China
Africa
Explore chapters and articles related to this topic
Evolution of Long Short-Term Memory (LSTM) in Air Pollution Forecasting
Published in Monika Mangla, Subhash K. Shinde, Vaishali Mehta, Nonita Sharma, Sachi Nandan Mohanty, Handbook of Research on Machine Learning, 2022
Satheesh Abimannan, Deepak Kochhar, Yue-Shan Chang, K. Thirunavukkarasu
Floods cause a significant loss of life and property and can cause huge social and economic losses. It strongly affects the locations around the river downstream. Flood mitigation and accurate forecasting become the need of the hour. The traditional methods involve simulation of the hydrodynamic process of the water’s flow through mathematical models. These models fail to gain decent accuracy due to data limitation and longer computational times. Daily discharge and rainfall measures can be used to perform a time series analysis for flood forecasting using LSTMs. The LSTM data-driven model can very efficiently learn linear and non-linear patterns of the data and thus offer a promising alternative to the existing mathematical models for the hydrological forecast of streamflow. In [14] Xuan-Hien Le et al. used LSTM to forecast one-day, two-day, and three-day flood flow at the Hoa Binh Station on the Da River, Vietnam. The LSTM model was not dependent on data such as land-use and topography for rainfall-runoff simulation which strengthens its ubiquitous applications. The LSTM model only used data collected at the target station and the upstream meteorological and hydrological stations to forecast the flovvrate at the target station. The LSTM model efficiently learned long-term relationships between the sequential series of data and strongly demonstrated reliable outputs and results.
Summary
Published in Jiri Marsalek, Blanca Jiménez-Cisneros, Mohammad Karamouz, Per-Arne Malmquist, Joel Goldenfum, Bernard Chocat, Urban Water Cycle Processes and Interactions, 2014
Jiri Marsalek, Blanca Jiménez-Cisneros, Mohammad Karamouz, Per-Arne Malmquist, Joel Goldenfum, Bernard Chocat
Urban drainage and flood protection practice has evolved dramatically during the last thirty years. With respect to drainage, the old policy of fast removal of surface runoff from urban areas has been largely abandoned and replaced by such concepts as low-impact development, sustainable urban drainage systems, or water sensitive urban design. All these concepts emphasize the need for distributed drainage systems maintaining on-site water balance and thereby reducing generation of runoff and its conveyance to downstream areas. This is achieved by minimizing imperviousness and employing green roofs, grassed drainage swales, infiltration facilities of various designs (wells, trenches and basins), storage and treatment in stormwater management ponds and constructed wetlands, and similar measures. For all these systems, design guidance and verification of performance is available. Essential issues include the overall system design, safe operation and proactive maintenance. In flood protection, the emphasis is shifting from the traditional structural or �do nothing� alternatives to more sustainable approaches combining a full range of such measures as keeping floodplains for flood conveyance, structural measures (dams and dykes), and non-structural flood management measures, including flood mapping, zoning, insurance, and real-time flood forecasting and warning. At the same time it is recognized that in the developing world, applications of many of these measures may be limited by political and economic conditions, and institutional arrangements.
Integration of Flood Losses in Risk Analysis
Published in Saeid Eslamian, Faezeh Eslamian, Flood Handbook, 2022
Dilani R. Dassanayake, Andreas Burzel, Hocine Oumeraci
First, the problem should be identified and clearly defined as in every other decision-making process. Floods may cause significant damage, including not only tangible (economic) damages, but also intangible (social and environmental) losses. In flood risk analysis, it is, however, not common to consider the intangible losses, mainly due to the difficulties in combining tangible and intangible losses within one frame. This study focuses on the integration of both tangible and intangible losses in flood risk analysis. Therefore, the goal of the MCA in this study is to determine the spatial distribution of integrated flood losses in order to estimate the future flood risk.
Quantitative study of water impact on land value in Jakarta
Published in Urban Water Journal, 2023
Ahmad Gamal, Lailatul Rohmah, Cynthia Adelina Perangin Angin, Widya Laksmi Larasati, Ahmad Aki Muhaimin, Risty Khoirunisa, Dwinanti Rika Marthanty
Flood, which is mainly caused by rain runoff, is a common disaster in Jakarta thus require a control system. A flood disaster system covers structural and non-structural methods (Kodatie, 2013 in Boatwright et al. 2014; Sörensen and Emilsson 2019; Wicaksono and Herdiansyah 2019). The structural methods include (a) flood control buildings such as dam, retention pool, check dam, groundsill, drop structure, retarding basin, and polder; (b) river improvement and regulation systems such as embankment, river improvement, by pass/short cut, floodway, and special drainage system. The non-structural methods of flood control system include watershed management, land use arrangements, erosion control, development and regulation of flood areas, emergency mitigation system, flood warning system, law enforcement, assurance, and information distribution. Additionally, more nature-oriented approaches such as vegetative conveyance systems and infiltration areas for rain runoff are included in the non-structural methods (Boatwright et al. 2014; Sörensen and Emilsson 2019).
Intricate flood flow advancement modelling in the krishna river sub basin, India
Published in ISH Journal of Hydraulic Engineering, 2023
Rangineni Pallavi, K. Rekha Rani, Kulkarni Shashikanth, P. Rajasekhar, Hiteshri Shashtri
Flood forecasting models essentially plan to estimate the water level and discharge of approaching flood. These forecasting models use data of flood flows and stages at strategic points in river basin to predict floods. The conventional method employs flood routing approaches such as river (channel) routing and reservoir routing. These routing techniques mainly aim to predict flood hydrographs at various sections. Furthermore, flood peak attenuation and duration of high water levels form essentially an efficient flood forecasting approach (Chow 1964). In river, during floods, the flow is non-uniform and unsteady. The hydraulic characteristics vary from stage to stage and also from channel to channel inclusive of later flows. The hydraulic routing method essentially involves the solution of Saint-Venant equation, whereas the hydrologic routing is traditionally based on the physical properties of the channel by applying methods such as Muskingum and Muskingum-Cunge (Chow 1964, Refsgaard 1997). However, with the advent of numerical methods and in combination with high-performance computing facilities, the flood modeling capabilities have increased and reliability of estimates has also increased in recent times (Timbadiya et al. 2014).
Big Data Driven Map Reduce Framework for Automated Flood Disaster Detection Based on Heuristic-Based Ensemble Learning
Published in Cybernetics and Systems, 2022
Abdallah Saleh Ali Shatat, Md. Mobin Akhtar, Abu Sarwar Zamani, Sara Dilshad, Faizan Samdani
Flood disaster is the most common natural disaster in the global region. Flood disasters can be caused by heavy rainfall, the severity of winds over water, high tides in water, etc. It creates a huge impact on human lives, the environment, and economic losses. Therefore, flood disaster detection is widely used to mitigate the damage and impacts of floods. Flood mitigation entails managing and controlling the movement of flood water. In past studies, the existing algorithm could not access a large dataset and required more computational cost. Since floods occur by a landslide, some traditional methods are futile for the early detection of floods. Because of these constraints, recent techniques can forecast flood damage. To enhance the prediction accuracy, a newly developed algorithm is proposed to detect flood disaster and attains high prediction accuracy, as shown in Figure 1.