ABSTRACT. Most hydrological phenomena like floods and droughts are modeled as stochastic processes defined on filtered probability spaces. The Markovian properties are crucial in establishing a feedback control system for a stochastic process. Therefore, using a Markov process in modeling these phenomena is often the first step in rational risk management. The authors have developed several Markov process models, encompassing different cases of stochasticity in the risks and designing feedback controllers for risk aversion based on dynamic programming. This study first reviews the mathematical foundations of Markov processes and their control, including the notions of filtered probability spaces, Bellman’s maximum principle, from which the governing time evolution equation of the value function is derived, and the viscosity solution to the equation. The celebrated WGEN model is used as the simplest but the most significant example of the Markov process in hydrology to explain how filtration is constructed. Reservoir operation is the principal application of the controlled Markov process. Using stochastic dynamic programming, we present a case study of the time-varying optimal release policy in a large multi-purpose reservoir under the tropical rainforest climate, comparing theoretical policy with the actual operation. A more critical case is the deterministic operation of a water tank for full irrigation, where discontinuities may appear in the value function. For drought risk management in arid and semi-arid environments, the Markov process of dry spell lengths serves as an index or an underlying weather variable. The main difficulty is identifying transition probabilities of the Markov process from limited data of extreme events, for which we propose a regularization technique using total variation minimizing flow. Measures of drought risk aversion in agriculture based on feedback control are typically crop management and supplementary irrigation, as common in empirical knowledge. However, with the progress of climate change, the probability distribution of extreme weather variables can become heavy-tailed, implying that conventional approaches to risk management, including financial instruments, may not work effectively.

ABSTRACT. This research evaluates water balance and allocation strategies in the Ghriss Basin, located in the province of Mascara. As the most productive agricultural area in the West-Northern part of Algeria, the Ghriss Basin boasts abundant biodiversity and a highly productive ecosystem despite its semi-arid climate. Covering an area of 1,849 km², the basin relies primarily on underground water to meet the needs of domestic, agricultural, and industrial sectors. This critical resource underpins regional development but faces sustainability challenges due to overexploitation driven by agricultural expansion, population growth, and competition among users. A robust water management policy is essential to preserve the basin's water resources. This requires forecasting future water resource mobilization and utilization. To address these challenges, the WEAP (Water Evaluation and Planning) system was employed to assess future water supply and demand projections up to 2052 in the Ghriss Basin. The results are promising and provide a decision support system for efficient water resource management. They also offer valuable insights for policymakers and water resource managers regarding best practices for managing water resources in the basin. Under the first scenario, rapid population growth significantly increases water demand, surpassing the available resources. The second scenario, which involves measures to reduce individual water consumption, leads to an overall drop in demand but does not fully offset the increased needs caused by population growth. The third scenario aims to bolster water resources and improve domestic water supply, yet agricultural water needs continue to face shortages. Ultimately, the fourth scenario identifies that a stable allocation of 43.8 m³ per capita per year for domestic use, combined with 2,000 m³ per hectare per year for agriculture, would meet the region's water requirements during the study period. The "Reinforcement & Allocation" scenario emerges as the most suitable approach to address unmet demand and optimize water allocation and conservation in the Ghriss Basin.

ABSTRACT. Runoff patterns and water accumulation in arid regions vary between events due to the spatial and temporal distribution of rainfall, as well as significant sediment transport, impacting flow routes. Identifying these zones is crucial for recognizing groundwater recharge after flash floods.

The high infiltration rates of bare, arid soils over alluvial plains result in short-lived runoff and water accumulation. Consequently, satellite imagery often struggles to detect water bodies due to the long revisit periods of low Earth orbit satellites. However, aeolian fine deposits prevalent in arid regions enhance runoff with turbidity, leaving traces detectable via satellite.

This study proposes a methodology to map water body occurrences following major rain events and delineate recharge zones in the alluvial deposits of selected catchments in the Batinah plains, Northern Oman. The approach involves utilizing multispectral infrared red to visual imagery products of Landsat 8 and 9, as well as Sentinel-2, after specific rain events to monitor short-term surface water dynamics. Validation of this method is conducted through comparison with hydrodynamic numerical models. 3Di modeling is employed for Bani Kharus catchments in the Batinah Plain by using Digital Elevation Models (DEMs) with high resolution 5x5m. DEM were used in 3Di modeling to incorporate terrain features that influence water flow and accumulation. Raster maps of Maximum water depth and maximum flow velocity were analyzed and compared with classified multispectral raster provided through remote sensing (RS). Sentinel 2 imagery captured on 18 April 2024 over Batina plain was classified into 20 classes using K-mean clustering algorithm. Two classes helped detect what is proposed to be sediment-laden flood traces or siltation. The remote sensing output was validated against 3Di modeling outputs on the coastal alluvial plain. The slowdown of velocity and flow in the braided channel part, where multiple branches reduce flow speed cause silt sedimentation which can be detected by RS. The comparison showed that 96% of 3Di model output in the braided part of Wadi Bani Kharus was captured through remote sensing. Although the maximum inundation map covers an area of 3.02 km², while the selected classes have an area of 5.8 km², there are differences in resolution, which are 5m for 3Di product and 30m for Sentinel-2. This allows satellites to more easily detect turbidity streamlines, which are helps detecting recharge zones. However, in higher terrain with faster flow rates, the satellite was unable to identify these zones. Results demonstrate the feasibility of analyzing satellite imagery to identify flood traces and assist in mapping groundwater recharge zones in the Batinah plains. This study emphasizes the value of integrating multi-sensor satellite data with advanced modeling approaches to enhance water resource management in arid regions.

ABSTRACT. The conventional approach to design rainfall standards has typically been based on adherence to Intensity-Duration-Frequency (IDF) curves set forth by a national authority in a Design Manual expressing the relation between rainfall intensity, duration, and average recurrence interval. Extreme rainfall may vary significantly within the same geographical region or areas sharing the same geomorphic/landcover characteristics. Qatar represents a land area of 11,571 km2 with an average annual maximum rainfall in the north, approximately 25% higher than the country's south. Before 2015, Qatar employed a uniform IDF curve nationwide; however, leading up to the World Cup, the Qatari Authority examined the spatial distribution of intense rainfall, with attention given to the location of stadiums and World Cup events. Eventually, Qatar developed a “rainfall portal” to incorporate significant spatial disparities in extreme rainfall. This software was designed to obtain local IDF curves based on project locations. This article details the creation and implementation of the rainfall portal and the flood portal. These tools have enhanced the efficiency of drainage infrastructure spending in Qatar. They have also facilitated improved economic distribution throughout the country, bolstering climate resilience and readiness for events like the World Cup. The authorities expanded the capabilities of the rainfall portal after the World Cup. These enhancements included features related to urban expansion and associated variations in land use and land cover over time. These features are powered by artificial intelligence (AI) algorithms, analyzing historical aerial and satellite imagery alongside future demographic projections. Downstream of the rainfall portal, the authorities utilised design hyetographs, a digital terrain model, and a TuFlow flood model run on supercomputers to produce flood inundation maps for the entire country. Additionally, the focus includes protecting the Wadi streams near urban areas, such as Wadi Lusail north of Doha and Wadi Al Jalta north of Al Khor. The upgraded version of the rainfall portal and the flood portal has transformed drainage design practices in Qatar, fostering greater awareness of flooding, climate impacts, and shifts in land use due to urbanization. The Qatar Highway Design Manual, used for road drainage design, will be moved online and integrated with the rainfall portal, allowing digital tools to standardize the drainage design process. Furthermore, the rainfall portal will be connected to real-time meteorological data, enabling the automated updating of design criteria at regular intervals as more climate information becomes available. The rainfall portal can export the required hyetographs for the flood modeling tools, such as Infoworks and TuFlow.

ABSTRACT. Urbanization and Flood Risk: A Novel Approach to Hydrological Modeling in Imphal, India Using High-Resolution LULC Data and GSSHA Abstract: Effective hydrological modeling is crucial for managing water resources and mitigating flood risks, particularly in urbanized watersheds like Imphal in India. This study aims to enhance the understanding of hydrological responses in the Imphal watershed by utilizing high-resolution Land Use and Land Cover (LULC) data and integrating these into the Gridded Surface Subsurface Hydrologic Analysis (GSSHA) model within the Watershed Modeling System (WMS). Imphal's rapid urbanization and resulting land cover changes significantly affect runoff characteristics, exacerbating flood risks. High-resolution LULC maps were developed using advanced remote sensing techniques and classified using the Random Forest algorithm, achieving high classification accuracy. These maps reveal significant changes in land cover over recent years, primarily due to urban expansion. The GSSHA model, known for its comprehensive groundwater-to-surface-water interaction capabilities, was employed to simulate hydrological responses under different LULC scenarios. Given the challenge of obtaining ground truth data for model calibration and validation in the Imphal watershed, this study proposes a novel approach by using consistent temporal scenarios. By applying synthetic rainfall events or historical rainfall data, we ensure comparability and reliability across all simulations. Additionally, we integrate daily flood reports, spanning several years, for further calibration and validation. This approach allows us to analyze relative changes rather than absolute values, making it possible to infer the impacts of urbanization and other land cover changes on flood behavior. To the best of our knowledge, this is the first study to apply the GSSHA model in the Imphal watershed, focusing on high-resolution LULC mapping and its integration into hydrological models. This novel methodology addresses the gap in ground truth data by ensuring model consistency through synthetic or historical rainfall data and utilizing flood reports, providing a robust framework for urban flood risk management and planning. By leveraging advanced satellite imagery and hydrological modeling techniques, this research offers significant contributions to the field of urban flood management, particularly in humid environments experiencing rapid urbanization. The findings underscore the importance of high-resolution LULC data in improving flood prediction accuracy and urban planning. This research supports sustainable development goals by promoting resilient urban infrastructure and disaster preparedness. Future studies should build on these findings to incorporate ground truth data for further calibration and validation, potentially utilizing historical discharge data or remote sensing technologies. The methodology and insights gained from this study can serve as an outline for similar hydrological modeling efforts in other urbanized watersheds, contributing to more sustainable and resilient urban development globally.

ABSTRACT. Rating curves, which depict the relationship between water surface elevation and discharge in a river, are crucial for understanding river behavior, analyzing flood risks, and conducting accurate hydrological modeling. The development and analysis of rating curves, however, are subject to temporal variations influenced by both natural and anthropogenic factors. The present study incorporates Monte Carlo simulation to estimate the parameters of the rating curve at two gauging stations Kudurmal and Hati on river Hasdeo and Mand respectively. By generating synthetic datasets through the addition of random noise to observed discharge data, the parameters were re-estimated for each synthetic dataset using non-linear optimization techniques. This approach allowed to quantify the uncertainty and variability in the parameter estimates, providing robust and reliable final parameter values. Results show significant changes in rating curves over time, with shifts in stage-discharge relationships, altered curve shapes, and varying degrees of hysteresis. For the gauging station at river Hasdeo, the rating curve coefficients have exhibited significant changes over two different time periods. The scale coefficient (a) decreased from 17.4 (for the period 2001-2007) to 4.59 (for the period 2007-2018), while the shape coefficient (b) increased from 3.6 to 4.38 over the same periods. This variation in the coefficients indicates changes in the river's flow characteristics or channel geometry over time. The Nash-Sutcliffe Efficiency (NSE) value, which measures the predictive accuracy of the developed rating curves, was 0.93 for the 2001-2007 period and 0.87 for the 2007-2018 period, indicating a high level of reliability for both periods. For the gauging station at river Mand, the scale coefficient (a) and shape coefficient (b) for the rating curves have varied across different time periods, reflecting changes in river flow characteristics. For the period 2001-2008, the scale coefficient (a) was 3.68, which decreased to 0.79 for the period 2009-2014 and further to 0.4 for the period 2015-2019. Meanwhile, the shape coefficient (b) increased from 3.39 (2001-2008) to 4.00 (2009-2014), and then to 4.12 (2015-2019). The Nash-Sutcliffe Efficiency (NSE) values for these periods were 0.93, 0.88, and 0.77, respectively, indicating a reliable predictive accuracy of the rating curves over time. These changes are attributed to natural and anthropogenic factors, including sedimentation, erosion, channel morphology modifications, and land use changes. The study highlights the importance of regularly updating rating curves to ensure accurate hydrological predictions and flood risk assessments. This research contributes to improved hydrological monitoring and management of the River Hasdeo and Mand, with implications for similar rivers globally.

ABSTRACT. In Najran Valley, Saudi Arabia, a region prone to both climatic and hydrological hazards, the relationship between precipitation and the floodwater reaching the dam is significantly influenced by the area's topography, soil saturation, and rainfall patterns. Intense or heavy rainfall rapidly generates surface runoff due to the steep terrain, leading to a rapid rise in floodwater levels at the dam. Understanding this relationship is crucial for effectively managing flood risks and dam operations in the valley. This study provides a comprehensive assessment of climate and flood risks in the Najran region. The assessment integrates advanced climate modelling, hydrological analysis, and risk evaluation frameworks to provide a holistic understanding of the risks and inform mitigation strategies. The findings reveal that the Najran region is likely to experience more frequent and severe flooding due to projected increases in extreme rainfall events. The risk maps identify key areas within the region that are particularly vulnerable, necessitating targeted intervention strategies. The study recommends a multi-faceted approach to flood risk management, integrating advanced technological solutions with community engagement and policy support to build long-term resilience. Overall, this assessment provides essential insights for policymakers, urban planners, and disaster management authorities to develop effective strategies to mitigate climate and flood risks in the Najran region. By adopting the recommended measures, the region can better prepare for future climate impacts and protect its communities and infrastructure from the adverse effects of flooding.

ABSTRACT. This study harnessed the formidable predictive capabilities of three state-of-the-art machine learning models—extreme gradient boosting (XGB), random forest (RF), and CatBoost (CB)—applying them to meticulously curated datasets of topographical, geological, and environmental parameters; the goal was to investigate the intricacies of flood susceptibility within the arid riverbeds of Wilayat As-Suwayq, which is situated in the Sultanate of Oman. The results underscored the exceptional discrimination prowess of XGB and CB, boasting impressive area under curve (AUC) scores of 0.98 and 0.91, respectively, during the testing phase. RF, a stalwart contender, performed commendably with an AUC of 0.90. Notably, the investigation revealed that certain key variables, including curvature, elevation, slope, stream power index (SPI), topographic wetness index (TWI), topographic roughness index (TRI), and normalised difference vegetation index (NDVI), were critical in achieving an accurate delineation of flood-prone locales. In contrast, ancillary factors, such as annual precipitation, drainage density, proximity to transportation networks, soil composition, and geological attributes, though non-negligible, exerted a relatively lesser influence on flood susceptibility. This empirical validation was further corroborated by the robust consensus of the XGB, RF and CB models. By amalgamating advanced deep learning techniques with the precision of geographical information systems (GIS) and rich troves of remote-sensing data, the study can be seen as a pioneering endeavour in the realm of flood analysis and cartographic representation within semiarid fluvial landscapes. The findings advance our comprehension of flood vulnerability dynamics and provide indispensable insights for the development of proactive mitigation strategies in regions that are susceptible to hydrological perils.

ABSTRACT. Floods are considered one of the most common and dangerous natural disasters, due to their negative effects in rural and urban areas. El Bayadh city in Algeria has witnessed devastating floods that caused significant human and economic losses. Identifying areas at risk of flooding has become critical to reducing this risk and protecting lives and property. A flood risk map was produced to reduce and mitigate flood risks in El Bayadh city based on a multi-criteria decision-making (MCDM) process. The main objective of this study is to produce a flood risk map of El Bayadh city with the determination of the risk degree, using the Analytic Hierarchy Process (AHP) and geographical information system (GIS). Five factors were selected in this study, including flood hazard map, buildings close to the wadi, distance from the wadi, road type, and slope level. The results indicate that 8.2 % of the areas in El Bayadh city are between medium and very high risk, 2.4 % are medium risk, 5.3 % are high risk, and 0.5 % are very high risk, representing buildings on the edges of the wadi.

ABSTRACT. Across the Middle Eastern regions, the inherent challenge associated with monitoring surface water runoff in arid and semi-arid wadis (also known as arroyos) is due to their localized spatial distribution of large-scale precipitation events, as well as inhospitable and ever-changing climatic conditions, such as sandstorms. Within these environments, hydrological phenomena occur mostly as heavy rainfall and sudden flash floods with a time span of years between reoccurrences of significant events. The periodicity of these events highlights the continuous evolving nature of the region’s hydrologic landscape.

Wadis are well-known ephemeral drainage courses widespread across the Middle East (i.e., Saudi Arabia, United Arab Emirates and other near Eastern countries). Wadis act as primary conduits for channeling of surface water and depression focused recharge. These watercourses typically remain dry except during rainfall periods, triggering quick and intense flash floods, where the peak values happening within the first few minutes of the rainfall, signifying the challenges in monitoring these unforeseeable events.

DischargeKeeper (DK) is a new image-based system for comprehensive flow monitoring integrated with a modern Pan-tilt-zoom (PTZ) camera. DK measures volumetric flow during day and night times and all year-round. The system can function in rivers or canals of different sizes and flow velocities and under different visibility conditions.

The installation and testing of DK device were carried out in April 2021 along the banks of Wadi Naqab, which is located in the north of United Arb Emirates (UAE). This wadi has an approximate width of 50 meters and has been dried for most of its existence. Between 31st December 2021 and 4th January 2022, a significant event triggered by heavy rainfall recorded a discharge peak reaching 78 m3/s within 15 minutes just after the initiation of water flow.

DK system validation was performed by a correlation of camera-based flow readings with those obtained from an Acoustic Doppler Current Profiler (ADCP) as a standard. Calibration of the system requires the adjustment of certain input parameters: (i) river or channel cross-section, (ii) four detectable reference points to be located far-off the shore within the camera's field of view, and (iii) initial water level corresponds to the current waterline position. These parameters are needed only during the initial establishment of a new monitoring site. The deviation found between ADCP and DK measurements was less than 5% in 40% of performed measurements (a total of 11 comparisons). Deviations of less than 10% and 15% were obtained for 60% and 80% of measurements, respectively. The difference mainly ranged between 0.01 m3 /s and 0.1 m3 /s.

ABSTRACT. Groundwater Recharge Dams in the Sultanate of Oman Ahmed Said Al Barwani1* and Ayisha Al Khatri2 1Oman Water Society, P.O. Box. 877, P.C 119. Al Amrat, Sultanate of Oman. 2Ministry of Agricultural, Fisheries and Water Resources, Al Khuwair, Ministries District, Sultanate of Oman. *Corresponding author: E-mail: ahmed.albarwani@gmail.com

Keywords Groundwater recharge dams, alluvial valleys.

Abstract The Sultanate of Oman is situated in the south eastern part of Arabian Peninsula, boarded by United Arab Emirates (U.A.E) from North West, Saudi Arabia from the west, Oman Sea and Arabian Sea from the east and south east. Groundwater recharge dams are used in arid countries to enhance the natural groundwater recharge by a controlled infiltration of stored flood events (Haimerl, Gerhard 2004). The main purpose of these dams is to enhance ground water aquifers through making use of flood water which is often lost to the sea and desert. This can be done by storing such water under the ground in order to use it later for various purposes. Moreover, these dams provide some degree of protection against floods as well as in curtailing intrusion of sea water into groundwater aquifers. Recharge Dams are constructed on alluvial valley channels for storing flood water in dam reservoirs for a temporary period of not more than fourteen days to avoid evaporation losses and health risks. Then the stored water is released slowly through controllable culverts. The idea of the possibility of implementing underground recharge dam facilities in the Sultanate of Oman began in 1978. The importance of these dams is highlighted through many objectives, such as retaining part of the floodwaters, using them to recharge groundwater reservoir, and to reduce seawater intrusion into the groundwater recharge reservoirs. These dams also provide more groundwater for the continuity of various development projects. Recharge dams also provide a degree of protection from flood risks and prevent traffic disruption. During the period (1985-2024), the government completed the implementation of (190) dams, where 70 groundwater recharge dams were implemented with a total storage capacity estimated at 108,769 million cubic meters. In conclusion we will address the Omani experience in groundwater recharge dams and whether we replicate these dams in order to increase aquifers recharge.

ABSTRACT. The Kagera aquifer in the equatorial region of Africa primarily consists of low-lying areas with alluvial deposits surrounding the Kagera River. A bank infiltration Managed Aquifer Recharge (MAR) approach has been recommended to enhance the conjunctive use of surface and groundwater. This strategy aims to improve access to clean and safe water for the local population. The proposed intervention involves constructing a subsurface high-permeability ditch to collect and store groundwater in sufficient quantities to meet the domestic water demand of targeted communities. A pilot structure was designed to serve about 500 people, providing approximately 20 cubic meters per day. The MAR structure will be oriented within the native aquifer, perpendicular to the natural flow gradient, and designed with adequate dimensions to ensure sufficient yield under varying hydrologic conditions. It will alter the flow regime towards the collection reservoir from a line flow (as in wells) to a plane flow (as in trenches). Additionally, the structure will function as a subsurface storage reservoir. The system will be equipped with wells and a solar pump to lift water from the underground reservoir to an elevated tank, which will then be connected to the community's water distribution point.

ABSTRACT. Flash floods in wadi systems pose significant risks to lives, infrastructure, and the environment, particularly in arid and semi-arid regions. This study investigates the application of machine learning techniques to enhance the accuracy and reliability of flash flood prediction. Key aspects of the study include the integration of diverse datasets encompassing meteorological, hydrological, and topographical data, alongside historical flood events. Rigorous preprocessing techniques, including normalization and feature engineering, ensure high-quality data inputs. Machine learning models such as Random Forest, Gradient Boosting, and Deep Learning are employed to analyse complex relationships and predict flash floods. Performance evaluation involves assessing model accuracy using metrics such as confusion matrices and validation against historical data. Results demonstrate high prediction accuracy in both historical and simulated real-time scenarios. This research paper contributes valuable insights into leveraging big data and machine learning for effective disaster risk reduction in wadi systems, advancing early warning systems' capabilities to mitigate flash flood impacts.

Key Contributions: Data Utilization: • Integration of diverse datasets including meteorological, hydrological, and topographical data, alongside historical flood events. • Application of rigorous preprocessing techniques to enhance data quality and relevance. Model Development: • Utilization of machine learning algorithms (Random Forest, Gradient Boosting, Deep Learning) to analyse complex relationships and predict flash floods. • Evaluation of multiple models to identify the most effective approach for accurate prediction. Performance Evaluation: • Assessment of model accuracy using metrics like confusion matrices and validation against historical data. • Demonstrated high prediction accuracy in historical and simulated real-time scenarios. Impact and Contribution: • Enhancement of early warning systems through predictive modelling, improving preparedness and response to flash flood events. • Contribution of insights into leveraging big data and machine learning for effective disaster risk reduction in wadi systems. This research advances the field of flash flood prediction, aiming to mitigate the adverse effects of these natural disasters on communities and infrastructure in arid and semi-arid regions.

ABSTRACT. Prediction of sediment loads is crucial for effective water resource management. This study aims to estimate soil loss and sediment yield potential in the Hasdeo river basin using an integrated RUSLE-SDR modeling approach. The analysis utilized gridded IMD rainfall data, soil data from NBSS & LUP, Cartosat DEM, and IRS LISS-IV satellite images with a 5.8 m resolution to develop soil loss parameters over three decades, from 1993 to 2023. The results show that rainfall, soil erosion rates, and sediment yield in the basin vary significantly, ranging from 845 mm to 1762 mm, 40.32 tons/ha/yr to 78.17 tons/ha/yr, and 1.87 million tons to 3.63 million tons, respectively. Notably, the northern part of the basin, characterized by high rainfall and rugged terrain, generates substantially more sediment than the southern part. The analysis reveals that rainfall is the dominant factor influencing sediment yield, surpassing other contributing factors. These findings provide valuable insights into the complex relationship between rainfall and sediment dynamics, establishing a framework for comparable studies in other river basins worldwide.

ABSTRACT. Jeddah city has recently experienced several devastating flood events that have severely impacted the local communities. This study utilized hybrid data mining techniques, specifically classification and association rules, to investigate the complex relationship between large atmospheric-scale teleconnection and extreme precipitation events in Jeddah. The study consisted of two main processes. The first stage involved classification, where the detrend of the surrounding seas' sea surface temperatures (SSTs) of the Mediterranean, Red, Arabian, and Gulf seas, as well as the Southern Oscillation Index (SOI) and the Oceanic Niño Index (ONI), were classified into five classes. The observed monthly precipitation data from 1970 to 2024 was also classified into high and low. The second stage involved the Apriori algorithm, which uncovers the nonlinear features and hidden relationship of the extreme precipitation event, the detrend SST of the surrounding seas, and additional climate indices for the analysis period (1970 – 2024). This approach was applied with different time lags to identify the most representative association rule based on confidence value. The results revealed several extracted rules that shed light on the relationship between detrended SSTs of the surrounding seas and extreme precipitation events. For instance, a significant association rule highlighted the influence of the La Niña phenomenon, low detrended SSTs of the Arabian and Red Seas, and very low detrended SSTs of the Gulf, which led to heavy rainfall and triggered a devastating flash flood event in November 2017 with a confidence value of 100%. This association rule recurred four times (in November 2000, 2008, 2017, and 2021), with a noticeable decrease in the frequency of occurrence from around nine years to five years. This approach can serve as a robust tool for decision-makers, enabling them to make proactive, knowledge-driven decisions to mitigate the impact of flood risks.

ABSTRACT. In Algeria, particularly in arid and semi-arid regions, flash floods represent a catastrophic phenomenon with significant economic and human consequences. The rapid climate changes and rising temperatures are likely to directly influence the frequency of these flash flood events. This research focuses on the Wadi M'zab catchment in Ghardaia city, situated in the arid Saharan region of Algeria. Ghardaia, constructed along the Wadi system (Wadi M'zab), has experienced several extreme events resulting in devastating inundations. A particularly catastrophic flood occurred on October 1st, 2008, claiming 50 lives, injuring 86 individuals, and leaving thousands homeless. Flash Flood Susceptibility Mapping (FFSM) serves as a crucial flood mitigation strategy, employed to predict areas prone to flooding. This study aims to compare three machine learning techniques for identifying flood-vulnerable zones and generating flood susceptibility maps, based on multiple factors that directly influence susceptible areas. By evaluating various algorithms, the research seeks to determine the most effective methods for predicting flood-prone regions. Specifically, this work will evaluate the performance of the newly applied Vision Transformer (ViT) model for flash flood susceptibility mapping against traditional machine learning techniques such as Random Forest and XGBoost.The research utilizes a comprehensive dataset extracted from 597 points within the catchment area, incorporating thirteen distinct features that influence flood susceptibility. To ensure a robust evaluation of model performance, the dataset was strategically divided into two subsets: a training set comprising 70% of the data, and a test set containing the remaining 30%. This study's findings demonstrate the superior performance of the Vision Transformer (ViT) model in flash flood susceptibility mapping compared to traditional machine learning techniques. The results reveal that ViT achieved an impressive Area Under the Curve (AUC) of 99.32 % on the test subset, significantly outperforming both Random Forest and XGBoost, which attained AUC’s of 97.22 and 94.44, respectively. The flood susceptibility maps generated by these models consistently indicate that densely populated areas face a very high risk of flooding. The research conclusively establishes the efficacy of the Vision Transformer against Random Forest, and XGBoost - for Flash Flood Susceptibility Mapping (FFSM) in arid and semi-arid regions of Algeria.

ABSTRACT. Essaouira basin (Western coast of Morocco) is facing significant overexploitation of its water resources over decades and especially over the past 20 years. The karstic nature of the aquifers in this basin worsens complexity to the identification and modeling processes of groundwater dynamics. The regional climate is classified as arid to semi-arid, reflecting an intense deficit in water resources, both surface and underground. Results from several sampling campaigns have revealed high concentrations of heavy metals, such as lead (Pb), iron (Fe), arsenic (As), and cadmium (Cd). However, the region lacks any developed industrial activities that could generate these high amounts of heavy metals. The Essaouira basin contains large reserves of phosphates, which, according to our hypothesis, could be a major natural source of heavy metal contamination. This hypothesis is supported by the potential for phosphate-rich areas to release heavy metals through dissolution in meteoric waters and direct infiltration into the aquifer. In order to explore this hypothesis, we employed remote sensing techniques and the Analytical Hierarchy Process (AHP) method to map potential areas of groundwater pollution by heavy metals. The AHP method provides a systematic and quantitative approach by structuring the problem into a hierarchy, allowing us to compare various criteria pairwise and synthesize the results to identify high-risk contamination zones. Criteria included proximity to phosphate deposits, geological susceptibility, hydrological network density and historical contamination records. Remote sensing techniques complemented this approach by providing spatial data necessary for mapping and analysis. Upon creating a preliminary map of potential pollution zones, we will validate our approach through targeted sampling and physicochemical characterization campaigns in these high-risk areas. This approach not only deals with immediate contamination concerns, but also contributes to wider attempts to understand and minimize sources of natural contamination in similar arid and semi-arid regions around the world.

ABSTRACT. Coastal estuaries hold a prominent economical position due to the abundance of resources in their environment. However, these ecosystems are intricate and face mounting challenges due to population growth, economic activities, climate change impacts, and ecological pressures. The environmental issues arising from these factors necessitate a unique focus on the distinctive characteristics of these critical areas. This requires a comprehensive understanding of the interplay between surface water and groundwater, as well as an analysis of the dynamics of flooding, sedimentation, and marine intrusion. The objective of this research is to obtain a comprehensive model that allows a clear understanding of this interaction, which depends on a judicious selection of approaches and various tools. The methodology begins with a diachronic study using digital earth Africa platform, python code and DSAS method for analyzing historical data and remote sensing imagery (Sentinel 2 and Landsat 8) to examine long-term changes in coastal morphology, sediment distribution, and land cover. Sediment core samples, collected from the fields, provide valuable insights into historical sedimentation trends. This study will conduct a comprehensive analysis of hydro-sedimentary dynamics and associated coastal hazards in Oum-Errabia estuary, that originates from the Middle Atlas Mountains in central Morocco and flows westward, discharging into the Atlantic Ocean near the city of Azemmour. MIKE 21 model is used to analyze hydro- sediment interactions and monitor hydrodynamic and sedimentary behaviour of this estuary based on samples collected through sediment coring, followed by simulations that account for wave and current variations, as well as tidal influences. The validation of these simulations will be achieved through field measurements using Acoustic Doppler Current Profilers (ADCP), ensuring the accuracy of the model outputs, and providing insights into the dynamics of estuarine closure that offers valuable information for understanding and managing coastal environments. Furthermore, the study incorporates an analysis of flood dynamics using MIKE Flood software. This involves integrating high-resolution topographic data from drone images (2m), bathymetry with a spatial resolution of 30m, rainfall and historic hydrodynamic data over a 30-year cycle, wave and wind direction, tidal current and Sea Level Rise (SLR). The findings of the diachronic study based on the analysis of shoreline changes over the period 2000 to 2024, indicated an average distance of 178.33 meters, with the maximum distance being 519.19 meters at transect ID 89, and the minimum distance being 75.4 meters at transect ID 1. The Net Shoreline Movement (NSM) showed an average negative distance of -140.37 meters, highlighting consistent erosion across all transects. This subdivision allows for a more granular analysis of the spatial variability in erosion and accretion, underlying factors driving these changes. Integrating granulometric analysis with advanced modeling techniques provides valuable insights into the hydro-sedimentary dynamics of the Azemmour estuary enabling the assessment of coastal hazards.

ABSTRACT. Floods are among the most destructive natural disasters impacting urban areas worldwide. Predicting floods is challenging due to the complex behavior of water during river overflows. Environmental factors such as terrain indices, topographic features, land cover, and hydrological characteristics further complicate the prediction process. Recently, machine learning algorithms and statistical methods have been employed in numerous research studies to generate flood susceptibility maps using topographical, hydrological, and geological conditioning factors. However, a significant challenge researchers face is the complexity and number of features required as input in a machine-learning algorithm to produce acceptable results. In this research, the Random Forest algorithm was utilized to analyze the effect of various combinations of 13 different flood conditioning factors. The model was tested against real flood data collected from the Japanese Hazard Map Portal Site for four basins: Sapporo, Sendai, Tokyo, and Saga. To determine the most critical conditioning factors for flood mapping, the factors were systematically tested to observe their impact on the model's accuracy. The results indicated that the most significant conditioning factors, which yielded the highest accuracies, were a combination of altitude, slope, distance from the river, river density, and rainfall. This combination achieved an overall accuracy of 96.99%, a mean absolute error of 0.0306, and an area under the curve (AUC) of 0.996.

ABSTRACT. Predicting floods and creating accurate flood risk maps is challenging due to environmental and physical factors. Effective flood modeling requires the calibration of models with observed data, such as rain gauge readings and discharge flow measurements, to ensure accuracy and reliability. Given these complexities, it is necessary to use advanced machine learning models to improve flood susceptibility predictions. This study evaluates the performance of various machine learning classifiers in predicting flood susceptibility using 13 different flood conditioning factors to enhance flood risk mapping. The study employed a dataset of normalized and balanced flood data for training and testing six machine learning classifiers: Random Forest, Gradient Boosting, Logistic Regression, Support Vector Machine (SVM), K-Nearest Neighbors (KNN), and Decision Tree. The models were tested against real flood data collected from the Japanese Hazard Map Portal Site for four basins: Sapporo, Sendai, Tokyo, and Saga. Key performance metrics, including the Receiver Operating Characteristic (ROC), accuracy, and Mean Absolute Error (MAE), were used for model comparison. The Random Forest, Decision Tree, and Gradient Boosting classifiers demonstrated the best performance, with area under the curve (AUC) scores above 0.96, accuracy exceeding 0.91, and MAE below 0.1, indicating superior predictive capabilities. Logistic Regression and SVM performed well but had slightly lower accuracy and AUC scores. KNN exhibited a moderately high MAE. Confusion matrices for each model revealed the distribution of true positives, false positives, true negatives, and false negatives. These results highlight the varying effectiveness of different machine learning classifiers in flood susceptibility modeling. Random Forest, Decision Tree, and Gradient Boosting demonstrated strong predictive power, making them suitable for high-accuracy flood risk mapping. The study underscores the importance of selecting appropriate machine learning models for disaster risk management. The findings provide valuable insights for achieving accurate and reliable flood predictions, emphasizing the critical role of model choice in effective flood susceptibility mapping.

ABSTRACT. The increasing frequency of flash floods due to climate change poses significant challenges for urban planning and disaster management. Rapid urbanization and associated land use changes in Wadi Ham, Fujairah, have heightened flood risks by increasing impervious surfaces and altering runoff characteristics. Accurate Land Use Land Cover (LULC) mapping is critical for effective hydrological modeling and flood forecasting. However, LULC mapping accuracy is influenced by various factors, including sensor resolution and data processing techniques, necessitating region-specific studies. This research aims to: 1. Evaluate the potential of high-resolution PlanetScope imagery for improving LULC mapping and classification. 2. Compare the performance of Maximum Likelihood Classification (MLC) and Random Forest (RF) classifiers using different remote sensing data sources. 3. Assess the implications of accurate LULC representation in hydraulic modeling. Our findings indicate that PlanetScope’s 3m resolution and superior temporal capabilities significantly enhance monitoring and mapping accuracy. The RF classifier outperformed MLC, achieving over 97% accuracy for the arid, ungauged watershed. These high-resolution LULC maps were crucial for improving hydrological models and flood forecasting accuracy. This research supports Abu Dhabi’s 2030 vision for resilient community development and aligns with the UN Sustainable Development Goals 11 (Sustainable Cities and Communities) and 13 (Climate Action). The insights gained will equip urban planners and policymakers with essential data to mitigate flash flood impacts effectively.

ABSTRACT. As global climate change intensifies, its impact on water resources poses an ever-growing challenge, particularly in the Ruzizi River Basin (RRB), a vital hydrological region within the Albertine Rift in the Great Lakes area of Africa. This study meticulously evaluates the implications of climate change on the basin's hydrological dynamics using the Soil and Water Assessment Tool (SWAT), supported by simulations under the SSP2-4.5 and SSP5-8.5 climate scenarios from the CMIP6 dataset. Employing advanced climate models for historical and future projections, alongside sensitivity analyses and rigorous calibration and validation phases, the study achieved a Nash-Sutcliffe Efficiency (NSE) of 0.64 and a Coefficient of Determination (R²) of 0.76 during calibration, with validation results showing an NSE of 0.70 and an R² of 0.74. Results indicated a projected decrease in mean annual precipitation by up to 35% under both scenarios by mid and late century, leading to significant reductions in water yield by nearly 50% and marginal variation in evapotranspiration, impacting water availability for various uses within the basin. This underscores the necessity for proactive and adaptive water resource management strategies to cope with anticipated hydrological changes. The study highlights the importance of developing responsive policies and infrastructure investments to bolster water resource resilience in the RRB, emphasizing the need for stakeholders to implement adaptive measures that mitigate the adverse effects of climate change. By providing a critical foundation for sustainable water management, this research calls for integrating climate projections into planning and adopting innovative approaches to enhance the region's adaptive capacity. Overall, the study offers valuable insights and practical recommendations for ensuring long-term water security in the RRB amidst an increasingly uncertain future.

ABSTRACT. This study investigates the collapse of diversion tunnels at an under-construction hydropower project site in Pakistan during an extreme flood event in August 2022. The flood, peaking at 7,370 cubic meters per second and delivering 790 million cubic meters of water over 72 hours, caused severe structural failures, including the overtopping and breaching of a temporary rockfill dyke and the collapse of headworks. The failures led to the submersion of tunnels, extensive debris inflow, and the destruction of key structures, including a 66-meter section of Diversion Tunnel-1.

Several root causes of the collapse were identified: the flood event's magnitude exceeded the flood frequency analysis, which did not account for such extreme conditions, leading to the dyke breach and tunnel failure. The protection dyke was inadequately designed for such high flood levels; a more robust concrete dyke might have mitigated the damage. The tunnels, lacking full concrete lining and proper hydraulic conditions, experienced irregular flow that eroded the rock and shotcrete. Insufficient ground support, including the absence of long tensioned rock anchors, and incomplete concrete lining further contributed to the collapse. Additionally, the inlet slope of Diversion Tunnel-1 had only partial installation of Tensioned Ground Anchors (TGAs), which was inadequate for maintaining stability during the flood and these factors collectively led to the failure of the diversion infrastructure. Recommendations include updating flood frequency analysis to include extreme events, implementing concrete linings for better protection, improving air inlet provisions, and ensuring full installation of TGAs for enhanced slope stability. The findings highlight the need for better design and construction practices to handle unprecedented climatic events and safeguard critical infrastructure.

ABSTRACT. Al-Ain experienced three significant flash flood events in 2024, with the February 12th storm recording over 100mm of rainfall. These events caused widespread flooding and property damage, highlighting the region's vulnerability to extreme weather. This study encompasses advanced remote sensing processing by applying a spectral index to Sentinel-2 images, along with a post-analysis using GIS by integrating precipitation records from 16 meteorological stations to identify and analyze flash flood areas in Al-Ain. The precipitation data allowed us to calibrate the event with the total flooded area to determine if more significant events would result in larger flooded areas. Additionally, the precipitation records helped determine the return period and identify the locations with the most significant rainfall. The data from the 16 meteorological stations were crucial in pinpointing the source of the floodwaters, the timing of the rainfall, and the differences between rainfall and runoff. The index extracts turbid waters common after flash floods due to heavy sediment loads, especially in highly deforested and desert lands like the Al-Ain region. During floods, turbid waters indicate inundation, while existing water bodies like oases and dams, which lack sediments, do not. Optical remote sensing differentiates floodwaters and other waters, unlike radar remote sensing. The Sentinel-2 images were used to map the flooded areas in Al-Ain city after the February 12th, 2024, event. The preprocessing of Sentinel images (imaged on February 17th, 2024) includes a resampling process, which transforms 20m pixel size data to 10m pixel size. This transformation does not affect the radiometric information contained in the resampled bands. Index band selection is based on water's spectral response: turbid waters reflect highly in the NIR band, the red band reduces vegetation effects, and the SWIR2 band makes soil and rocks appear dark like vegetation. Thus, the index was defined as follows: ITW=2*(Red +NIR)/ SWIR2 The index highlights all flooded areas, which appear bright, while other objects appear in shades of gray. By integrating these datasets, we were able to accurately map flooded areas from this event and elaborate future scenarios for more significant events. Our findings reveal that a total area of 7.19 km² was flooded. These flooded areas are divided into different types, 11% (0.79 km2) roads in the city, 27% (1.94 km2) obstacles formed by roads outside the city, 33% (2.37 km2) artificial obstacles to prevent flooding, 8% (0.57 km2) in the streams of the wadis and 21% (1.5 km2) natural depressions. The floods can be more significant if there is no early intervention to address topographic obstacles in the highlands over the opening of alluvial fans near the Oman borders. These results highlight the critical importance of effective urban planning and flood mitigation strategies to minimize future impacts in Al-Ain. The effects of climate change, urbanization, and changes in topography due to numerous public works exacerbate the region's susceptibility to flooding.

ABSTRACT. Finding areas that are susceptible to flooding is becoming more and more important as floods become more severe and frequent as a result of climate change and human activities. Using innovative modeling techniques, this study examines the susceptibility to flash floods in the Sedrata basin in north-eastern Algeria. Flash floods are a common occurrence in the area because of its steep topography, fast urban growth, and intense rainfall patterns. Accurate mapping of regions vulnerable to flash floods is necessary to reduce risks and offset flooding-related expenses. Using GIS-based metaheuristic optimization algorithms—more especially, the Grasshopper Optimization Algorithm (GOA) combined with the Random Forest (RF) machine learning algorithm, this work intends to map the regions of Sedrata that are susceptible to flash floods. 30 % of the twenty flash flood areas that were found in the flood inventory were used for testing, while the remaining 70%were used to train the model. Ten conditioning elements were taken considered, including aspect, curvature, elevation, slope, Topographic Wetness Index (TWI), Terrain Ruggedness Index (TRI), rainfall, stream density, distance to stream, and land use/cover. Using ArcGIS, estimated weights for these variables were assigned to produce the final flash flood susceptibility maps. With 96.6% accuracy, the RF-GOA model was implemented with remarkable success. The obtained results offer valuable guidance to decision-makers on where to focus future research efforts and how to generate practical flood control plans. This study adds to the growing body of information on flash flood risk assessment and shows how machine learning and optimization methods may be used to improve natural hazard mapping. The findings have implications for infrastructure development, emergency preparedness, and urban design in regions at risk of flooding.

ABSTRACT. Coastal zones comprise a large variety of ecosystems. They play a crucial role in the balance and regularity of the Earth's climate, but are often vulnerable to climate change particularly sea level rise. The negative effects of SLR are becoming a major concern in coastal areas around the world due to the challenges posed through climate change. This phenomenon poses a major threat to coastal systems, causing flooding and the risk of coastal erosion. Morocco has 3,500 km of coastline on both the Atlantic and the Mediterranean. It’s confronted to the problems of rising sea levels. Already, many Moroccan beaches are undergoing intensive erosion or have disappeared on the Atlantic coast. According to research already carried out by 2100, 72% of beaches could be eroded. Our study aims to assess the vulnerability of an area of the Atlantic coast to SLR. To this end, we used a qualitative model and a geographic information system to simulate and map coastal vulnerability in order to identify the hottest spots on the coast. This model is based on a qualitative index of the relative vulnerability of each coastal point to erosion and marine flooding. The model calculates exposure indices using spatial representations of bio-geophysical variables. These parameters are preprocessed and classified according to hazard level, ranging from 1 (very low) to 5 (very high). Coastal vulnerability modeling has been carried out in the region to determine the coast's exposure to climate change. The results showed that the highest hazard classification was > 3, 25, the classification ranged between 3.25 and 2.41, and the lowest hazard classification ranged from < 2, 41. The research results can identify the districts along the coastline at greater risk of marine disasters and provide scientific and theoretical support for coastal protection and sustainable development.

ABSTRACT. This research aims to modele the relationship between flow and rainfall in the Zardaza catchment in the north-east of Algerian using artificial intelligence techniques; namely Support Vector Regression (SVR) and Long Short Term Memory (LSTM). The challenge is to predict the nonlinear correlations between parameters. To manage this non-linearity we use minimal amount of observed data such in Zardezas basin. After standardization of the input features and comapring with the observed flow, one notes a strong correlations between rainfall and flow rates. The Nash coefficient and root mean square errors (MSE) confirm it.

ABSTRACT. The Medjerda basin in North Africa faces significant challenges related to soil erosion and sediment transport, impacting dam efficiency and reservoir sedimentation. The area experiences frequent flash floods, has fragile soil composition, and lacks vegetation cover, all contributing to high sediment output rates. These factors make the watershed more susceptible to erosion, emphasizing the need to understand and predict these patterns for effective management.This research is crucial for establishing measures to mitigate the effects of erosion and sedimentation on water resource infrastructure, especially considering that these challenges may worsen due to climate change.Advanced machine learning techniques, namely: Gradient-Boosted Decision Trees (GBDT) and Memetic Programming (MP) were employed to address this. These approaches were used to forecast the vulnerability to erosion and potential sediment yield in the Medjerda basin under different climate change scenarios, including changes in vegetation cover management. The study integrated eight primary variables into its research: digital elevation model, stream power index, slope length, topographic wetness index, rainfall erosivity, soil erodibility factor, cover management, and distance from the river basin. These variables were derived from extensive field research, grid data analysis, and examination of digital elevation models (DEMs). The study utilized the most recent CMIP6 data to incorporate different climatic scenarios, focusing on monthly total precipitation for three selected future periods: 2021-2040, 2041-2060, and 2061-2080. The investigation considered the SSP126 and SSP370 scenarios obtained from three prominent General Circulation Models (GCMs): MPI-ESM1-2-HR, EC-Earth3-Veg, and IPSL-CM6A-LR. According to the forecasts, climate change is expected to intensify current conditions by increasing the frequency and severity of hydrologic extremes, worsening concerns related to soil erosion and sediment transport. The research conducted a systematic dataset partitioning, allocating 70% of the data for model training and reserving the remaining 30% for testing. An assessment of model performance using Receiver Operating Characteristic (ROC) curves showed high precision for both models. The GBDT model achieved the highest level of accuracy, with an AUC of 0.975, followed by the MP model, with an AUC of 0.960. The study identified slope length and cover management as the most significant predictors of erosion vulnerability.These findings underscore the importance of the GBDT model in predicting soil erosion in the Medjerda basin, providing valuable insights for dam operators and water resource managers. The study highlights the urgent need for adaptive management techniques to address the escalating risks associated with climate change and mitigate the consequences of erosion and sediment transport in this watershed, promoting sustainable water resource management.

ABSTRACT. Sediment Bypass Tunnels (SBTs) have been demonstrated to effectively mitigate or eliminate reservoir sedimentation by diverting sediment-laden flows around reservoir dams, directing them to downstream river reaches. Typically, SBTs convey high-velocity water-laden floods, sometimes exceeding velocities of 10 m/s, to fulfill their intended function. However, these elevated velocities result in significant abrasion damage to tunnel floors. The prediction of such abrasion is further complicated by the complex interaction between flow hydraulics and sediment transport processes, coupled with the limited availability of high-quality data. This study explores the predictive capabilities of novel machine learning (ML) models for assessing abrasion damage in SBT floors. We focus on evaluating the influence of input features for both novel models: the 2D Abrasion Susceptibility Model (ASM) and 3D Abrasion Susceptibility Depth Model (ASDM) which indicate the damage extent and detail damage extent plus abrasion depth, respectively. We focused on a case study on the Koshibu SBT in Japan. The tunnel, approximately 4000 meters in length and 7 meters in width was scanned using advanced laser technology, capturing abrasion damages with cloud points every 2 cm across the entire floor. We investigated the influence of these parameters through three experimental setups: the entire tunnel, the straight part only (Part A), and the curved part (Part B). Each experiment was further dissected into three evaluation scenarios: using only geometric features (S1), only hydraulic features (S2), and a combination of both (S3). The ASM model evaluations employed indices such as accuracy, sensitivity, and Specificity while the ASDM model assessments utilized the correlation coefficient (R) and Root Mean Squared Error (RMSE) yielding values of 0.86, 0.87, 0.86, 0.86, 0.04 for accuracy, sensitivity, Specificity, R, and RMSE respectively for the entire tunnel. The findings indicate variances in models’ performance between straight and curved tunnel sections. In part (A), the models exhibit a slightly lower accuracy, sensitivity and R, alongside RMSE suggesting that the ML models are hardly adept at capturing all complex abrasion patterns. Specificity remained consistently high, while RMSE was minimal across all scenarios, signifying a low rate of false predictions. While, in part (B), all indices exhibited enhanced performance across all scenarios, which may be attributable to the more obvious patterns. The relative impact of geometric versus hydraulic features on model accuracy reveals that models incorporating both types of features (S3) outperform those using only one type (S1 or S2). Geometric features alone (S1) provide markedly enhanced accuracy than hydraulic features alone (S2), underscoring the significance of the tunnel’s outline configuration in abrasion damage prediction. Our findings reveal significant variances in the impact of input parameters across different tunnel sections and events, highlighting the complex interplay between geometric and hydraulic factors in predicting abrasion damage. The results contribute to the growing body of knowledge on SBTs and present a robust framework for future research and practical applications in reservoir sedimentation management.

ABSTRACT. Recently, remote sensing has been considered an indispensable tool in hydrology applications and water resources management. The instrumental role of remote sensing implicates adeptly delineating land use and land cover, a foundational layer of flash flood management. The first and primary step of this study is the quality assurance of the land use and land cover utilizing Sentinel 2 and Landsat 8 satellite imagery. The accuracy assessment procedure will be employed to discern and compare the efficiency of these two prominent satellite products. The Gharbia Governorate, Egypt, discerned by its four distinct aspects encompassing Built Area, Water, Crops, and Agricultural Land, was chosen for this phase of the study. Subsequent, a vital application of remote sensing, this study presents a comprehensive analysis of flash flood vulnerability in Wadis adopting the satellite imagery of the superior product, GIS, and a precise land use land cover (LULC) valuation to ensure reliability. Morphometric assessments, considering soil, geology, drainage, slope, LULC, and elevation layers, provide an inclusive understanding of vulnerability factors. All these layers are clustered and over-weighted installing the Analytic Hierarchy Process (AHP) to delineate the flashflood vulnerability zones in Wadi El-Assiuty, Egypt. Wadi El-Assiuty, the rainfall catchment area, is particularly susceptible to flash floods due to its topographical, climatic and geographical features, proximity to the red mountains. The integration of these layers simulates the topographical and hydrological features influencing flash flood vulnerability. The results revealed the superiority of the Sentinel satellite over its Landsat counterpart, highlighting the significant accuracy of Sentinel-2 in interpreting and analyzing the LULC layer. Furthermore, the flash flood risk map indicated that most areas in Wadi El-Assiuty exhibit moderate to low flooding susceptibility. The findings contribute to scientific knowledge and offer practical insights for decision-makers and planners, emphasizing the importance of LULC accuracy for reliable vulnerability assessments. The integrated approach provides a robust framework for managing flash flood vulnerability in wadis, with implications for land-use planning and disaster preparedness in arid regions.

ABSTRACT. Flash floods, influenced by seasonal factors and climate change, present significant challenges in various regions worldwide. A notable dam break occurred at the Derna dam in Libya in September 2023, resulting in flooding downstream and the loss of over 10,000 lives. In the Middle East, Iran is one of the countries facing severe flash floods and potential damage to hydraulic structures. Reservoir dams, such as the Golestan Dam in the river basin, played a vital role in lessening the impact of flash floods, as seen in the August 2001 flood. In March 2019, a severe flash flood, the worst in a decade, caused the Maruak Dam in Lorestan and Voshmgir Dam on the Gorganrud River in Golestan to overtop, leading to substantial financial losses, structural damage, and fatalities. Dam weirs play a significant role in efficiently managing excess water during floods, releasing it safely downstream to prevent dam failure. Careful management is required when water flows rapidly through weirs to prevent damage; this involves reducing its speed and distributing it evenly to ensure safe and efficient use. Sharp-crested weirs in rivers and irrigation channels are important in regulating water flow and can also be used to dilute water for intake. Slit weirs, a type of sharp-crested weir, have a contraction ratio (b/B) less than 0.25, where b represents the opening width, and B is the total width. While research on discharge coefficients of slit weirs exists, there is no study regarding energy dissipation mechanisms downstream of slit weirs. This study aims to explore energy dissipation in rectangular slit weirs, focusing on how contraction ratio (b/B) and flow discharge (Q) affect relative energy dissipation (ΔEr) and residual energy (E1/E0). The study was conducted in a laboratory setup, testing models with various contraction ratios, including b/B=1/12, 1/8, 1/6, and 5/24, under different flow discharges. Dimensional analysis was applied, and b/B (channel width to the slit width) and h/P (head on the weir to the slit weir height) were found to be effective parameters for ΔEr. Graphs of ΔEr and residual Energy (E1/E0) about h/P and discharge per unit width (q) were compared with b/B. The results demonstrate that relative energy dissipation (ΔEr) decreases as the head over the weir increases. A higher discharge rate (q) correlates with a reduction in ΔEr, indicating more efficient discharge. Higher flow rates and larger contraction ratios lead to lower energy losses in rectangular slit weirs, highlighting the importance of weir geometry and flow conditions on energy dissipation efficiency. Additionally, increased discharge rates result in higher relative residual energy (E1/E0), suggesting better energy retention downstream, which aligns with previous findings. Maximum values for E1/E0 were observed at contraction ratios of b/B = 5/24 and 1/6, with values of 0.72 and 0.619, respectively, indicating that wider slit weirs are more effective at minimizing energy losses. The results emphasize the relationship between head, discharge, and contraction ratio in managing flow energy and ensuring stability. Incorrect weir design can exacerbate downstream flood damage.

ABSTRACT. Flash floods in agricultural areas such as Punjab province in Pakistan are of increasing concern. Their consequences are often dangerous for human lives and have devastating effects on the local economy. There is often severe loss of property, land, and agricultural production. Assessing the impact of hazards such as flash floods can help estimate probable loss in crops. In this study, we investigated the efficiency of remotely sensed microwave data to map the croplands affected by the flash flood that occurred over the past four years in Nankana district, Punjab, Pakistan. The Nankana district had 10.3% of its total geographical area under flood. We used two machine learning algorithms, random forest and support vector regressor, to forecast wheat crop production and potential loss during the rainy seasons of 2020–2024. A regression algorithm with five predictor variables, including cropland area, two vegetation indices, and two climatic parameters, was applied to forecast the wheat production in the area. Comparing the two algorithms, the random forest showed significant performance. The random forest regressor estimated the production of wheat with an R² of more than 0.8 in the district. The mean absolute error (MAE) was 12.5 quintals per hectare, and the root mean squared error (RMSE) was 15.7 quintals per hectare. The maximum production loss of wheat was estimated at 54.13%, equating to approximately 4,500 metric tons of wheat lost over the affected areas. Further analysis indicated that approximately 15,000 hectares of cropland were directly impacted by the floods, leading to an estimated loss of 20,000 metric tons of wheat production annually. These findings underscore the utility of the proposed approach for a quick in-season forecast on crop production loss due to climatic hazards. By integrating remote sensing data with machine learning techniques, this study provides valuable insights that can enhance the accuracy of agricultural forecasts. The results can significantly contribute to disaster preparedness and mitigation strategies. By providing timely and precise information on potential crop losses, this approach can aid farmers and policymakers in making informed decisions to minimize economic damage and improve resilience against future climatic events.

ABSTRACT. Urban geomorphology is considered one of the most important applied aspects of geomorphology, as it deals with the study of different landforms, the study of the natural geomorphological processes that appear in the study area (Salaha, 1994), and the study of the dangers of floods to which areas of urban expansion are exposed in terms of type, their effects on the region, and the possibility of predicting them to ward off their dangers (El-Delemy, 2012). The city of Safaga occupies a distinguished location with unique geomorphological features. Its location as a coastal city on the Red Sea coast has given it great importance, as the city is distinguished by the presence of several international ports in it, which represent the main artery of maritime transport for the Egyptian state in the Red Sea region. The city is also located on the international trade line of the Suez Canal, which Making it the focus of attention of the Egyptian state in the Golden Triangle project for mineral wealth (Public Authority of Urban planning, 2014) The city is considered the main port of the Egyptian state on the Red Sea coast and is considered one of the most important commercial, logistical and mining cities (El-zokaa, 1997), in addition to its central location in relation to the Red Sea Governorate (Public Authority of Urban planning, 2014). The UAE is also investing more than $200 million in the Safaga mining port as direct investments in the infrastructure of the Safaga mining port (A D Ports Group, 2023). This research paper aims to present a scientific study to planners and decision makers about the dangers to which the Safaga mining port is exposed and the dangers it may be exposed to in terms of sudden floods that could lead to the destruction of the port’s infrastructure and what it could cause to the loss of huge investments for either the Egyptian state or the United Arab Emirates. The research paper also aims to present geographical expertise as an important and helpful factor in analyzing the dangers of floods to which the infrastructure of the Safaga mining port is exposed, and to provide the best suggestions to confront that danger, reduce its effects, or avoid it if it was not possible to confront it.

ABSTRACT. Floods are classified as one of the hydrological hazards that affect many countries across the world. With most weather-related disasters occurring in developing countries, demographics and socio-economic pattern changes of these countries have contributed too many losses relating to water-related disasters such as floods. South Africa is among the developing countries affected by natural disasters most frequently, in particular, flood disasters. Thus, this study assessed the root causes of flood risks and the impacts of these floods on the communities of Bronville and Hani-park in Welkom in the Free State Province in South Africa. The study adopted a mixed method approach that included both quantitative (semi-structured questionnaire used to collect the data) and qualitative approaches. In addition, photographic evidence was used to demonstrate the environmental impacts of floods in the study area. The results of this study revealed that lack of stormwater drainage systems, absence of drainage in some areas, groundwater flooding and sewage were the root causes of flood in the study areas. The findings of this study underscore the urgent need for comprehensive and sustainable flood mitigation strategies in vulnerable communities like Bronville and Hani-park. The root causes identified, including inadequate stormwater drainage systems and groundwater flooding, illuminate critical areas for intervention and investment. Furthermore, the documented impacts of floods on these communities serve as a poignant reminder of the disproportionate burden borne by developing regions in the face of hydrological hazards. By integrating quantitative and qualitative methodologies, this research provides a nuanced understanding of the complex dynamics at play, facilitating informed decision-making for policymakers, urban planners, and community stakeholders alike. Moving forward, concerted efforts must be made to prioritize resilience-building measures, enhance infrastructure resilience, and foster community engagement to mitigate the devastating effects of floods on vulnerable populations. Only through collaborative action and proactive measures can we aspire to safeguard lives, livelihoods, and the environment against the ever-looming threat of flood disasters.

ABSTRACT. Floods and flash floods are considered severe natural disaster hazards and have had significant impacts on human life throughout history. They can destroy infrastructures, cause landslides, damage agricultural fields, and cause injury or death to both livestock and humankind. The number of flood and flash flood events has significantly increased all around the world over the last three decades. As a result of insufficient measures adopted by the local governmental authorities in the Petra region such as preparedness, planning, and mitigation measures, lack of knowledge, awareness, and capacity-building capacities among local communities; the Petra Region has been exposed to many decades to floods and flash flood risk, which posed severe impacts at the human life and economic situation in the region. While the frequencies and impacts of flash floods might not be controlled easily, the need for more effective efforts has become extremely important. Therefore, the development of a Community-Centred Flash Floods Early Warning System (EWS) Guidance as a tool that might enhance the local authorities and communities to mitigate flash flood risks and impacts in the Petra Region. This tool is a key factor in which the mitigation plan of the flash flood of the concerned parties might succeed. EWS a system used by the Petra Development and Tourism Region Authority to issue warnings to local concern parties in the region and the local community of the risk of flash floods so that appropriate actions can be taken to avoid the impacts that may be caused by flash floods. Rain gauge stations measure the amounts of precipitation and send them to the main control station which in turn analyzes and issues the appropriate warning. EWS for flash flood guidance is probably the most important tool that might contribute to reducing flash flood impacts and vulnerability at all levels in the Petra region. The guidance would be a key element that can provide the necessary information and strategies to the concerned authorities and communities that are at risk to enable them to be proactive, better prepared, equip them of the knowledge of the hazards at the locality, community vulnerabilities, and imminent risks and disasters, to receive warning messages, and mobilize their response capabilities to reduce risks. Furthermore, it helps to reduce economic losses by allowing people to better protect their assets and livelihoods. Early warning information allows people to make decisions that contribute to their economic self-sufficiency and their countries’ sustainable development.

ABSTRACT. Mediterranean drylands are vulnerable to land degradation, desertification, and drought. Implementing water harvesting technologies in diverse living labs (LL) and outscaling these solutions across the Mediterranean is crucial for sustainable water and land management. This study presents a comprehensive, science-based approach to the potential outscaling of three water harvesting technologies: subsurface water retention technology (SWRT), managed aquifer recharge (MAR), and levelled terraces (LT), implemented in Moroccan, Tunisian, and Egyptian LLs, respectively, under the SALAM-MED PRIMA project. We conducted spatial analysis by first engaging with experts responsible for each technology and LL to identify relevant implementation criteria, environmental problems related to land and water management, and the effectiveness of the technologies. Using the Land System Archetypes methodology, we combined maps of problem areas and suitability with effectiveness scores assigned by technology experts through a weighted sum approach. The resulting outscaling maps indicate that SWRT, MAR, and LT are generally recommended to highly recommended for most of Morocco, Tunisia, northern Algeria, and parts of Egypt, and moderately recommended for Spain and parts of Italy and Greece. Our findings will be presented to stakeholders involved in the SALAM-MED project across eight Mediterranean countries, aiming to promote the uptake of these technologies based on the out-scalability maps. This valuable insight can be useful for all stakeholders and decision-makers in the Mediterranean region to enhance sustainable land, soil, and water management.

ABSTRACT. Understanding the fundamental elements of time series data, such as trends, periodicities, leaps, and randomness, is essential to make effective analysis. For a variety of applications in management, operation, planning, and maintenance works, these elements are essential. Finding trends in time series data becomes crucial in the context of climate change impact research because long-term fluctuations are the main emphasis. In this study, a new hybrid approach called Probabilistic Analysis of Innovative Trends (PITA) which is a combination of Innovative Trend Analysis technique along with statistical analysis is used to identify and evaluate the annual precipitation trend over 41 years (1982-2023) at four meteorological stations located on the northern coast of Algeria. The study found that two stations named Annaba and Skikda present a relatively stable precipitation regime with slight decrease in the received annual amounts. This decrease should be further analyzed especially in the context of climate change and its direct impact on droughts in the North of Africa including Algeria. In contrast, the other two stations, Jijel and Bejaia showed significant increasing trends with flagrant perturbation in the annual precipitation regime with a higher likelihood of having extreme events. This study serves as a scheme, demonstrating the PITA method's application which could be used to analyze various data types across different regions without any restrictive assumptions.

ABSTRACT. Effective sediment management techniques are crucial for long-term dam utilization. A sediment bypass tunnel (SBT) is an innovative method for passing through sediment from upstream to downstream during floods. Accurate predictions of sediment inflow (Suspended Sediment Concentration SSC) and timing are important for efficiently operating the SBT. The Miwa Dam in Japan is one of the dams where SBT operates. Due to the high sediment production in its basin, sediment control facilities such as the SBT and stockyard are operating. The SBT at Miwa Dam is designed to transport suspended sediment downstream. There are two challenges in the SBT operation. First, the water inflow (Q) and sediment concentration do not correspond to one-to-one. The second challenge is the trade-off with water use. When the water storage recovery is prioritized, the SBT cannot be operated. For the second challenge, ensemble rainfall forecasts, which provide longer prediction periods and allow for the assessment of forecast reliability, could be a potential solution. We set the objectives of this study as follows. The first is to analyze the dynamics of SSC during floods. The second is to enhance the operation of the SBT by predicting Q and SSC using ensemble rainfall forecasts. The research methodology is as follows. To understand sediment dynamics during floods, we analyzed turbidity data. We plotted Q and SSC and analyzed hysteresis loops and developed Q-SSC correlation formulas. Next, we constructed the SWAT model as the runoff model and used ensemble rainfall forecasts to predict Q. We also predicted SSC by applying the correlation formula to the predicted Q. In this study, we utilized the JWA ensemble rainfall forecasts. Forecast members were ranked based on cumulative rainfall and selectively averaged. We used the average forecasts to predict Q and SSC. Operational rules based on these predictions were established, and the efficiency of the SBT operation was evaluated. The study focused on the flood event of June 2, 2023. The analysis of Q-SSC hysteresis loops found that small-scale floods exhibited high SSC in the earlier period of floods. In large-scale floods, more complex behavior was observed. Ensemble forecasts have confirmed that it is possible to detect floods early. Furthermore, these forecasts can help determine the timing of increases in Q. However, predictions for SSC using the correlation formula underestimated the peak values. As a result of examining SBT operation rules, it was found that by deciding the start of SBT operation based on Q predictions, the risk of delayed operation during the early stages of floods can be avoided. Additionally, by ending the SBT operation based on predictions of water level recovery, it is possible to ensure water level recovery after the operation ends and maximize the duration of the SBT operation. This study confirmed the characteristics of sediment dynamics in small-scale and large-scale floods by analyzing hysteresis loops. It was also demonstrated that operational rules based on ensemble rainfall forecasts enable rational SBT operation. Future challenges include improving prediction methods, as the current approach could not forecast complex sediment dynamics.

ABSTRACT. Amidst the escalating impacts of climate change, the strategic operation of reservoirs in the Ruzizi River Basin (RRB), Africa, presents a pivotal mechanism for mitigating the transition from meteorological to hydrological extremes, thereby ensuring regional water security and supporting sustainable development. Utilizing the Soil and Water Assessment Tool (SWAT) and climate projections from CMIP6 scenarios SSP2-4.5 and SSP5-8.5, this study analyzes the historical (1983 – 2020) and future (2040 – 2100) impacts of reservoir management on hydrological responses. The research aims to quantify the influence of reservoir operations on the correlation between meteorological and hydrological extremes, employing the Standardized Precipitation Index (SPI) and the Standardized Streamflow Index (SSI). Our findings reveal that reservoir operations significantly modulate the transmission of extremes, effectively delaying and altering the progression from meteorological to hydrological extremes. During periods of reservoir influence, substantial control over short-term extremes was observed, while the impact on long-term events was more limited. In upstream areas of the basin, where reservoir influence is more pronounced, a mitigative effect on long-term dry extremes was evident. This study underscores the critical role of integrated reservoir management in enhancing water-related hazard and risk management under future climatic uncertainties, providing a quantifiable assessment of how strategic reservoir operations can stabilize hydrological extremes and sustain regional water security. By integrating advanced climate models and detailed hydrological simulations, the research offers valuable insights into optimizing reservoir operations to mitigate the adverse effects of climate change on water resources, highlighting the necessity of adaptive and proactive water management strategies to ensure sustainable development and water security in the RRB.

ABSTRACT. The aim of this project is to use artificial intelligence techniques, namely Decision of Tree (D-T) and Gaussian process regression models (GPR), to predict the link between flow and rainfall in the watershed of Zardeza in northern Algeria. Predicting nonlinear correlations between the components under investigation is the goal at hand, and gathering the necessary data might be time-consuming. It becomes necessary, therefore, to employ models that can handle this non-linearity and need a small number of parameters. This argument argues that flow rates in gauged male rivers, like the Zardezas basin, be anticipated using the artificial intelligence technique Longue Short Term Memory, which has shown effectiveness in various scientific domains.The rains and flows show a strong association with mean square errors (MSE) and NASH, supporting.

ABSTRACT. Accurate flood prediction is essential for effective water resource management in arid regions characterized by wadi systems, where zero-inflated runoff data presents significant challenges. This study presents a novel hybrid model that integrates classification and regression techniques to enhance predictive accuracy. The model's two-stage approach first utilizes a classifier to identify runoff events, followed by a regressor applied to the identified runoff instances. To further refine performance, Variational Mode Decomposition (VMD) was used for preprocessing the runoff data, optimizing it for the Extreme Learning Model (ELM). The evaluation of the model using Nash-Sutcliffe Efficiency (NSE) revealed a baseline NSE of 0.49 for the standalone regressor. The hybrid model demonstrated a marked improvement with an NSE of 0.55. Notably, VMD with varying mode numbers achieved exceptional results: NSE of 0.64 for VMD2, 0.97 for VMD4, 0.98 for VMD6 and VMD8, and a peak NSE of 0.99 for VMD10, underscoring its superior capability in managing complex runoff data. In summary, the proposed hybrid model, when enhanced with VMD, significantly advances flood prediction accuracy in arid wadi systems, effectively mitigating challenges posed by zero-inflated datasets.

ABSTRACT. Natural disasters are considered one of humanity's most challenging issues, with devastating consequences from natural phenomena such as floods, which cause enormous human and material losses. This research focuses on the city of Derna on the Libyan coast, tragically famous for the unprecedented September 11, 2023 flood disaster. Hurricane Daniel caused heavy rains, which broke up the two dams and caused catastrophic flooding in the region. This breach resulted in loss of life, significant demographic losses, and intense psychological distress among the local population. This study examines the dynamics of this event by estimating the runoff volume generated in the Derna river basin and exploring the factors that led to the dam failure and their role in Derna's devastating floods. The storm's origin and rainfall distribution in the watershed were analyzed. Surface runoff was simulated using satellite products and the Hydrological Modeling System (HEC-HMS). The study reveals that even if the dams had survived, they would have been ineffective in controlling the sudden floods in the city due to the extreme nature of the runoff from Storm Daniel, which is estimated at 311.02 mm of rain in 24 hours. The results highlight the importance of conducting detailed hydrometeorological assessments and adopting flood management practices in areas sensitive to the effects of climate change. They also recommend reducing excessive reliance on structural flood control measures.

ABSTRACT. The Flood risk modelling 2D have come an important tool to evaluate and determine flood prone area. The main input to 2D flood models is the Digital Elevation Model (DEM). The resolution and quality of DEMs have an impact on flood modelling outputs. This study focus on the analysis of the impact of DEM resolution on flood modelling using DEMs with resolutions 1m (LiDAR data) and 30 m (the freely available NASA Shuttle Radar Topography Mission). Data obtained during the flood event caused by the Storm in year 2015 in the section Jendouba-Bousalam from Medjerda river (northern Tunisia) was used for this study. All of the simulations were conducted using the HEC-RAS 2D model. Further, a satellite image radar Sentinel 1 was used to determine the flood extent. The models obtained with 1m and 30m resolutions DEM were compared with the flood extent obtained from processing of Satellite image radar Sentinel 1. Results show that the flood extent simulated by HECRAS 2D model for 1m and 30m resolutions DEM and Sentinel 1 are 2350 ha, 1500 ha and 1350 ha respectively. The validation of the hydraulic model with 1m DEM resolution was conducted by comparison with the flood extent obtained from Sentinel 1. The statistical index F1 and F2 values are 0.8 and 0.65 respectively. This work constitutes an important support to decision makers for building a Flood Prevention Plan.

ABSTRACT. Flash floods are characterized by short durations, small areal extents, high flood peaks, rapid flows, and heavy disastrous consequences (human and material losses). They usually develop at space and time scales that conventional observation systems are not able to monitor for rainfall and river discharge. Consequently, the atmospheric and hydrological generating mechanisms of flash-floods are poorly understood, leading to highly uncertain forecasts of these events. This study is interested in the characterization of flood hazard in the Gabes Catchment (Southeastern Tunisia), considered as an important step for flood management in the region. A spatial database was first developed based on geological map, digital elevation model, land use, and rainfall data in order to evaluate the different factors susceptible to affect flood analysis. Next, the Analytical Hierarchy Process (AHP) method was applied for flood risk mapping. AHP analysis considered six variables, including elevation, slope, land use, drainage density, litho-facies, and rainfall. Monte Carlo simulation-aided analytic hierarchy (MC–AHP) was also used to quantify the sensitivity and minimize uncertainty and subjectivity associated with the AHP model. Both AHP and MC–AHP models gave similar results. However, compared to AHP approach, MC–AHP confidence intervals (95%) of the overall scores had small overlaps. Results obtained were validated by remote sensing data for the zones that showed very high flood hazard during the historic extreme rainfall event of June 2014, characterized by a return period of 50 years. The obtained MC–AHP results show that 33.5% of the basin area is characterized by a high to a very high flooding hazard. Land use was shown to be the most important factor contributing to flood hazard. The examined flooded areas proved that among 226 observed flood samples, 209 zones are characterized by a high to a very high susceptibility hazard. 96.3% of the observed flooded areas are located downtown in the highly urbanized area of Gabes City, close to streams and surrounding anarchic urban settings. The obtained results are considered as decision-aid tools for proper flood management.

ABSTRACT. Machine learning (ML) techniques, including genetic programming, boosting methods like gradient boosting, and random forests, are proving highly effective in addressing water-related challenges. This paper explores the application of these advanced ML algorithms for flood susceptibility mapping, flood depth prediction, suspended sediment concentration (SSC) prediction, flow discharge modeling, and rainfall analysis. By processing large-scale hydrological data and satellite imagery, these techniques offer accurate, real-time predictions, providing critical insights for early warning systems. Genetic programming allows for the automatic evolution of models capable of identifying complex patterns in water systems, improving prediction accuracy in diverse environments. Boosting methods further enhance model performance by reducing bias and variance, making them ideal for predicting flood susceptibility and flood depth mapping. By integrating machine learning with hydrological based models, this AI approaches also supports the development of real-time monitoring systems, crucial for disaster preparedness and flood risk management. Overall, this paper demonstrates the transformative role of ML, genetic programming, and boosting methods in delivering data-driven, efficient solutions to water-related hazards, enhancing both real-time and future prediction.

ABSTRACT. This study examines rainfall storm characteristics in the Al-Qassim administrative region of Saudi Arabia, with a focus on storm duration, start times, and cumulative rainfall depth distribution. The primary objective is to compare the developed SCS Type II curve for the region with the standard SCS Type II curve, which is widely utilized in the United States, and to assess the implications of these differences for hydrological modeling. Storm duration data were analyzed in 2-hour increments, and storm start times were divided into 3-hour intervals. It was observed that a significant proportion of storms initiated during the early morning hours, particularly between midnight and 3 AM, indicating a distinct temporal pattern in storm activity in the region. This temporal concentration of storms may have implications for flood risk management and hydrological predictions. A comparison between the developed SCS Type II curve for Al-Qassim and the standard SCS Type II curve reveals notable deviations. The developed curve demonstrates a higher percentage of cumulative rainfall occurring earlier in the storm event, indicating a more rapid accumulation of rainfall compared to the standard curve. This suggests that the application of the standard SCS Type II curve to the region may not accurately represent local storm dynamics, potentially leading to under- or overestimation of runoff. In addition, the cumulative rainfall depth distributions for storms of varying durations were examined. Storms were categorized into four groups based on duration (0-6 hours, 6-11 hours, 11-18 hours, and 18-24 hours). For each group, cumulative depth percentage was plotted against cumulative duration percentage. The results show that shorter storms tend to have a steeper rainfall accumulation curve, while longer storms exhibit a more gradual and evenly distributed rainfall pattern. The findings indicate that storm characteristics in the Al-Qassim region differ significantly from those represented by the standard SCS Type II curve. These regional variations highlight the necessity of localized hydrological models to improve accuracy in predicting stormwater runoff and flood risks. It is recommended that further research be conducted to refine these models and develop region-specific stormwater management strategies.

ABSTRACT. Sedimentation is a persistent challenge in the management of reservoirs and dams, significantly impacting their storage capacity, operational efficiency, and lifespan. One effective method to address this issue is the use of low invert level structures, such as under-sluices, for sediment flushing. The flushing of sediment through low invert level structures like under-sluices is a critical aspect of water infrastructure management. Despite its importance, the topic remains poorly understood, and a comprehensive review of existing literature is lacking. This paper provides a systematic review of current knowledge on sediment flushing through under-sluices, synthesizing findings from physical modelling, numerical simulations, and field studies. The review examines the factors influencing sediment transport, the impact of structure design and operation on flushing efficiency, and the environmental implications of sediment removal. Gaps in current understanding are identified, and directions for future research are proposed. This review aims to provide a foundation for the development of evidence-based guidelines for the design and operation of low invert level structures, ultimately supporting the sustainable management of water resources.

ABSTRACT. The Integrated Early Warning System (IEWS) for Water-Climate-Agriculture-Energy Security is a multifaceted approach aimed at enhancing the resilience and sustainability of interconnected systems critical to human and environmental well-being. Our developed IEWS leverages advanced technologies and data analytics to monitor, predict, and respond to environmental and socio-economic challenges. By integrating real-time data from climate models, hydrological forecasts, agricultural monitoring, and energy consumption patterns, the IEWS provides comprehensive and timely information to stakeholders. This facilitates proactive decision-making, risk mitigation, and adaptive management strategies across sectors. The system's primary objectives include improving the accuracy of forecasts, promoting efficient resource use, and safeguarding food and water security amidst changing climate conditions. Through collaborative efforts and continuous innovation, we aim to support sustainable development and enhance the resilience of communities worldwide.

ABSTRACT. the United Arab Emirates. Machine learning helps with flood mapping and prediction, while artificial intelligence improves early warning systems and satellite-based weather forecasts. Natural techniques for mitigating floods, like permeable pavements and wireless sensor networks, provide long-term advantages for society and the environment. The UAE is committed to addressing these issues, as evidenced by programs like the AI-powered GIS and the National Early Warning System, even in the face of obstacles like resource scarcity . Notwithstanding obstacles, these innovative methods provide encouraging prospects to enhance flood resistance in the United Arab Emirates I. Innovative Solutions for Enhancing Flood Risk Management The United Arab Emirates (UAE), though situated in a hot and arid environment, has been increasingly threatened in recent years by flash floods and strong rainfall. Innovative strategies have been created

and put into action to solve these issues and support flood risk control initiatives. Enhancing preparedness, mitigation, response, and recovery efforts, these solutions make use of technological innovations and natural methods. The application of artificial intelligence (AI) to several facets of flood risk management is a significant area of innovation. According to studies by Dewitte et al. (2021) and reports from the World Economic Forum (2023), AI-powered satellites make weather forecasting faster and more precise. Artificial intelligence (AI) improves early warning systems by evaluating massive volumes of data from sources like satellite imaging and weather monitors, enabling prompt notifications and proactive steps to reduce the danger of flooding. Flood mapping and prediction capacities have also been transformed by AI-integrated geographic information systems (GIS) and machine learning predictive modeling. Research by Hagos et al. (2022) and Razali et al. (2020) show how these technologies II. Application and Benefits of Innovative Solutions in the UAE The United Arab Emirates has introduced multiple inventive approaches to augment its capacity for managing flood risks. AI tools are used by the National Emergency Crisis and Disaster Management Authority (NCEMA) to run the National Early Warning System, which protects people and property by detecting storm motions and issuing timely warnings. According to Husain (2023), this method shows the UAE's dedication to using technology to improve planning and response for disasters. In order to increase flood mapping and prediction accuracy, the UAE has also adopted AI-powered GIS and machine learning predictive modeling. The government of the United Arab Emirates has improved its ability to recognize and reduce flood risks in susceptible areas by collaborating with technology companies and research centers. These programs support the UAE's overarching objectives of encouraging sustainability and adaptability in the face of climate change. In the United Arab Emirates, initiatives to protect and restore natural ecosystems in order to reduce the danger of flooding have also gained support as nature-based solutions. According to Naismith (2023), the UAE's investment in NBS demonstrates its dedication to environmentally responsible development and sustainable development. The United Arab Emirates (UAE) aims to protect its natural heritage and improve flood resilience by incorporating NBS into infrastructure and urban development projects

ABSTRACT. When the population of a growth region increases the areas need to expand to provide for the rapidly growing population and this results in the availability of a variety of residential buildings. As such, new accommodations are packed in a way that entails enhanced embodied carbon, water, and energy. This research employs satellite and optical imagery to examine construction disparities in the Melbourne study area and a Victorian characterization growth area to ascertain embodied carbon, water, and energy. These are the areas of concern, and they are measuring 5 km2 with the differences in construction types. The growth region is predominantly represented by second-generation low-rise residential buildings: 80% of the built area – while Melbourne trend is mixed-purpose industrial construction: 30% of the built area is built in this type. The approach applied in this work defines the utility of satellite imagery data gathered from the open source and converted to the spatial database in QGIS. Applying this approach allowed us to fulfill visual investigation of constructions in the regions which were interesting to make conclusions about a generational identity of objects and materials, used for constructions. To enrich the readers, I will give additional descriptions on the remote sensing products available, and the data used in the manuscript. The embodied carbon in Melbourne stands at 32,895 tonnes while the water and energy amounts to 4,192 ML and 3,694,412 GJ, respectively. Whereas the totals for the growth region is the emission of 179376 tonnes of carbon, consumption of 2533 ML of water, and energy utilization of 2243567 GJ. Even though the overall footprint of Melbourne is much higher than the growing region whose population is equivalent to that of Melbourne, it is perceived that the growing region with the current construction type has a higher embodied carbon, water and energy per capita. Melbourne’s averages are equal to 226 TL per capita. 7 Giga joules of energy, 257 thousand litres of water, and 20 tonnes of carbon. In the growth region though the embodied energy, water and carbon per capita has been determined to be 287. 4 GJ, 324. 6 kL for the first one and twenty-two tones for the second one. The current performance per capita for the growth region, however, is considerably less efficient than Melbourne. One way to change this may be using different types of residential construction which might enable reducing the impact of density of the population to the density of the materials used.

ABSTRACT. Water and other natural resources within the Cagayan River Basin (CRB) are recognized as critical resources that must be appropriately managed and maintained to provide greater benefits to communities, farmers, and other water users in the region and surrounding areas. In order to achieve water security in the region, water assessment and resource planning need to be strengthened and integrated. Through WEAP as an analytical modeling tool, the research aimed to generate information on water resources, its present use, and future demands. It also aimed to establish indicators of the water security index under present and future climate scenarios.

The WEAP model was customized to evaluate its applicability in the basin. Data on water resources, including hydrologic variables and water demand, were gathered and used as inputs in model development. The model was calibrated and validated until it satisfactorily simulated the hydrological processes. The local WEAP model was used in the analysis of the present and future water supply and demand under various scenarios. Future scenarios include additional Local Government Units (LGUs) as water users, population growth, irrigation system improvement, climate change, and forest loss.

Simulations showed that the population growth resulted in significant increase in unmet water demand. In order to meet the demand for water supply for both agricultural and domestic surface water, the combination of high conveyance efficiency and moderate alternate wetting and drying irrigation management interventions is required. Additional scenarios like climate change and forest loss were proven to also increase the unmet water demand. Irrigation improvements such as canal concrete lining that will improve conveyance efficiency to 95 percent and irrigation system modernization that will result to 25 percent water savings.

Further in this research study, a water security index (WSI) was developed using water security dimensions which define the overall WSI of the basin. It considers a variety of driving forces that impact water security, including domestic water, economic water, environmental water, water-related disaster, and water governance. The study revealed that water-related disaster, as a key dimension, is a serious concern in the basin. Recommendations to achieve resilience in water-related disasters are to prioritize investments in disaster risk reduction infrastructure, increase in public finance for water-related disasters infrastructure, promotion of integrated flood risk mitigation including nature-based solutions, and improvement of data collection and associated systems for proactive disaster risk management.

On policy-making and implementation of interventions, the River Basin Control Office, in close collaboration with the Cagayan River Basin Management Council, will facilitate and spearhead the crafting of policies, programs, projects and activities along water resource management and oversee its implementation.

ABSTRACT. Over the past two decades, floods have become one of the most frequent, severe and devastating major natural disasters caused by climate change. These phenomena are responsible for thousands of deaths every year, as well as economic losses estimated at billions of dollars. The aim of this work is to present innovative technologies developed by DryUp Swiss Technology to deal with natural disasters such as flooding. Developed through years of research and expertise, DryUp Swiss technologies are designed to offer rapid and effective responses to mitigate the effects of flooding. Among these technologies stands out the FloodBlock system, which represents a significant advance in flood protection. The FloodBlock system was validated through a multidisciplinary methodology that involved advanced 3D numerical modeling and prototype testing with 50 tons of water in various flood scenarios, including both flash floods and gradual rising waters, at the Artelia laboratory in Grenoble, France. The results demonstrate that this system is designed to be deployed rapidly, offering a fast and effective flood barrier that significantly reduces response time to flood hazards. Furthermore, the barrier has proved a superior resistance to water pressure tests. Its flexibility and robustness make it an effective solution for protecting vulnerable areas, reducing the disastrous impact of flooding without sacrificing respect for the environment. The use of innovative DryUp technology and its installation in areas at risk of flooding offers valuable assistance to decision-makers and disaster risk managers in fostering a resilient future for coastal communities.

ABSTRACT. One major cause of flash floods downstream of water-retaining structures is their failure, often due to internal erosion. A primary issue with earthen embankment dams is that internal erosion mechanisms such as concentrated leak, backward erosion, or suffusion can lead to soil seepage within embankments, dykes, and levees. This seepage can form a pipe that expands under a high hydraulic charge. Piping, in particular, is a rapid and dangerous form of internal erosion that can swiftly lead to the failure of earthen structures. At the laboratory scale, researchers use various piping erosion testing apparatuses to investigate and simulate the piping mechanism. These allow for the quantification of key erodibility properties (critical erosive shear stress and soil erosion coefficient), which aid in assessing piping erosion behavior and developing effective erosion control measures. The present study aims to investigate the possibility of induced piping erosion in controlled laboratory settings using soil compaction permeameter apparatus for different soil textures and ultimately assess their erodibility. The experiments were conducted using a permeameter compaction apparatus. A central hole (pipe) was drilled axially through the compacted soil sample, which was then subjected to a controlled water flow system. This setup included an upstream pressure sensor and a downstream collection jar. Five fine-grained soil types were tested under varying hydraulic heads. During the tests, hydraulic pressure and flow rates were continuously recorded, and the final diameter of the hole (pipe) was measured. The preliminary results from the conducted piping erosion tests allowed for a qualitative classification of each soil sample as either resistant or susceptible to erosion under the testing conditions. Here, the assessment is done in terms final diameter of the hole (the pipe)measured at the end of the each test and the test’s duration (where under a given hydraulic headand for the same duration,the larger the final hole diameter is, the more susceptible to erosion the soil is, and vice versa). Furthermore, technical limitations of the by default equipment settings were identified (e.g., water inflow and outflow limitations, outlet clogging), and a need for modified testing configuration has been judged necessary.

ABSTRACT. Water scarcity and access to safe drinking water are pressing global challenges aggravated by climate change. This article addressed the specific case of Nyala City in West Sudan, where water shortages occur during the summer due to a decline in groundwater levels. This study aims to develop a comprehensive rainwater harvesting system as a potential solution to alleviate water scarcity in the city. Also, the goals involved in investigating the social and cultural factors that influence the acceptance and adoption of rainwater collection, to ensure water quality it is necessary to design an efficient sand filtration system, an evaluation of the effectiveness of the filtration system and the economic viability were conducted. The Data were collected through structured questionnaires and interviews conducted with the local community experiencing water scarcity and the relevant staff of the Water Corporation responsible for water supply, with staff from the Water Corporation revealed a moderate level of familiarity with rainwater filtration systems, with 70% of participants, comprising 60% male and 40% female and similarly, the questionnaire results for citizens indicated mixed perceptions of the respondents, 55% were male and 45% were female. Additionally, rainfall data spanning the period of 2011-2020 were obtained from Climatic Research Units (CRU data) and Nyala Shapefile from Data-Interpolating Variational Analysis Geographic Information System (DIVA-GIS). The findings indicate significant variations in rainfall patterns across different city regions. These findings were instrumental in informing the design of the rainwater harvesting system, which incorporates five filtration layers, including coarse sand, charcoal, and gravel, to effectively filter and enhance the quality of collected rainwater. Two case studies were presented to provide practical insights: Elshahid Hamza Basic School and a residential house. These case studies illustrated the layout and components of the rainwater harvesting system, emphasizing the integration of plastic tanks and PVC pipes for efficient collection, storage, and distribution of rainwater. Furthermore, calculations were performed to estimate the annual water harvesting potential for the case study buildings based on the roof area and average rainfall.

ABSTRACT. Dams’ construction is an ancient engineering practiced by humans for thousands of years. Historically, dams show great innovations in the field of water resources management that includes water supply for cities and settlements, supply of irrigation water, protection of cities and villages from flashflood hazards, flow regulations, storage of water, and even hydropower generation. Worldwide, the register of the International Commission on Large Dams (ICOLD) indicates a dramatic increase of the number of such dams within the past thirteen years to about 60,000 large dams (between 1950 – 2020). In the GCC countries and the MENA region, for many decades, dams have played an essential role in augmentation of groundwater resources (by harvest of flashflood water), protection or urbanized and rural areas, storage of surface water, and in some areas for hydropower generation. In the Sultanate of Oman, 187 dams have been constructed as of 2023 for different purposes: flood protection (5 dams), recharge (67) and surface storage (115). Similarly, there are more than 600 dams in Saudia Arabia (with a strategic plan to reach 1000 dams by 2030), and about 140 dams in United Arab Emirates. The total storage capacity of such dams in the GCC countries reached more than 2.8 billion m3. Considering the MENA region, the total number of dams constructed before 2013 reaches more than 2000 dams (Alataway et al., 2019, Fouli et al., 2015, Al-Ismaily et al, 2013, Bouwer et al, 2001, Casas et al., 2021, Djuma et al., 2017, Luo et al., 2020, Missimer et al., 2015, Sumi et al., 2022, Mohammadzadeh-Habili and Khalili, 2020, Prathapar and Bawain, 2014, Zaidi et al., 2020, Ajjur and Baalousha, 2021). In recent years, the so-called “community dams” have become very popular in Oman. Community dams are an initiative by society to build dams (mainly small to mid-size –storage capacity in the range of 0.01 – 0.35 Mm3). Since 2017, the total number of community dams reached 38, with a total storage capacity of 4.3 Mm3 with a total cost of 6,734,692 US$(Al-Hamdi, 2024). Available statistics show a boom in building dams during the past century and indicate at the same time many failures that resulted in losses in human lives, damage of infrastructure, and devastating impact to the economy and the society. This indicates gaps in human knowledge (or consideration) of safety measures that could have stopped such failures. For example, the failure of Wadi Ahen dam in Oman, collapse of Wadi Derna dam in Libya, and very recent catastrophes in Sept 2024, of several dams in Poland due to torrential rains in Eastern Europe. The danger of potential dam’s failure becomes even more possible with the impact of climate change (extreme rainfall events become more frequent inducing more associated hazards). Moreover, absence of regular and rigorous inspection procedures and protocols for dam’s safety. Thus, dam safety emerged as an issue that must be considered at all levels, especially when dam’s failure/collapse cases start to be of concern or subject of litigations in forensic geotechnical investigations.