Sediment yield estimation along with identification of soil erosion mechanisms is essential for developing sophisticated management approaches, assessing, and balancing different management scenarios and prioritizing better soil and water conservation planning and management. At the watershed scale, land management practices are commonly utilized to minimize sediment loads. The goal of this research was to estimate sediment yield and prioritize the spatial dispersion of sediment-producing hotspot areas in the Nashe catchment using the Soil and Water Assessment Tool (SWAT). Moreover, to reduce catchment sediment output, this study also aims to assess the effectiveness of certain management practices. For calibration and validation of the model, monthly stream flow and sediment data were utilized. The model performance indicators show good agreement between measured and simulated stream flow and sediment yields. The study examined four best management practice (BMP) scenarios for the catchment’s designated sub-watersheds: S0 (baseline scenario), S1 (filter strip), S2 (stone/soil bunds), S3 (contouring), and S4 (terracing). According to the SWAT model result, the watershed’s mean yearly sediment output was 25.96 t/ha. yr. under baseline circumstances. The model also revealed areas producing the maximum sediment quantities indicating the model’s effectiveness for implementing and evaluating the sensitivity of sediment yield to various management strategies. At the watershed scale, treating the watershed with various management scenarios S1, S2, S3, and S4 decreased average annual sediment yield by 34.88%, 57.98%, 39.55%, and 54.77%, respectively. The implementations of the soil/stone bund and terracing scenarios resulted in the maximum sediment yield reduction. The findings of this study will help policymakers to make better and well-informed decisions regarding suitable land use activities and best management strategies.

A combination of a deterministic approach and fragility analysis is applied to assess tsunami damage caused to buildings. The area selected to validate the model is Imwon Port in Korea. The deterministic approach includes numerical modeling of tsunami propagation in the East Sea following an earthquake on the western coast of Japan. The model is based on the linear shallow-water equations (LSWE) augmented with Boussinesq approximation to account for dispersion effects in wave propagation, and coastal wave run-up is modeled by non-linear shallow-water equations (NLSWE). The output from the deterministic model comprises inundation depth. The numerical output is used to perform fragility analysis for buildings vulnerable to flooding by tsunamis in the port area. Recently developed fragility curves—based on the ordinal regression method—are used for damage probability estimates. The extent of structural damage in the areas under a tsunami hazard is identified by the numerical modeling of tsunami features. Our numerical model offers high bathymetric resolution, which enables us to capture flow features at the individual structure level and results in improved estimation of damage probability. This approach can serve as a measure of assessing structure vulnerability for areas with little or no records of tsunami damage and provide planners with a better understanding of structure behavior when a tsunami strikes.

Scouring around bridge piers is a highly nonlinear process making its prediction by deterministic and stochastic models challenging. This study explores the application of inferential models for predictions of bed elevations around bridge piers. The objective is to get a generalized machine learning model with an interpretable structure. The historical data comprise a detailed record of streamflow and bed elevations that were captured by sensors installed at the 5th Street Bridge piers over Ocmulgee River at Macon, GA. We investigate the accuracy and efficiency of various tree-based machine learning algorithms, including a single tree as well as homogeneous ensemble models for simultaneous predictions of bed elevation at multiple sensors installed at piers. The ensemble models were based on bagging and boosting techniques. Special attention is given to balancing between overfitting and underfitting without compromise on the model's robustness. Observation of the performance metrics showed that tree-based models have excellent predictive capacity. It was observed that boosting models, including a gradient based regression model, and adaptive boosting outperformed the bagging model. Among all the models investigated in this study, the adaptive boosting method was observed to be most generalizable. The performance of developed models shows the potential of tree-based ensemble models in providing rapid and robust predictions for complex nonlinear fluid flows.

The Fukushima nuclear disaster highlights the importance of accurate and fast predictions of tsunami hazard to critical coastal infrastructure to devise mitigation strategies in both long-term and real-time events. Recently, deep learning models allowed us to make accurate and rapid forecasts on high dimensional, non-linear, and non-stationary time series data such as that associated with tsunami waveforms. Thus, this study uses a one-dimensional convolutional neural network (CNN) model to predict waveforms at cooling water intakes of nuclear power plant at Uljin in South Korea. The site is particularly vulnerable to tsunamis originating from the west coast of Japan. Data for the CNN model are generated by numerical simulation of 1107 cases of tsunami propagation initiating from fault locations. The time series data for waveforms were predicted at 13 virtual gauges located in the nearshore region of the study area, 10 of which were classified as observation points and 3 gauges situated at the cooling water intakes were categorized as target locations. The performance assessment of the model's forecasts showed excellent results with rapid predictions. The study highlights two main points: (i) deep learning models can be based on sparse waveform in situ data (such as that recorded by deep-ocean assessment and reporting of tsunamis or any locally operating monitoring stations for ocean waves) or numerically simulated data at only a few points along the dominant wave propagation direction, and (ii) deep learning models are fully capable of accurate and fast predictions of complex geo-hazards that prompt rapid emergency response to coordinate mitigation efforts.

Due to overexploitation and lower rainfall rates, it is essential to study the detailed water balance of the Keenjhar lake by considering the climate change impacts and higher water demands linked with the population growth. A hydrological model of Keenjhar Lake is developed based on a system dynamic approach using STELLA (Structural Thinking and Experiential Learning Laboratory with Animation). The model (STELLA) developed in the current research study comprises the following three sub-systems: population, water supply, and water demand. The hydrological and climate data for the period of seventeen years (2000–2016) is used in the current study. The monthly water budget of the Keenjhar Lake is determined by inflow components such as rainfall and the Kalri-Baghar Feeder (K.B.F) (upper) and outflow components such as evaporation, the K.B. Feeder (lower), and the Keenjhar-Gujju (K.G) canal from the lake. The water balance results revealed that the contribution of direct rainfall and the annual inflow components to the lake are 22.03% and 77.91%, respectively. Whereas the evaporation, outflow to K.B.F lower and water abstraction to the K.G. Canal constituted about 5.78%, 92.55%, and 1.57% of the total annual outflow from the lake, respectively. Moreover, the annual inflow components of the water budget of the lake showed a declining trend while the outflow components (water abstraction) intimated an increasing trend. The study results also acknowledged that the demand for water can increase from 3 × 10¹⁰ ft³/yr up to 1.2 × 10¹¹ ft³/yr by the year 2050 (influence of overdrawing of water due to population growth), and water supply may decrease to 9.066 × 10¹⁰ ft³ (rainfall depletion due to climate change). A detailed water balance explains the main water loss components and will help in developing better water management practices and well-informed policy decisions.