Rapid urbanization and increasing population in cities with a large portion of them settled in deprived neighborhoods, mostly defined as slum areas, have escalated inequality and vulnerability to natural disasters. As a result, monitoring such areas is essential to provide information and support decision-makers and urban planners, especially in case of disaster recovery. Here, we developed an approach to monitor the urban deprived areas over a four-year period after super Typhoon Haiyan, which struck Tacloban city, in the Philippines, in 2013, using high-resolution satellite images and machine learning methods. A Support Vector Machine classification method supported by a local binary patterns feature extraction model was initially performed to detect slum areas in the pre-disaster, just after/event, and post-disaster images. Afterward, a dense conditional random fields model was employed to produce the final slum areas maps. The developed method detected slum areas with accuracies over 83%. We produced the damage and recovery maps based on change analysis over the detected slum areas. The results revealed that most of the slum areas were reconstructed 4 years after Typhoon Haiyan, and thus, the city returned to the pre-existing vulnerability level.

Floods are frequent hydro-meteorological hazards which cause losses in many parts of the world. In hilly and mountainous environments, floods often contain sediments which are derived from mass movements and soil erosion. The deposited sediments cause significant direct damage, and indirect costs of clean-up and sediment removal. The quantification of these sediment-related costs is still a major challenge and few multi-hazard risk studies take this into account. This research is an attempt to quantify sediment deposition caused by extreme weather events in tropical regions. The research was carried out on the heavily forested volcanic island of Dominica, which was impacted by Hurricane Maria in September 2017. The intense rainfall caused soil erosion, landslides, debris flows, and flash floods resulting in a massive amount of sediments being deposited in the river channels and alluvial fan, where most settlements are located. The overall damages and losses were approximately USD 1.3 billion, USD 92 million of which relates to the cost for removing sediments. The deposition height and extent were determined by calculating the difference in elevation using pre- and post-event Unmanned Aerial Vehicle (UAV) data and additional Light Detection and Raging (LiDAR) data. This provided deposition volumes of approximately 41 and 21 (10³ m³) for the two study sites. For verification, the maximum flood level was simulated using trend interpolation of the flood margins and the Digital Terrain Model (DTM) was subtracted from it to obtain flooding depth, which indicates the maximum deposition height. The sediment deposition height was also measured in the field for a number of points for verification. The methods were applied in two sites and the results were compared. We investigated the strengths and weaknesses of direct sediment observations, and analyzed the uncertainty of sediment volume estimates by DTM/DSM differencing. The study concludes that the use of pre- and post-event UAV data in heavily vegetated tropical areas leads to a high level of uncertainty in the estimated volume of sediments.

Extreme rainfall events often trigger landslides, debris flows, and sediment-laden floods that cause severe damage in built-up areas, yet sediment deposition is rarely quantified in hazard assessments. This study evaluates the capability of the physically based catchment model LISEMHazard to reconstruct sediment generation, transport, and deposition during Hurricane Maria (2017) in two catchments in Dominica (Coulibistrie and Grand Bay). Simulations were performed at 10 m resolution using rainfall, topography, soil, and land-use data. Model calibration and validation used mapped landslides and debris flows, field measurements of deposition height, and DEMs of Difference (DoDs). LISEMHazard reproduced the general magnitude of sediment volumes and the frequency–area distribution of medium and large landslides but showed poor ability to predict their exact locations and overestimated landslide depth and deposition height. Agreement between modeled and observed debris-flow patterns was good in major channels but weak in minor ones. Sensitivity analysis indicated that soil depth and cohesion dominate uncertainties, whereas saturated hydraulic conductivity and surface roughness exert minimal influence. Despite substantial data and model limitations, physically based modeling remains a practical approach for spatial estimation of sediment deposition needed for risk assessment, structural damage evaluation, and cleanup cost estimation.

Floods and associated landslides account for a large number of natural disasters and affect many people wherever they occur (Hong et al., 2007). Sediment-free floods are rare, and in most cases, floods carry a notable amount of sediments (Acreman, 2016). Mass movement processes also transport a huge amount of sediments within a short period (Varnes, 1978). The mobilized sediments cause significant costs and damages as soon as they reach urban or rural environments. These damages and costs include (but are not limited to) cleaning or dredging cost, damage to contents of buildings (e.g. furniture, electric appliances), and blockage of drainage and sewer systems which get filled up with sediments (Einstein, 1950; Merz et al., 2010; Rodríguez et al., 2012).This study aims to achieve a reliable sediment deposition quantification which is useful for assessing the risk of such events. Three methods were implemented for this purpose. First, the sediment deposition height was determined through in-situ investigation and the average height was estimated. Second, the deposition surface was simulated using trend interpolation and DEM was subtracted from that to get deposition height. Third, the deposition height and extent were determined by calculating the difference in elevation using pre- and post-event drone and LiDAR flights.Dominica has experienced sediment deposition events in the past. It is significantly vulnerable to tropical storms and hurricanes. Dominica is a mountainous island covered by tropical rainforests and located about halfway between the French islands of Guadeloupe and Martinique in the Eastern Caribbean sea (Knutson et al., 2015; Malhotra et al., 2007; Wilkinson, 2018). Hurricane Maria made landfall on this island on September 18th, 2017 and it heavily impacted the housing, transport infrastructure, tourism, agriculture, and education sectors (Dominica News Online, 2018). The intense rainfall caused flash floods, landslides, and debris flows resulting in a massive amount of sediments being deposited in urban and rural areas. The overall damages and losses are estimated at approximately USD 1.3 billion (The Government of the Commonwealth of Dominica, 2017). Dominica’s Ministry of Public Works reported that the total cost related to deposition of sediments (e.g. dredging rivers, cleaning streets and main roads, and clearing of airports and seaports)  exceeds USD 92 million which is a considerable portion of total damages and costs. This implies the significance of the risk imposed by sediment deposition.The results of this research were compared with each other and with the findings of in-situ investigations. They indicate similar deposition heights and volumes, however, the pattern and extent of deposition are not the same. The practicality of the third method depends on the availability of data, but when data is available the outcomes provide a reliable assessment of sediment deposition volume. However, this cannot be trusted unless an in-situ investigation is performed.