Calculation and visualization of flood inundation based on a topographic triangle network
Introduction
Floods are among the most devastating natural disasters in the world, each year claiming more than 20,000 lives and adversely affecting about 75 million people world-wide, mostly through homelessness (Smith, 2001). Consequently flood protection systems are necessary for most countries, with the rapid development of computer technology, non-engineering measures are becoming more important in flood management and decision-making (Ding et al., 2004, Gouldby et al., 2010, Liao et al., 2011, Qi and Altinakar, 2011). Establishing flood and inundation simulation systems could extend information that would be helpful in dealing with contingencies and emergencies, and thereby help to alleviate risks and losses of life and property (Chang et al., 2010). Moreover, characterizing floodplains and other areas with a high potential for flash floods can help authorities produce management strategies for flood mitigation, such as designing water control structures (reservoir levee projects), decision making for flood insurance and facilitating emergency preparedness to cope with flooding (Sarhadi et al., 2012, Sanders, 2007, Srinivas et al., 2008). Among these systems, flood inundation models play a central role in the evaluation, selection and in some cases the implementation of flood protection measures (Ernst et al., 2010, Merz et al., 2010). During past decades, predictions of flood inundation extent have been made possible by advances in numerical modeling techniques and the synergistic use of radar and optical remote sensing in conjunction with GIS modeling (Chang et al., 2010, Hsu et al., 2002).
Accuracy and efficiency are two major indicators used in evaluating the performance of flood inundation models (Chen et al., 2012). These two aims (accuracy and efficiency) are often in conflict with each other. On one hand, recent developments in remote sensing technology have increased the availability of digital terrain models (DTMs) and digital elevation models (DEMs). Data availability no longer represents a significant limitation to improvements in the accuracy of numerical modeling (Chen et al., 2012). At the same time, many physically-based inundation models have used these data to develop inundation simulations with improved accuracy (Bates et al., 2010, Jiang et al., 2011). On the other hand, inundation models based on hydrodynamics need long computational times for iterative solutions to carry out high-resolution spatial discretization (Imrie et al., 2000, Yu, 2010). The increase of model resolution results in the growth of grid cell numbers, and as a consequence, the number of steps required for temporal progress increases because the time computation step decreases with grid size to ensure model stability. To deal with the efficiency problem, many parallel computing techniques, including the Message Passing Interface (MPI), Open Multi Processing (Open-MP) and the Graphical Processing Unit (GPU), have been applied to speed up 2D inundation modeling (Hankin et al., 2008, Neal et al., 2010, Yu, 2010). Nevertheless, the growth in high-resolution terrain data has been much quicker than the increase of computing capacity for real-time or large-scale modeling. The limitations of available computing resources, therefore, still restrict the applications when very detailed information or risk-based analysis is required over large areas. Consequently, it is very difficult to reach on-line simulation of the extent of inundation using a complex inundation models based on hydrodynamics under current conditions (Chang et al., 2001, Chang et al., 2010, Rajurkar et al., 2004).
Besides parallel computing, another way to solve this problem is to simplify the flood inundation model. There are many rapid flood inundation models available in recent years, for example, the rapid flood inundation models RFIM (Krupka, 2009), an urban flood inundation model called GUFIM (Chen et al., 2009). Another good example of rapid flood inundation model is ISIS FAST, developed by Halcrow Group Ltd., which allows quick assessment of flood spreading using simplified algorithm. These tools work by pre-processing DEM data to identify depressional subcatchments in the ground surface. Hydrological input data, such as rainfall, inflows, tidal/water level boundaries are then routed through the models to generate flood extent and depth results (Paul et al., 2012).
The goal of this study is to find a method that can reach rapid simulation of flood inundation extent with high-resolution data. To improve flood model accuracy, high-quality topographic data and triangle network have been used. Also, to improve efficiency, a connected domains searching algorithm has been used instead of a hydraulic model to simulate flood inundation. Two cases have been used to expand on the usage and effect of the method.
Section snippets
Study area
Harbin, an important industrial base and the biggest transportation junction in northeastern China, is located in the southeastern part of Songnen plain (Fig. 1). The city covers an area of 53,000 km2, including a central urban area of 1637 km2, and has a population of 9.68 million. The rivers of the region all belong to the Songhua River system. The main path of the Songhua River originates in the Tianchi in Changbai Mountain, in Jilin province, and crosses the central part of Harbin from west to
River flood inundation method
In flood inundation simulations, terrain elevation and flood level are the two basic elements. Calculation accuracy and efficiency of flood inundation simulations are directly affected by the resolution and storage format of the terrain model. To acquire a more accurate simulation result, high-quality terrain data should be used, and as a consequence, can be divided into regular (e.g., GRID) and irregular (e.g., TIN) nets. Regular net data storage structure is relatively simple, but it has
River flood inundation
The Songhua River is used as an example to illustrate the use of the FCDC flood inundation method. In the Songhua River, dikes provide the main flood control. Because the river is wide and shallow and has a large shoal area, dikes with many different standards of construction have been used from place to place on the shoals to address flood control in the region. One of the challenges of a river flood inundation analysis in such a setting is to calculate the range of possible inundation areas
Discussion and conclusions
In this study, the FCDC method was presented to simulate source flooding such as river or dike breach flood inundation. Compared with a 2D hydrodynamic flood model, the FCDC method is relatively simple. As long as a flood flows into an area and flood levels are higher than the terrain elevation, the area will be inundated. Therefore, we simply need to provide the inundation water level and terrain elevation data to quickly determine the inundation area and depth distribution. The following
Acknowledgements
This study was supported by the National Natural Science Foundation of China (51009064, 51379076) and the 12th Five-Year National Key Technology R&D Program (2012BAB05B05).
References (32)
- et al.
A simple inertial formulation of the shallow water equations for efficient two-dimensional flood inundation modelling
J. Hydrol.
(2010) - et al.
Clustering-based hybrid inundation model for forecasting flood inundation depths
J. Hydrol.
(2010) - et al.
A coarse-grid approach to representing building blockage effects in 2D urban flood modelling
J. Hydrol.
(2012) - et al.
A GIS-based model for urban flood inundation
J. Hydrol.
(2009) - et al.
Two-dimensional, high-resolution modeling of urban dam-break flooding: a case study of Baldwin Hills, California
Adv. Water Resour.
(2009) - et al.
Multiscale modelling in real-time flood forecasting systems: from sand grain to dike failure and inundation
Proc. Comput. Sci.
(2010) - et al.
River flow prediction using artificial neural networks: generalisation beyond the calibration range
J. Hydrol.
(2000) - et al.
Numerical modeling of dam break-induced flood and inundation using smoothed particle hydrodynamics
J. Hydrol.
(2012) - et al.
River environmental decision support system development for Suzhou Creek in Shanghai
J. Environ. Manage.
(2011) - et al.
A comparison of three parallelisation methods for 2D flood inundation models
Environ. Modell. Softw.
(2010)
A GIS-based decision support system for integrated flood management under uncertainty with two dimensional numerical simulations
Environ. Modell. Softw.
Modeling of the daily rainfall-runoff relationship with artificial neural network
J. Hydrol.
Evaluation of on-line DEMs for flood inundation modeling
Adv. Water Resour.
Probabilistic flood inundation mapping of ungauged rivers: linking GIS techniques and frequency analysis
J. Hydrol.
Regional flood frequency analysis by combining self-organizing feature map and fuzzy clustering
J. Hydrol.
Impact of the Gezhouba and Three Gorges Dams on habitat suitability of carps in the Yangtze River
J. Hydrol.
Cited by (46)
A 3D-Panoramic fusion flood enhanced visualization method for VR
2023, Environmental Modelling and SoftwarePluvial flooding: High-resolution stochastic hazard mapping in urban areas by using fast-processing DEM-based algorithms
2022, Journal of HydrologyCitation Excerpt :Consequently, RFMs are adequate for conducting probabilistic analyses. Some RFMs are the Rapid Flood Spreading Model (RFSM) (Bernini and Franchini, 2013), the flood-connected domain calculation (FCDC) method (Zhang et al., 2014), the rapid urban flood inundation and damage assessment model (RUFIDAM) model (Jamali et al., 2018), and the cellular automata fast flood evaluation (CA-ffé) model (Jamali et al., 2019). Hierarchical filling-and-spilling algorithms (HFSAs) or puddle-to-puddle dynamic filling-and-spilling approaches (P2Ps) have been recognised as promising techniques for characterising pluvial floods (Chu et al., 2013; Zhang and Pan, 2014).
Modeling spatial distribution of flow depth in fluvial systems using a hybrid two-dimensional hydraulic-multigene genetic programming approach
2021, Journal of HydrologyCitation Excerpt :Hydrological models are currently the most popular choice for inundation forecasting, especially when a timely prediction is required and the study area is a large-scale catchment (Bermúdez et al., 2018). Traditional hydrological models mainly calculate the time series of discharges in fluvial systems, but recent distributed hydrological models based on digital elevation models (DEM) can also delineate the inundation extends and depths (Zhang et al., 2014; Li et al., 2020; Ming et al., 2020), making them widely used nowadays for flood forecasting. However, hydrological models typically ignore the strict conservation of momentum, and can only provide a simplified representation of the hydrological responses.
Approximate calculation of flash flood maximum inundation extent in small catchment with large elevation difference
2020, Journal of HydrologyCitation Excerpt :The damages caused by flash floods mainly include strong scouring and erosion and widespread inundation. It is the basis of flash flood disaster control planning (FFDCP) to calculate and analyze the inundation region (Chang et al., 2010; Zhang et al., 2014; Burn and Whitfield, 2015). Many efforts have been made to study flood inundation, which can be categorized into three types from the perspective of research methods: remote sensing extraction, hydrodynamic simulation, and inundation calculation based on Digital Elevation Model (DEM).
Urban flood risk assessment and analysis with a 3D visualization method coupling the PP-PSO algorithm and building data
2020, Journal of Environmental Management