinate data from existing remote sensing satellites to meet the needs of urban observation. In this study, we abstracted the urban remote sensing observation process and proposed an urban spatio-temporal-spectral observation model, filling the gap of no existing urban remote sensing framework. In this study, we present four applications to elaborate on the specific applications of the proposed model: 1) a spatio-temporal fusion model for synthesizing ideal data, 2) a spatio-spectral observation model for urban vegetation biomass estimation, 3) a temporal-spectral observation model for urban flood mapping, and 4) a spatio-temporal-spectral model for impervious surface extraction. We believe that the proposed model, although in a conceptual stage, can largely benefit urban observation by providing a new data fusion paradigm.

Urban flood hazard mapping using machine learning models: GARP, RF, MaxEnt and NB

Rapid urban development, increasing impermeable surfaces, poor drainage system and changes in extreme precipitations are the most important factors that nowadays lead to increased urban flooding and it has become an urban problem. Urban flood mapping and its use in making an urban development plan can reduce flood damages and losses. Constantly producing urban flood hazard maps using models that rely on the availability of detailed hydraulic-hydrological data is a major challenge especially in developing countries. In this study, urban flood hazard map was produced with limited data using three machine learning models: Genetic Algorithm Rule-Set Production, Maximum Entropy (MaxEnt), Random Forest (RF) and Naive Bayes for Kermanshah city, Iran. The flood hazard predicting factors used in modeling were: slope, land use, precipitation, distance to river, distance to channel, curve number (CN) and elevation. Flood inventory map was produced based on available reports and field surveys, that 117 flooded points and 163 non-flooded points were identified. Models performance was evaluated based on area under the receiver-operator characteristic curve (AUC-ROC), Kappa statistic and hits and miss analysis. The results show that RF model (AUC-ROC = 99.5%, Kappa = 98%, Accuracy = 90%, Success ratio = 99%, Threat score = 90% and Heidke skill score = 98%) performed better than other models. The results also showed that distance to canal, land use and CN have shown more contribution among others for modeling the flood and precipitation had the least effect among other factors. The findings show that machine learning methods can be a good alternative to distributed models to predict urban flood-prone areas where there are lack of detailed hydraulic and hydrological data.

Urban Flood Mapping With Bitemporal Multispectral Imagery Via a Self-Supervised Learning Framework

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.

Testing Urban Flood Mapping Approaches from Satellite and In-Situ Data Collected during 2017 and 2019 Events in Eastern Canada

The increasing frequency of flooding worldwide has driven research to improve near real-time flood mapping from remote-sensing data. Improved automation and processing speed to map both open water and vegetated area flooding have resulted from these research efforts. Despite these achievements, flood mapping in urban areas where a significant number of overall impacts are felt remains a challenge. Near real-time data availability, shadowing caused by manmade infrastructure, spatial resolution, and cloud cover inhibiting optical transmission, are all factors that complicate detailed urban flood mapping needed to inform response efforts. This paper uses numerous data sources collected during two major flood events that impacted the same region of Eastern Canada in 2017 and 2019 to test different urban flood mapping approaches presented as case studies in three separate urban boroughs. Cloud-free high-resolution 3 m PlanetLab optical data acquired near peak-flood in 2019 were used to generate a maximum flood extent product for that year. Approaches using new Lidar Digital Elevation Models (DEM)s and water height estimated from nineteen RADARSAT-2 flood maps, point-based flood perimeter observations from citizen geographic information, and simulated traffic camera or other urban sensor network data were tested and verified using independent data. Coherent change detection (CCD) using multi-temporal Interferometric Wide (IW) Sentinel-1 data was also tested. Results indicate that while clear-sky high-resolution optical imagery represents the current gold standard, its availability is not guaranteed due to timely coverage and cloud cover. Water height estimated from 8 to 12.5 m resolution RADARSAT-2 flood perimeters were not sufficiently accurate to flood adjacent urban areas using a Lidar DEM in near real-time, but all nineteen scenes combined captured boroughs that flooded at least once in both flood years. CCD identified flooded boroughs and roughly captured their flood extents, but lacked timeliness and sufficient detail to inform street-level decision-making in near real-time. Point-based flood perimeter observation, whether from in-situ sensors or high-resolution optical satellites combined with Lidar DEMs, can generate accurate full flood extents under certain conditions. Observed point-based flood perimeters on manmade features with low topographic variation produced the most accurate flood extents due to reliable water height estimation from these points.

Patch Similarity Convolutional Neural Network for Urban Flood Extent Mapping Using Bi-Temporal Satellite Multispectral Imagery

Urban flooding is a major natural disaster that poses a serious threat to the urban environment. It is highly demanded that the flood extent can be mapped in near real-time for disaster rescue and relief missions, reconstruction efforts, and financial loss evaluation. Many efforts have been taken to identify the flooding zones with remote sensing data and image processing techniques. Unfortunately, the near real-time production of accurate flood maps over impacted urban areas has not been well investigated due to three major issues. (1) Satellite imagery with high spatial resolution over urban areas usually has nonhomogeneous background due to different types of objects such as buildings, moving vehicles, and road networks. As such, classical machine learning approaches hardly can model the spatial relationship between sample pixels in the flooding area. (2) Handcrafted features associated with the data are usually required as input for conventional flood mapping models, which may not be able to fully utilize the underlying patterns of a large number of available data. (3) High-resolution optical imagery often has varied pixel digital numbers (DNs) for the same ground objects as a result of highly inconsistent illumination conditions during a flood. Accordingly, traditional methods of flood mapping have major limitations in generalization based on testing data. To address the aforementioned issues in urban flood mapping, we developed a patch similarity convolutional neural network (PSNet) using satellite multispectral surface reflectance imagery before and after flooding with a spatial resolution of 3 meters. We used spectral reflectance instead of raw pixel DNs so that the influence of inconsistent illumination caused by varied weather conditions at the time of data collection can be greatly reduced. Such consistent spectral reflectance data also enhance the generalization capability of the proposed model. Experiments on the high resolution imagery before and after the urban flooding events (i.e., the 2017 Hurricane Harvey and the 2018 Hurricane Florence) showed that the developed PSNet can produce urban flood maps with consistently high precision, recall, F1 score, and overall accuracy compared with baseline classification models including support vector machine, decision tree, random forest, and AdaBoost, which were often poor in either precision or recall. The study paves the way to fuse bi-temporal remote sensing images for near real-time precision damage mapping associated with other types of natural hazards (e.g., wildfires and earthquakes).