An official website of the United States government
The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.
The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.
- Account settings
- Advanced Search
- Journal List
- Sensors (Basel)
Remote Sensing Methods for Flood Prediction: A Review
Hafiz suliman munawar.
1 School of Built Environment, University of New South Wales, Kensington, Sydney, NSW 2052, Australia; [email protected]
Ahmed W. A. Hammad
S. travis waller.
2 School of Civil and Environmental Engineering, University of New South Wales, Kensington, Sydney, NSW 2052, Australia; [email protected]
No new data were created or analysed in this study. Data sharing is not applicable to this article.
Floods are a major cause of loss of lives, destruction of infrastructure, and massive damage to a country’s economy. Floods, being natural disasters, cannot be prevented completely; therefore, precautionary measures must be taken by the government, concerned organizations such as the United Nations Office for Disaster Risk Reduction and Office for the coordination of Human Affairs, and the community to control its disastrous effects. To minimize hazards and to provide an emergency response at the time of natural calamity, various measures must be taken by the disaster management authorities before the flood incident. This involves the use of the latest cutting-edge technologies which predict the occurrence of disaster as early as possible such that proper response strategies can be adopted before the disaster. Floods are uncertain depending on several climatic and environmental factors, and therefore are difficult to predict. Hence, improvement in the adoption of the latest technology to move towards automated disaster prediction and forecasting is a must. This study reviews the adoption of remote sensing methods for predicting floods and thus focuses on the pre-disaster phase of the disaster management process for the past 20 years. A classification framework is presented which classifies the remote sensing technologies being used for flood prediction into three types, which are: multispectral, radar, and light detection and ranging (LIDAR). Further categorization is performed based on the method used for data analysis. The technologies are examined based on their relevance to flood prediction, flood risk assessment, and hazard analysis. Some gaps and limitations present in each of the reviewed technologies have been identified. A flood prediction and extent mapping model are then proposed to overcome the current gaps. The compiled results demonstrate the state of each technology’s practice and usage in flood prediction.
Recently, there has been an increase in natural as well as man-made disasters in the world. Hydrological extremities caused by human activities, increased urbanization, global warming, and weather change can be attributed to the dramatic rise in the global flood risks [ 1 ]. Among the natural disasters, flooding is the most devastating natural hazard. Floods are common in all parts of the world. However, their characteristics and intensity vary from region to region [ 2 ].
Apart from the destruction of infrastructure and agricultural lands, floods cause in-tense impacts on the people who may drown or who have severe injuries due to hypothermia, for instance [ 3 ]. Some additional fatalities may occur due to indirect effects of floods which include the destruction of the health infrastructure, the spread of infectious diseases, psychological distress, and starvation [ 4 , 5 ]. Being the most commonly occurring natural disaster, floods have caused approximately 53,000 fatalities on a global scale [ 4 ]. In Europe, the economic losses resulting from floods that occurred in the year 2005, 2007 and 2010 surpassed EUR 1 billion [ 5 ]. Floods that occurred in Europe in the year 2002 caused material damage worth EUR 20 billion [ 6 ]. The impact of floods on people and communities can be devastating. Similarly, in Australia, more than 900 fatalities occurred due to floods with an estimated cost of infrastructure damage to about AUD 5 billion [ 7 ]. Recent flooding in north-eastern China has caused catastrophic impacts in many cities within Shanxi province, including in its capital, Taiyuan, with 29 fatalities in the region. The event occurred over 14 days (1 to 14 October 2021), with the heaviest rainfall occurring over 5 days, from 2–7 October [ 8 ]. At 185 mm, the rainfall in Taiyuan over 12 h was the highest recorded rainfall in the area. This was seven times the pre-2010 overall average for the same month [ 9 ]. Around 1.76 million people were affected, and economic losses reached CNY 5.02 billion, or about USD 780 million [ 9 ].
Regions influenced by the disaster and the extent of damage can, in certain instances, go undetected due to the large geographical size of regions affected by rain and flood, which hamper immediate relief response activities [ 7 ]. Due to extent of the destruction, some regions become inaccessible, and the relief groups fail to provide their services. Currently, images of a disaster region are extracted with satellite and aerial imagery [ 10 ]. These images undergo various image processing techniques to make predictions regarding the possibility of flood occurrence in a particular region. The remote sensing technology extracts the characteristics of the flood region and gives information about the upcoming disaster and event challenges. With the help of obtained image and remote sensing data, flood risk maps can be produced. High-quality images are derived from remote sensing mechanisms such as synthetic aperture radar (SAR) technology that provides high-resolution images of the land and water reservoirs, even in bad weather conditions and low light. Using recent technologies such as artificial intelligence (AI) and image processing can assist in automatic flood risk mapping [ 10 , 11 ]. These methods can assist in the alarm systems and framework development that estimate the water level of a certain region and predict the upcoming floods. Coupling all these technologies with GPS increases the accuracy of results and provides the precise location of an upcoming disaster [ 11 , 12 ]. The use of such technologies for disaster prediction and management helps in lowering the destruction risk by issuing immediate warnings and formulating appropriate strategies for conducting emergency responses [ 11 ]. Through early disaster prediction, timely disaster response strategies can be adopted which include arranging for prompt evacuations and helping to build resilience in the long run. More awareness about flood-related disasters can be spread among the public. Communication gaps can be reduced, infrastructure failures can be avoided, and immediate contacts can be made with relief commissionaires [ 11 ]. Early planning and effective communication can help save mankind from hazardous situations. This can ultimately help in supporting economic and social development of the country [ 12 ].
Remote sensing technologies acquire data about objects and infrastructure on the surface of the Earth without being in direct contact by using various recording instruments [ 13 ]. Therefore, it is helpful in areas where no physical or close contact is possible [ 14 , 15 , 16 ]. Examples of such technologies include SAR, space-based imaging platforms, and satellites. This technique helps in faster data collection [ 17 , 18 ]. The data cannot be collected accurately with such efficiency using ground-based observations, whereas the remote-sensing technology gathers the same data by covering large spatial areas in the least time and provides a comprehensive view of the target objects [ 19 ]. It can capture images of distant objects despite bad weather conditions. The aerial photography and satellite images obtained using remote-sensing help in visualizing the topography and other terrain properties [ 17 ]. Such features help in locating natural disasters and evaluating their proportion. A wide variety of relief operations could be conducted by timely visualization of data obtained from the resources [ 20 ].
Several surveys have been conducted to review and evaluate the flood prediction systems [ 13 ], machine learning models for flood forecasting [ 13 ], disaster management using big data [ 14 ], flood mapping and damage assessment technologies [ 15 ]. A significant number of developed flood prediction systems use machine learning and image processing techniques for flood prediction. However, the use of various techniques and analyses associated with remote sensing for flood forecasting has not been assessed explicitly. Hence, a significant research gap was observed for assessing remote sensing-based technologies in the pre-disaster phase. Recent research in the field of remote sensing demonstrates its immense potential to make accurate predictions of an upcoming disaster and to play a vital role in flood risk analysis [ 18 ]. To overcome these research gaps, this paper presents a systematic literature review on the use of remote sensing techniques to manage floods in the pre-disaster phase by conducting accurate flood prediction. The classification framework presented in this paper categorizes the flood prediction technologies that use remote sensing based on data capturing and analytical approaches. The importance of each technique along with its current use, performance requirements, application and limitations have been highlighted to understand how it can be used to avoid the disastrous effects of floods in real-time by reliable flood prediction. This study also conducts a comprehensive comparison between these technologies based on various metrics [ 19 , 20 , 21 ]. The results of this study would help the concerned authorities in better understanding the use of technology for dealing with floods and choosing the most suitable model to manage floods in their target area. Precisely, this research is focused on answering the following research questions:
How can remote sensing technologies for flood detection be categorized?
How are remote sensing technologies being used for flood prediction?
What are the current gaps in remote sensing-based flood prediction technologies?
The rest of the paper is organized as follows: Section 2 describes the methods used to acquire materials to conduct this study, while Section 3 presents the results of the study by providing a comprehensive analysis of the remote sensing technique. Section 4 discusses the results of this research, identifies the research gaps, and presents a solution to overcome them. The paper closes with a conclusion in Section 5 . Table 1 shows the list of acronyms used in the article.
List of acronyms used in the article.
2. Materials and Methods
The main aim of this paper is to conduct a comprehensive systematic review of the recent remote sensing technologies for flood prediction and to identify the existing research gaps when it comes to flood prediction. For this purpose, a detailed study of the literature based on this domain was carried out to identify recent developments in the timely prediction of floods. To achieve these objectives, the foremost step was to gather a set of the most relevant, recent, and authentic research articles published in top tier journals. We divide the search process of research articles into two main phases, which are: article retrieval and screening phases. In the following sections, we discuss each of these phases in detail.
To retrieve the research articles for this study, the chosen search engines were Google Scholar, Web of Science, ERIC, IEEE Explore, Science Direct and Scopus. The next step was to formulate a set of queries to be used in each of these search engines to retrieve the articles. The major aim was to fully exhaust the search database and retrieve a maximum number of articles matching our domain of interest. We used three categories of terms representing the subdomains to extract a variety of research articles. The process of formulation of terms to be used as keywords in the search engines is demonstrated in Figure 1 . In this figure, the notation TC denotes the term category to which the set of keywords or terms belongs. The AND operation indicates that the final set of search queries used in the search engine of each website must contain keywords from all three categories. After entering the search queries, a set of articles ranked based on their relevance were retrieved. The first category of phrases was formulated to retrieve articles that proposed flood prediction models using remote sensing technologies that utilized multispectral sensors. In Figure 1 , “M” denotes the terms in this category. The phrases were formed by using keywords related to flood prediction which include “flood prediction”, “flood forecasting”, “flood risk analysis” and “flood hazard mapping” along with phrases such as “multispectral remote sensing” and “optical remote sensing”. The second category of terms was formulated to retrieve articles that proposed flood prediction methods using LIDAR remote sensing. For this purpose, we used flood prediction keywords along with the keyword “LIDAR”. In Figure 1 , “L” represents the terms in this category. The third term category was aimed towards retrieving articles that used remote sensing technologies having radar sensors which include the Synthetic Aperture Radar (SAR). The search strings used to extract such articles used keywords specific to flood prediction along with the string “radar”. In Figure 1 , “R” denotes the terms in this category. A total of 147 articles were retrieved after this search phase.
Query formulation for the retrieval of articles.
The number of articles retrieved from each category of search keywords is shown in Figure 2 . The number of articles from each term category that passed the screening phase is shown in Figure 2 . From the multispectral domain, initially, a total of 55 articles were retrieved because of the first phase. After analysing these papers based on screening criteria, 25 papers were excluded, resulting in the selection of 30 papers from this category. Similarly, 72 papers related to LIDAR were retrieved initially. This number was reduced to 20 after continuing through the second phase, as 52 articles were omitted. From the third term category, which is radar, initially, 30 papers were retrieved, out of which 20 passed the screening test. Hence, overall, 70 papers were finally collected as an output of the screening phase.
The Detailed Screening Process.
Figure 3 shows the year-wise distribution of articles retrieved from each category. This shows a significant increase in the use of multispectral remote sensing techniques for flood prediction as compared to radar or LIDAR based technologies in the past decade. Conversely, a comparatively smaller number of articles focused on the use of multispectral imaging technologies individually, for dealing with flood forecasting. An even smaller number of papers focused on radar-based remote sensing technologies. The search was extended to include reports, magazine articles, and web pages from authentic websites, thus increasing the scope and collecting a wide range of articles based on the subject matter. All the articles published before 1 January 2010, were discarded. This occurred to include the most recent technologies in the review. One exception to this rule was keeping some earlier papers that introduced basic concepts and definitions related to the technologies discussed in this study. This occurred to prevent any misrepresentation of information or modification of the key terms.
Year-wise distribution of articles from each category.
After the first phase based on article retrieval, the articles were passed through a screening phase to further narrow down the selection criteria. Four assessment criteria were defined to evaluate the articles:
- No duplicates;
- Time interval: 2010–2021;
- Document type: research article, abstract, book chapter;
- English language only
Thus, by filtering the articles based on these metrics, we were able to extract the most recent, applicable, and unique research articles written in the English language. From the 147 articles retrieved in the first phase, 70 articles passed all the four selection criteria. Hence, this review is based on these screened articles.
RQ-1. How can remote sensing technologies be categorized?
As a result of the article retrieval and screening process, a total of 70 recent research articles were selected to be reviewed in this study. Remote sensing technologies can be categorized into two types which are active and passive. The active remote sensors use their light source to gain data from the target located on the surface of the Earth, while the passive remote sensing methods rely on natural light or the sun to gain this data. The technologies such as radar and LIDAR use their light sources; hence, they are categorized under the active class. The remote sensing methods that use satellites to capture imagery of the target on Earth can be classified as passive methods because the satellites rely on sunlight to capture these images. The satellite sensors may be able to capture multispectral or hyperspectral images. Hence, two further types of passive remote sensing can be defined, which are multispectral and hyperspectral. Figure 4 shows the classification framework of remote sensing technologies. Most remote sensing technologies belonged to the multispectral category and hence rely on satellites. Flood prediction models belonging to the multispectral, LIDAR and radar-based remote sensing domains are analysed in detail in the subsequent sections along with their advantages and potential drawbacks.
Classification of remote sensing methods.
RQ-2. How are remote sensing technologies being used for flood prediction?
Multispectral remote sensing stores the emitted or reflected energy from objects present on the surface of the Earth through sensors that can recognize specific spectral bands [ 22 ]. The spectral bands form a thin portion of the electromagnetic spectrum, specified by the lowest and highest wavelength that is recognizable by the sensor. As a result, one raster image is saved for each of the spectral bands [ 22 , 23 , 24 , 25 ]. Examples of current satellites making use of such sensors include Sentinel-2, Landsat 7, Landsat 8, and MODIS. In this section, a review of these technologies for flood prediction is presented. Wieland and Martinis presented a framework to perform flood prediction on multispectral data gained from Landsat TM and Sentinel-2 images [ 26 , 27 , 28 ]. A convolutional neural network (CNN) was trained using these data to perform segmentation to determine water extent levels. Biases that may occur during downstream analysis are overcome by especially handling noise data such as clouds, shades, and frost. It outperforms the Random Forest classifier and a Normalized Water Index (NDWI) threshold function. Massari et al. [ 29 ] retrieved readings of soil moisture using the Advanced Scatterometer (ASCAT) to develop a rain-fall-runoff model that forecasts floods. The direct association between the satellite, soil moisture and rainfall are utilized in the model to make decisions regarding the future occurrence of floods. The study took place in the Mediterranean Sea, where readings from ten catchment locations in the ocean were recorded. These observations were acquired using the ASCAT satellite. These data are given as input to a rainfall-runoff calculation method called MISDc to obtain rainfall estimates. The rainfall data were used to predict the high-water flows in the Mediterranean Sea. Shahabi et al. [ 30 ] identified flood-prone areas using multispectral data acquired from the Sentinel-1 satellite of the Haraz watershed located in Iran. A machine learning-based ensemble method was used to perform flood susceptibility mapping. This model was composed of a combination of K-Nearest Neighbour (KNN), bagging, and a cubic classifier. Ten conditioning factors were gathered to train the model. Validation of the model showed that this ensemble method performs well and outperforms many other ensembles. The bagging approach significantly improved the accuracy of the KNN-cube ensemble for flood management and mapping problems. Noymanee [ 31 ] experimented with linear regression, ANN, boosted decision trees, Bayesian linear model, and decision forest to forecast floods in the district of Pattani in Thailand [ 31 ]. Bayesian Linear model demonstrated the best performance among all selected models, and was, therefore recommended for flood detection. A mathematical model was designed to model the upper and lower portions of the river stream. For example, to model the upstream part of a river, the following formula was used:
In the above equation, M represents the machine learning operation, W is the water level, the symbol * denotes predicted value, TX 347 and TX 33 are labels of the stations which are assigned to various river portions and R represent the rain value [ 31 ]. The flood mapping results in an input multispectral aerial image. The system classifies the flooded (Red) and non-flooded (Blue) regions and highlights them using different colours in the output such that the rescue workers can easily distinguish between them. Zhang et al. compared the flood prediction results for both spatial and temporal resolutions of existing sensors [ 32 ]. Landsat and MODIS images were collected for real-time prediction of floods. The models achieved a high level of accuracy which proved that for Landsat images both spatial and temporal models generate similar results for real-time prediction of floods. Cenci et al. evaluated the ability of Sentinel-1 to acquire soil moisture data for flood forecasting [ 33 ]. The soil moisture readings recorded by the satellite were used in a hydrological model called “Continuum” to predict flash floods. The study area was the Mediterranean Sea. The hydrological assimilation of different GEO SAR-like soil moisture products was evaluated using the SAR images. The results showed the effectiveness of Sentinel-1 derived soil moisture data to improve the flood predictions, especially for heavy flows. The Sentinel-1 data need the application of proper pre-processing methods before assimilating the data. Another finding was that apart from the need for high spatial resolution of the satellite, the temporal resolution of the satellite also plays an important role in the acquisition of correct data for the hydrological model. Ogilvie et al. [ 34 ] combined flood events data and satellite imagery to build a numerical model that monitors the water level in reservoirs. For this purpose, the rainfall run-off model and water level models were built for seven reservoirs. The data were collected between 1999 and 2014. An Ensemble Kalman Filter was applied to reduce the rainfall run-off errors and classification outliers. This method was able to reduce the root mean squared error by 54% when compared with flood forecast results provided by the previous hydrological model. Optical imagery was used for the measurement of water levels which helps in defining the scope of a flooded area [ 35 ]. The water level of a wide area can be measured in consecutive events. Analysing the change in water level can help in the easy prediction of flood events. This technology also takes the data of absolute water elevations. The data help to develop protocols for flood management and gives immense information for environmental science research. Remote sensing measures the accuracy of water up to decimeter level and shows real-time transmission [ 36 ]. Meng et al. [ 37 ] presented an approach to predict snowmelt floods in the Juntanghu watershed in China. A weather research and forecasting (WRF) model were used along with a snowmelt run-off model known as Tianshan Snowmelt Runoff Model (TSRM) which contains the snowmelt readings recorded during multiple years. Image data gathered from MODIS and DEM were used to predict floods using these hydrological models. The TSRM model driven by WRF was able to achieve 80% of condition ratios and determination coefficients of 0.85 and 0.82 for 2 years, respectively [ 37 ]. Boni et al. [ 38 ] combined data collected from Sentinel-1 and SAR to monitor floods in the Po River situated in Northern Italy. Image processing techniques such as thresholds, classification, and Region Growing Algorithm (RGA) were applied for the mapping of flood-prone areas [ 38 ]. The model achieved an overall user accuracy ranging from 60% to 80%. Li et al. used Sentinel-2 data along with data obtained from DEM having 90 m of spatial resolution. The noise data produced due to the presence of clouds, shadows, and frost were reduced using a Modified Normalized Difference Water Index (MNDWI), Revised Normalized Difference Water Index (RNDWI), Automated Water Extraction Index (AWEI), and Otsu threshold [ 39 ]. Google Earth Engine framework was used to calculate the water index and to extract water features. A root means a square error of 16.148 m was recorded using the proposed approach. Airborne SAR was studied for real-time flood area observation as well. Mason et al. [ 40 ] studied a method for selecting a subset automatically and in near real-time, which would allow the SAR water levels to be used in a forecasting model. Distributed water levels may be estimated indirectly along the flood extents in SAR images by intersecting the extents with the floodplain topography. It is necessary to select a subset of levels for assimilation because adjacent levels along the flood extent will be strongly correlated. Table 2 compares the performance outcomes and functionality of the multispectral remote sensing technologies used for flood prediction.
Multispectral remote sensing technologies for flood prediction.
LIDAR stands for Light Detection and Ranging. It is an active remote sensing technology that uses laser pulses to measure the distance of an object present on Earth from the sensor [ 51 , 52 , 53 ]. The Lidar system records other data from the Earth’s surface and along with the returned light pulses, the data obtained are used to create a 3D model which represents the properties of the Earth’s shape and surface. A LIDAR system thus consists of a laser scanner and a GPS. This technology has been used in applications that monitor and examine the Earth’s surface. A more common application of LIDAR technology is to generate DEM to be used in GIS which facilitates the emergency response operations. Hence, it has an immense potential to monitor water levels in water bodies to predict any future occurrence of a flood. Recently, several research articles have proposed methods for flood risk assessment and prediction using LIDAR remote sensing. Webster et al. [ 54 ] employed LIDAR to acquire details related to the rise of sea level to produce food risk maps. LIDAR data are used to construct a Digital Surface Model (DSM) and DEM which show the ground and non-ground regions and highlight the elevated and normal sea levels in the study area. The results were validated using GPS technology which shows accuracy that exceeds 30 cm [ 54 ]. Lamichhane and Sharma [ 55 ] developed a flood warning system using a DEM derived from LIDAR. The acquired LIDAR data were also integrated with some field data related to flooding in the target area to determine the evacuation time required by the people. Flood risk maps were produced by an HEC-GeoRAS, a software that allows the processing of geospatial data in ArcGIS [ 55 ]. The flood risk maps were then combined with digital orthographic maps to construct a real-time online flood warning system for the public [ 56 ]. Fadi et al. [ 57 ] used three channels of geometrical data derived from LIDAR. The first channel consists of survey data, the second channel is based only on the data acquired by LIDAR and the third one consists of a combination of the riverbank locations derived from survey and cross-sections data acquired by LIDAR technology [ 57 ]. The study aimed to predict the return period of the storm in the target area. The data were processed in the HEC-RAS tool to make flood-related predictions. The results showed that geometries obtained from LIDAR predicted floods with higher widths as compared with the predictions made by survey-derived geometries. Makinano and Santillan [ 58 ] integrated data from several resources to construct an early flood warning system. These sources include LIDAR, an open-source flood model, meteorological data, real-time hydrologic data and geographic visualization tools [ 58 ]. The acquired data are used to construct a two-dimensional (2D) hydraulic model using the HEC-RAS tool that produces accurate flood risk maps and provide early flood warnings. Stoleriu et al. [ 59 ] used high-density LIDAR derived data to improve the accuracy of flood risk maps generated by DEM [ 59 ]. HEC-RAS software was used to construct the flood hazard maps. The system was used to predict the flood reoccurrence probabilities in the durations of 33, 100, and 1000 years. The system can measure water levels up to an accuracy of 0.5 m. Table 3 summarizes LIDAR technologies for flood prediction.
LIDAR Technologies for Flood Prediction.
Radar (Radio Detection and Ranging) [ 60 , 61 , 62 ] was first used in the year 1940 by the navy department of America. As its name suggests, this remote sensing technology makes use of radio waves to find various characteristics of objects such as their direction, speed, location, and range. The organization of radar is composed of a transmitter that generates electromagnetic waves in the domain of radio or microwaves, an antenna for transmission, an antenna for receiving, a receiver, and a workstation that processes the object characteristics. The transmitter emits radio waves which are reflected by the object and then return to the receiver, where it is analysed by the processor to determine different object properties [ 61 ].
Once the detection through optical remotely sensed data fails, the synthetic aperture radar (SAR) comes into action. A high-resolution synthetic aperture radar (SAR) has been frequently used in the detection of areas affected by floods [ 63 ]. The technology provides real-time assessment of devastated and flooded areas. The prime quality of this technology is its penetration capacity to clouds, rain, and haze [ 64 ]. It does not matter whether it is a bad or drastic environment or too much sunlight, the technology provides effective expertise. The technology can easily distinguish between light and water. Radar uses microwaves; thus, flooding surfaces can be easily detected by its sensors. The flat surface of the water reflects the signals away from the sensor. This causes a decrease in the intensity of returned radiation as compared to the incident radiation causing a darker pixel in the image [ 65 ]. Thus, areas with water show dark pixels as compared to the pixels formed by the deflection through land areas.
Mitigation and management of floods require the analysis of the spatial extent and progressive pattern of remotely sensed images. The spatial extents of flooding are necessary to save lives and to avoid destruction. Combining this information with GIS and satellite data can help in estimating the damage caused by a flood [ 66 ]. Satellite transmissions involving microwaves revolutionized data extraction even in bad weather conditions and sunlight [ 38 , 39 ]. Data assimilation techniques facilitated the real-time integration of SAR-derived water levels and developed forecast models for disasters [ 67 ]. The integration of sensing data with data assimilation provided 3D reports of the flood used for the prediction of the flood as well as organizing the warning system for the flood. A problem faced by this technology is its inability to measure the long-term and real-time water level at fixed points. This is because of the orbital cyclic movement of the satellite. Thus, regardless of its high accuracy and real-time monitoring, it does not fit as the best technology for urban flood prediction. However, it works best for large water bodies including oceans and rivers [ 68 ]. Garcia-Pintado developed a flood prediction model that used SAR-derived water level observations in the Severn and Avon rivers situated in the United Kingdom (UK). The authors proved that by applying an Ensemble Transform Kalman Filter (ETKF) directly, some divergence in the filter was caused due to false correlations. To overcome this problem, a spatial filter localization method is proposed. Overall results showed that this model is feasible to work as an independent flood forecasting model that uses Earth Observations (EO) [ 69 ]. Table 4 summarizes radar technologies for flood pre-diction reviewed in this paper.
Radar technologies for flood prediction.
Three categories of remote sensing technologies adopted for flood prediction are Multispectral, LIDAR, and Radar. After detailed content analysis and examination of the methods adopted in each study for flood forecasting, we further categorized these studies based on the processing method used by the authors. For example, the use of machine learning (ML) methods for flood prediction has been commonly found in the literature [ 44 , 45 , 46 , 47 ]. This includes the use of CNN, AI, KNN, Bayesian linear, SVM, etc. used for flood forecasting. Image segmentation has been applied to remotely sense images to determine water fluctuation levels to build an active real-time urban flood warning system [ 31 ]. Apart from that, many studies [ 45 , 49 ] have used hydrological models for flood forecasting using remotely sensed data. Numerical modelling techniques which include thresholds and filters have been commonly used in the literature [ 54 , 57 ]. The difference between machine learning-based methods and hydrological models is that machine learning methods are data-driven and mainly depend on the training data to produce accurate results, while hydrological models are knowledge-based which implies that the human experts already feed them the knowledge to make flood-related decisions. Numerical modelling is a commonly used technique in geology, which is used to solve complicated geological problems. Numerical modelling methods work by simulating geological states and scenarios. They use mathematical models to define the physical properties of scenarios related to geology using numbers, calculations, and equations. Other less commonly used domains included decision making [ 32 ], image processing [ 55 ], and electromagnetic modelling [ 41 ]. These domains are categorized into the “Other” class. Figure 5 shows the distribution of the reviewed flood prediction techniques to the identified domains. This pie graph shows that the use of machine learning and hydrological models has been equally and most frequently observed in the literature for flood forecasting.
Methods adopted in studies.
RQ-3. What are the current gaps in remote sensing-based flood prediction technologies?
A significant issue in remote sensing technologies is the constraint regarding orbital cycles and spaces between trajectories of satellites, which makes the continuous monitoring of fixed objects a difficult task [ 31 , 71 ].
Floods, being natural disasters, are uncertain and unpredictable, which makes flood modelling a complex task having numerous uncertainties. Hence, hydrological and numerical models seldom provide imprecise results and fail to give reliable predictions regarding floods. In addition, these models are case specific, which means that they depend on the physical properties and climate of the specific study area [ 45 , 51 , 72 ]. Hence, they cannot be applied to a new area without modification. In addition, the high-resolution images captured by satellites are stored in satellite databases, making the processing slow and time consuming, with expected delays in the final output [ 53 ]. Image processing techniques, conversely, have some limitations regarding robustness and consistency, as the results are greatly affected by environmental conditions such as cloud cover, fog, rain, and pollution. Hence, most of the algorithms do not perform up to the mark in less-than-ideal conditions. To overcome these problems, researchers are rapidly adopting machine learning or artificial intelligence-based methods. The machine learning techniques are robust to the quality of input images, as these algorithms are trained using a wide variety of images having different illumination conditions, scale, quality, and colour, which enables the model to handle varying inputs. A limitation of machine learning models is that they must be trained using discriminative and relevant data, as the presence of noise and irrelevant information decreases the performance of the model [ 55 ]. In addition, these models require structured data that is the data must be labelled, which is an extensive and tedious task. Hence, there is an inherent need to efficiently extract features from the remotely sensed data and use it for training.
Insights and Future Directions
Machine learning techniques manage the uncertainties related to natural disasters such as floods in an efficient way. The limitations of machine learning models can be overcome by providing reliable historic flood data and flood inventory maps to the machine learning model [ 31 , 73 ]. Machine learning models offer a cheap and time-efficient solution to predict floods and perform a flood risk assessment. They also inform the experts about the need for additional data such as data related to rainfall run-off and river flow. By providing this data, the model can generate more accurate prediction results.
The extensive labelling and feature extraction steps performed in machine learning methods can be reduced by using a deep learning approach. Deep learning models can use unstructured data and can automatically perform feature extraction. The research in this paper shows that researchers, to deal with disasters such as floods, have rarely adopted deep learning methods. Deep learning models have not been well experimented with or documented for flood risk analysis [ 42 , 74 ]. Hence, this domain needs to be further explored. This can be achieved by retrieving data from numerous sources, including disaster history, satellite imaging, and weather reports. The data gathered can be used to train the deep learning system. The deep learning model would be able to forecast the upcoming disaster events and be trained to perform real-time flood mapping. A proposed system for flood detection and extent mapping is demonstrated in Figure 6 .
Schematic diagram of proposed system for flood detection and extent mapping.
To overcome the issues faced by remote sensing technologies, the proposed system uses data from various sources along with satellite data such as past flood events, Google Earth Images, social media, and weather reports. The collected data would consist of both text and images, hence providing rich information about the nature, causes, locations, and effects of flood events. The data would be utilized to train a deep learning model such as CNN to predict future flood events and do real-time flood mapping in an input image. A team of experts can help in disaster management with the analysis reports produced by a deep learning system. In this way, proactive measures could be taken, and imminent devastation can be prevented. Natural disasters cannot be prevented; thus, one should take active steps to protect themselves from these situations. Using the deep learning system with drones such as UAVs, experts can gather real-time data from various areas and perform flood mapping at the same time. The drone technologies could track specific areas and deliver help in narrow territories. Deep learning, in today’s life, is ready to show its potential in disaster management. Classical machine learning models such as SVM, Naïve Bayesian, and Decision Trees require extensive steps of labelling the training data and selecting the relevant and discriminative feature for training the model. Deep learning models, conversely, have an inherent capability of efficiently handling unstructured data and automatically extracting features from it. Hence, the images in the training set do not have to be labelled and can be used directly for training. This saves time and reduces the complexity of the system. Some most common deep learning models are Convolutional Neural Networks (CNN), Recurrent Neural Networks (RNN), and Long Short-Term Memory (LSTM). CNN has been widely used for image classification and segmentation problems [ 43 , 75 ]. RNN models can efficiently handle temporal data and have been applied to video processing problems. LSTM model is an improved form of RNN, which can retain information for longer periods, unlike a standard RNN model. Hence, the future direction for using technological advancements for dealing with floods would be to investigate the use of deep learning for real-time flood mapping and prediction. To find the depth of floodwater in a region, DEM can be integrated into the system, such that rescue activities could be prioritized in the regions having deeper floodwater. The framework proposed in this study aims to bridge the gaps identified in various flood management technologies belonging to different domains. For real-time flood mapping in emergency scenarios, time is one of the most important factors to be considered, and the system should be efficient enough to map the flooded regions immediately, such that rescue operations can be initiated as soon as possible [ 76 ]. However, in most of the research studies, this measure was not evaluated in the performance results. The researchers should specify the time taken by their system to produce the results, as a slow and lagging system is not suitable to be implemented in an emergency. This would help in the better assessment of flood management technologies in the future. More research needs to be focused on using technologies to facilitate post-flood rescue and relief operations. This includes route finding, vehicle detection, and locating affected people such that the people stuck in flood-related crises could be identified and rescued by finding the available routes and transport facilities.
This article presents a systematic review of remote sensing technologies used for flood predictions. The review indicated that there is a rapid surge of studies implementing AI techniques coupled with remote sensing techniques for flood prediction. Based on a content analysis methodology, a review of 76 relevant papers on remote sensing technologies for flood prediction was presented. A classification framework for flood detection and mapping was proposed that aimed to answer the proposed questions: (1) How can remote sensing technologies for flood detection be categorized? (2) How are remote sensing technologies being used for flood prediction? (3) What are the current gaps in remote sensing-based flood prediction technologies?
To measure the depth of floodwater in a region, a high-resolution DEM can be used. Deep learning is a sub-domain of machine learning that uses neural networks that can undergo unsupervised learning from unstructured or unlabeled data. A framework has been proposed in this paper that uses a deep learning neural network to generate reports regarding the detection of flood from an input image and predict future flood events by learning from big data collected from various sources such as historic flood events, social media posts, Google and satellite images. A DEM module has been added to determine the extent of flooding in each area. This system is focused on disaster prediction and response. Detection of flooding from images can accelerate disaster relief services in the target region, thus assisting in the domain of disaster response. By finding the depth of floodwater in an area, rescue activities can be prioritized in the regions which have deeper flood water.
The outcomes of this research support the United Nations International Strategy for Disaster Reduction and Sendai Framework for Disaster Risk Reduction 2015–2030 [ 32 ]. As with the application of remote sensing technologies, the priority action of the Sendai framework can be met, which focuses on understanding the disaster risk, managing it, reducing disaster by building resilience and enhancing disaster preparedness through effective response and recovery. In addition, this study can assist the national disaster management authorities in the implementation of state-of-the-art technology for flood prediction, detection, and management. Other countries frequently hit by flood-like disasters can also benefit from this research. In the future, this study can be extended to include more techniques as well as identifying various domains, parameters and metrics for effective detection, prediction, and response to flood-related disasters and the assessment of various technologies. By defining a wide range of assessment measures, the techniques can be more thoroughly examined. We aim to implement the proposed flood management system proposed in real time to assess its limitations and practicability.
Methodology, H.S.M. and A.W.A.H.; investigation, A.W.A.H. and S.T.W.; writing—original draft preparation, H.S.M. and A.W.A.H.; writing—review and editing, A.W.A.H. and S.T.W.; supervision A.W.A.H. and S.T.W. All authors have read and agreed to the published version of the manuscript.
This research received no external funding.
Institutional Review Board Statement
Informed Consent Statement
Data availability statement, conflicts of interest.
The authors declare no conflict of interest.
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.
- View all journals
- My Account Login
- Explore content
- About the journal
- Publish with us
- Sign up for alerts
- Open access
- Published: 18 November 2022
Flood risk management through a resilience lens
- Karin M. de Bruijn ORCID: orcid.org/0000-0003-1454-5338 1 ,
- Bramka A. Jafino ORCID: orcid.org/0000-0001-6872-517X 1 , 2 ,
- Bruno Merz ORCID: orcid.org/0000-0002-5992-1440 3 , 4 ,
- Neelke Doorn ORCID: orcid.org/0000-0002-1090-579X 5 ,
- Sally J. Priest ORCID: orcid.org/0000-0003-2304-1502 6 ,
- Ruben J. Dahm ORCID: orcid.org/0000-0002-8667-3358 7 ,
- Chris Zevenbergen ORCID: orcid.org/0000-0003-0807-5253 8 , 9 ,
- Jeroen C. J. H. Aerts ORCID: orcid.org/0000-0002-2162-5814 10 , 11 &
- Tina Comes 5
Communications Earth & Environment volume 3 , Article number: 285 ( 2022 ) Cite this article
- Environmental studies
- Natural hazards
- Water resources
To prevent floods from becoming disasters, social vulnerability must be integrated into flood risk management. We advocate that the welfare of different societal groups should be included by adding recovery capacity, impacts of beyond-design events, and distributional impacts.
Societies have prospered in river valleys, deltas, and coastal areas thanks to effective strategies to cope with flood hazards. However, floods have been increasing in frequency and severity due to climate change and increasing exposure. Governments worldwide aim to develop strategies to reduce flood risks, usually favoring the measures with the largest risk reduction benefits and the lowest costs for a range of sufficiently likely hazard events. Here, the costs conventionally considered are the direct damages.
The high impact of recent extreme but rare events such as the 2022 floods in Pakistan and Malawi, the July 2021 flood in Northwestern Europe, the devastation due to Hurricane Iota in the Central Americas (2020), or the 2017 flooding of Houston, Texas, have brought us to rethink flood risk management. In conventional risk analyses rare, extreme events typically have little importance, because the expected annual damage–the indicator of conventional risk approaches–is often dominated by events that have a high probability but cause relatively low damage. Risk reduction measures conventionally aim to reduce direct impacts and total flood risks while minimizing costs. In contrast, it is rarer for measures to be implemented that enhance the ability to cope with flood hazards and to recover rapidly, to reduce indirect flood effects and to account for the distribution of impacts over wealthier and poorer communities 1 This may result in strategies that amplify existing inequalities, promote already wealthy societal groups 2 and neglect disastrous outliers.
Climate change and the related increase in flood hazards require additional investments into flood risk management. This opens a window of opportunity to ensure new investments contribute to a fairer and more resilient world. We argue that policy makers should adopt a resilience lens that utilises more comprehensive analyses, rooted in societal welfare.
Adopt a resilience lens
To develop flood risk management strategies, governments need to consider what really matters, namely how and over what period floods affect societal welfare. To do so, we advocate the adoption of a resilience lens in flood risk management. Here, resilience is understood as the ability of a society to cope with flood hazards by resisting, absorbing, accommodating, adapting to, transforming and recovering from the effects of floods on people’s welfare 3 , 4 . To analyze and enhance resilience, we need to consider how and over what period floods affect societies and how measures could affect flood impacts and society 5 . Questions to consider include whether floods will hamper economic activities; whether people can earn sufficient income or their livelihoods are destroyed and whether their health will be affected.
Adopting a resilience lens means taking societal welfare as our starting point. From there, the interaction with flood hazards and flood risks can be considered 6 . For frequent events resistance may be required to allow societies to continue functioning without facing frequent damage. Damage as a result of rare and extreme events may not be avoidable, but such events must be included in our considerations in order to make sure that those events, although damaging, do not turn into disasters. This requires a deep understanding of what makes people vulnerable to floods and how resilience can be improved. We offer four elements linked to this resilience lens to understand what makes a flood disastrous. We aim to enable an informed discussion on how to arrive at appropriate flood risk management strategies (see Fig. 1 ).
A welfare and recovery capacity (element 1 and 2): Different effects of floods on different areas or societal groups: some have a larger deterioration of welfare or a slower recovery than others. Both the maximum impact and the recovery together determine the impact of a flood disaster. B include beyond-design events (element 3). The grey curve shows the impacts as a function of event extremity. The standard assessment integrates over this curve and uses the resulting expected annual damage as risk measure; this aggregation undermines the role of high-impact but low-probability events. The extreme events must be given attention as well; ( C ) distributional impacts (element 4). Distributional impacts can be considered spatially or for different social groups. Welfare economics principles can be applied to capture the utility of different communities and vulnerable groups. By aggregating the effects, we may not see how some groups benefit from measures while others pay for them, or still face large risks. Therefore, next to total cost and benefits, also distributed impacts must be used and weighted to enhance equity.
Impacts on welfare, instead of on asset losses
Floods hit socially vulnerable people harder, because poorer communities often lack the capacity to recover quickly. Vulnerable people or communities have a lower capacity to anticipate, cope with, resist or recover from the impact of hazards 7 . They may be forced to live in hazardous places, have less access to flood warnings, a less effective network to enhance recovery, and fewer resources to protect their homes or livelihoods. Especially people that already live in poverty may need to shift to destructive strategies such as selling land or cattle or consume seeds to meet other short-term needs. Such strategies can lead to a vicious circle.
Using absolute asset-based damages as yardsticks, as is often done in flood risk management, largely underestimates the disproportionally large welfare impact relatively small absolute losses can have on poor people and may lead to biased planning 8 . As one dollar does not count equally for all people, flood risk planning should move beyond asset-based valuations and put the welfare of people at the core of the assessment 9 . This can be done, for example, by considering social impacts such as loss of houses (irrespective of their value), deprivation cost, loss of percentage of income, or considering the effect on income generating ability.
There are further merits to placing welfare upfront. First, it opens the possibility of better aligning flood risk management with the larger development agenda 3 , for instance by linking flood risk management to spatial and economic planning. Second, it allows for a better inclusion of non-structural measures in flood risk management strategies, such as adaptive social protection systems that can quickly disburse financial assistance to households when a disaster hits 10 . Such measures may not reduce asset-based damages but can have significant benefits of increasing recovery rate and dampening welfare losses.
When recovery from floods takes longer, the impact of the floods is more disastrous because of the many indirect and cascading effects, which often exceed the direct damage 11 . Differences in flood impacts across societal groups often link to differences in their ability to recover from flood impacts. To recover, physical damage must be repaired and income generating options must be restored. Accounting for disruption of services of critical infrastructure, cascading impacts 12 or addressing people’s recovery capacity are thus crucial to understand the impact of floods on societal welfare. If we consider recovery as part of flood risk management, the effect of recovery enhancing measures can be included to reduce longer-term welfare loss. Measures such as citizen training, micro-credits, affordable insurance to compensate for flood losses and improving critical infrastructure (enhancing its robustness, redundancy, or flexibility) then become relevant.
The July 2021 floods in Europe have shown the devastating impact of beyond-design events, events that exceed the known risks. The flood peak discharge in July 2021 in the Ahr valley was roughly five times higher than the extreme event scenario of the official flood map 13 and its return period was estimated to be around 500 years. Such an event was beyond the imagination of people and authorities, which led to high numbers of fatalities and massive destruction.
The complexity of flood risk systems, limitations of scientific knowledge but also motivational and cognitive biases in perception and decision making contribute to such surprises 14 , 15 . In many regions, climate change and other drivers of change, such as population growth or increasing vulnerability, lead to more frequent situations where current protection systems are overwhelmed. Our third element targets this blind spot of flood risk management: extreme events beyond current design standards to prevent disastrous surprises.
This can be done for example by using a storyline approach, narrative scenarios or training exercises and simulation games that stimulate decision-makers to think through the full disaster cycle. Such exercises are known to inspire discussion of potentially long-term unexpected or unintended cascading effects across different systems 16 . Outliers in ensemble forecasts may be used as a starting point for such scenarios. These explorations guide dialogues towards achieving the desired level of protection and preparedness for extreme events, to reduce the impact to the most crucial objects, locations, or groups of a society, and provide the basis for training of decision-makers.
Distributional impacts and equity
A resilience lens requires asking the distributional questions of “the five Ws“ 17 : for whom, when, what, where, and why? Most flood risk analyses aggregate risks and flood protection benefits and disregard their distribution across people, space and time. The resilience lens requires unpacking this aggregation by assessing the distributional impacts of alternative measures. Making explicit who wins and who loses can support distributive justice and prevent unintended distributional consequences. Additional measures for compensating worse-off groups can also be prepared. It is one option, for example, to target flood risk protection measures 18 at the most socially vulnerable instead of selecting measures based on utilitarian principles. To do so, a risk analysis that shows distributed impacts on a range of social groups and regions must be carried out. These distributional questions also play out between current and future generations (intergenerational justice).
The distributional performance of alternative plans can be assessed through a normative analysis. Various ethical principles drawn from theories of distributive justice can be operationalized to evaluate the fairness of alternative measures 19 . Multiple principles can also be combined. In the Netherlands, the flood protection standard is designed such that every person has at least a minimum level of safety (sufficientarian principle), while additional safety margin is allowed if it is economically sensible (utilitarian principle) 20 .
We make a plea for more comprehensive, better-informed and transparent decision-making which allows an open discussion of inherent trade-offs between different values or ambitions, and makes transparent the impact of flood risk management over space, time and population groups. Disparities in flood risk and in effects of risk on people’s welfare should be understood and transparently shown to enable decision-makers to take equitable and effective decisions and to prevent increasing inequity due to climate change.
We now have the appropriate tools and methods available to adopt a resilience lens by analyzing distributional impacts, by assessing impacts on welfare, and by including recovery and longer-term consequences for both design and beyond-design events. Using this broader perspective will lead to other flood measures that better serve our joint journey towards a more just and resilient world.
Barbier, E. B. & Hochard, J. P. The impacts of climate change on the poor in disadvantaged regions. Rev Environ. Econ. Policy 12 , 26–47 (2018).
Article Google Scholar
Hino, M. & Nance, E. Five ways to ensure flood-risk research helps the most vulnerable. Nature 595 , 27–29 (2021).
Article CAS Google Scholar
UNDRR. Global assessment report on disaster risk reduction. United Nations Office for Disaster Risk Reduction (UNDRR) (2019).
De Bruijn, K. et al. Resilience in practice: Five principles to enable societies to cope with extreme weather events. Environ. Sci. Policy 70 , 21–30 (2017).
Di Baldassarre, G. et al. Floods and societies: the spatial distribution of water-related disaster risk and its dynamics. Wiley Interdiscip. Rev.: Water 1 , 133–139 (2014).
Aerts, J. C. J. H. et al. Including Human Behavior in Flood risk assessment. Nat. Clim. Change 8 , 193–199 (2018).
Blaikie, P. et al. At Risk. Natural hazards, people’s vulnerability and disasters. Routledge, London and New York. 284 p (1994).
Hallegatte, S. & Walsh, B. Natural disasters, poverty and inequality: New metrics for fairer policies. In The Routledge Handbook of the Political Economy of the Environment (pp. 111-131). Routledge (2021).
Kind, J., Botzen, W. J. & Aerts, J. C. J. H. Accounting for risk aversion, income distribution and social welfare in cost-benefit analysis for flood risk management. Wiley Interdiscip. Rev.: Clim. Change 8 , 1–20
Bowen, T. et al. Adaptive social protection: Building resilience to shocks. World Bank Publications (2020).
Merz, M. et al. A composite indicator model to assess natural disaster risks in industry on a spatial level. J. Risk Res. 16 , 1077–1099 (2013).
Arrighi, C., Pregnolato, M. & Castelli, F. Indirect flood impacts and cascade risk across interdependent linear infrastructures. Nat. Hazards Earth Syst. Sci. 21 , 1955–1969 (2021).
Kreienkamp, F. et al. Rapid attribution of heavy rainfall events leading to the severe flooding in Western Europe during July 2021. World weather attribution, www.worldweatherattribution.org (2021).
Merz, B. et al. Charting unknown waters - On the role of surprise in flood risk assessment and management. Water Resour. Res. 51 , 6399–6416 (2015).
Dilling, L., Morss, R. & Wilhelmi, O. Learning to Expect Surprise: Hurricanes Harvey, Irma, Maria, and Beyond. J. Extreme Events 4 , 3, 1771001 Brief Report (2017).
Wright, G. & Goodwin, P. Decision making and planning under low levels of predictability: Enhancing the scenario method. Int. J. Forecast. 25 , 813–825 (2009).
Meerow, S. & Newell, J. P. Urban resilience for whom, what, when, where, and why? Urban Geogr 40 , 309–329 (2019).
Sayers, P., Penning-Rowsell, E. C. & Horritt, M. Flood vulnerability, risk, and social disadvantage: current and future patterns in the UK. Reg. Environ. Change 18 , 339–352 (2018).
Jafino, B. A., Kwakkel, J. H. & Taebi, B. Enabling assessment of distributive justice through models for climate change planning: A review of recent advances and a research agenda. Wiley Interdiscip. Rev.: Clim. Change 12 , 1–23 (2021).
Kaufmann, M., Priest, S. & Leroy, P. The undebated issue of justice – Silent discourses in Dutch flood risk management. Reg. Environ. Change 18 , 325–337 (2018).
Authors and affiliations.
Deltares, Department of Flood Risk Management, Delft, The Netherlands
Karin M. de Bruijn & Bramka A. Jafino
Global Facility for Disaster Reduction and Recovery, The World Bank Group, Washington, DC, USA
Bramka A. Jafino
GfZ German Research Centre for Geosciences, Hydrology, Potsdam, Germany
University of Potsdam, Institute for Environmental Sciences and Geography, Potsdam, Germany
Delft University of Technology, Faculty of Technology, Policy and Management, Delft, The Netherlands
Neelke Doorn & Tina Comes
Flood Hazard Research Centre, Middlesex University London, London, UK
Sally J. Priest
Deltares, Department of Catchment & Urban Hydrology, Delft, The Netherlands
Ruben J. Dahm
IHE Delft Institute for Water Education, Water Engineering Department, Delft, The Netherlands
Delft University of Technology, Department of Hydraulic Engineering, Faculty of Civil Engineering, Delft, The Netherlands
Deltares, Department of Information, Resilience & Planning, Delft, The Netherlands
Jeroen C. J. H. Aerts
VU IVM Amsterdam, Department of Climate and Water Risk, Amsterdam, The Netherlands
You can also search for this author in PubMed Google Scholar
K.dB.: Initiator, conceptualization, and writing – first draft. R.J.D.: Initiator, conceptualization, review and editing. B.A.J.: Conceptualization, visualisation, and writing – review and editing. B.M., N.D., S.J.P., R.J.D., C.Z., J.C.J.H.A. and T.C.: Conceptualization and writing – review and editing.
Correspondence to Karin M. de Bruijn .
The authors declare no competing interests.
Peer review information.
Communications Earth & Environment thanks Duran Fiack, Gyan Kumar Chhipi-Shrestha and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Primary Handling Editors: Joseph Aslin, Heike Langenberg.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/ .
Reprints and permissions
About this article
Cite this article.
de Bruijn, K.M., Jafino, B.A., Merz, B. et al. Flood risk management through a resilience lens. Commun Earth Environ 3 , 285 (2022). https://doi.org/10.1038/s43247-022-00613-4
Received : 19 April 2022
Accepted : 31 October 2022
Published : 18 November 2022
DOI : https://doi.org/10.1038/s43247-022-00613-4
Share this article
Anyone you share the following link with will be able to read this content:
Sorry, a shareable link is not currently available for this article.
Provided by the Springer Nature SharedIt content-sharing initiative
This article is cited by
Timewise: temporal dynamics for urban resilience - theoretical insights and empirical reflections from amsterdam and mumbai.
- Supriya Krishnan
- Nazli Yonca Aydin
npj Urban Sustainability (2024)
Harbingers of decades of unnatural disasters
- Friederike E. L. Otto
- Emmanuel Raju
Communications Earth & Environment (2023)
- Explore articles by subject
- Guide to authors
- Editorial policies
Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.
Click through the PLOS taxonomy to find articles in your field.
For more information about PLOS Subject Areas, click here .
GIS-based flood hazard mapping using relative frequency ratio method: A case study of Panjkora River Basin, eastern Hindu Kush, Pakistan
Roles Formal analysis, Software, Writing – original draft, Writing – review & editing
Affiliation Institute of Natural Disaster Research, School of Environment, Northeast Normal University, Changchun, China
Roles Conceptualization, Supervision, Writing – review & editing
* E-mail: [email protected]
Affiliations Institute of Natural Disaster Research, School of Environment, Northeast Normal University, Changchun, China, State Environmental Protection Key Laboratory of Wetland Ecology and Vegetation Restoration, Northeast Normal University, Changchun, China, Key Laboratory for Vegetation Ecology, Ministry of Education, Changchun, China
- Kashif Ullah,
- Jiquan Zhang
- Published: March 25, 2020
- Reader Comments
Flood is the most devastating and prevalent disaster among all-natural disasters. Every year, flood claims hundreds of human lives and causes damage to the worldwide economy and environment. Consequently, the identification of flood-vulnerable areas is important for comprehensive flood risk management. The main objective of this study is to delineate flood-prone areas in the Panjkora River Basin (PRB), eastern Hindu Kush, Pakistan. An initial extensive field survey and interpretation of Landsat-7 and Google Earth images identified 154 flood locations that were inundated in 2010 floods. Of the total, 70% of flood locations were randomly used for building a model and 30% were used for validation of the model. Eight flood parameters including slope, elevation, land use, Normalized Difference Vegetation Index (NDVI), topographic wetness index (TWI), drainage density, and rainfall were used to map the flood-prone areas in the study region. The relative frequency ratio was used to determine the correlation between each class of flood parameter and flood occurrences. All of the factors were resampled into a pixel size of 30×30 m and were reclassified through the natural break method. Finally, a final hazard map was prepared and reclassified into five classes, i.e., very low, low, moderate, high, very high susceptibility. The results of the model were found reliable with area under curve values for success and prediction rate of 82.04% and 84.74%, respectively. The findings of this study can play a key role in flood hazard management in the target region; they can be used by the local disaster management authority, researchers, planners, local government, and line agencies dealing with flood risk management.
Citation: Ullah K, Zhang J (2020) GIS-based flood hazard mapping using relative frequency ratio method: A case study of Panjkora River Basin, eastern Hindu Kush, Pakistan. PLoS ONE 15(3): e0229153. https://doi.org/10.1371/journal.pone.0229153
Editor: Mou Leong Tan, Universiti Sains Malaysia, MALAYSIA
Received: October 21, 2019; Accepted: January 30, 2020; Published: March 25, 2020
Copyright: © 2020 Ullah, Zhang. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the manuscript and attached in Supporting Information files. Landsat-8(OLI) imagery (Date: 19-September-2018) and ASTER DEM can downloaded from USGS official website ( https://earthexplorer.usgs.gov ).
Funding: This research is supported by the national key research and development program of china (2018YFC1508804); The Key Scientific and Technology Research and Development Program of Jilin Province (20180201033SF); The Key Scientific and Technology Research and Development Program of Jilin Province (20180201035SF); The Key Scientific and Technology Program of Jilin Province (20170204035SF). The funders provided support in the form of salaries for authors [Jiquan Zhang], but did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Flood is the most prevalent and devastating natural disaster among all natural disasters that have adverse impacts on human health, natural and artificial environments [ 1 , 2 ]. Flood is a major risk to human life (loss of life, injury), assets (agriculture area, yield production, homes, and buildings), communication systems (urban infrastructure, bridges, roads, and railway lines), culture heritage, and ecosystems [ 1 – 3 ]. Literature indicates that more than 2000 deaths occur every year due to flooding, and more than 75 million people are adversely affected in one way or another across the globe [ 2 , 3 ]. Many factors, including both natural and anthropogenic are responsible for catastrophic flood incidents. Flood occurs due to heavy rainfall or snow melt that overflows to adjacent areas, or flood plains, and temporarily inundates the surrounding areas [ 4 , 5 ]. Recent studies, indicating that climate change is a fundamental factor that induces flood in various parts of the world [ 6 , 7 ], Charlton et al. [ 8 ] indicate that flood disasters in a region can be considerably influenced by changes in land use patterns forming an impermeable surface, which may increase flow velocity. Aside from these, many other factors that trigger flood occurrence are: slope, elevation, land use, curvature, Normalized Difference Vegetation Index (NDVI), proximity to rivers, etc., [ 9 , 10 ]. Due to the complex nature of floods, their frequent occurrence and extensive destruction across the globe, a large number of scientists have devoted significant effort to investigate and understand flood hazard for better mitigation and management [ 4 , 11 – 14 ].
Flood is a natural phenomenon and its complete prevention is not possible; however, the risk of the flood can be minimized by appropriate planning and mitigation measures. Flood management is one of the key steps in mitigation and risk reduction. Various studies have indicated that identification of flood risk zones and application of essential risk reduction measures (structural and non-structural) can effectively reduce flood losses to an acceptable level [ 14 , 15 ]. Moreover, flood hazard mapping plays a significant role in flood planning, early warning systems, emergency response services, and design of flood risk reduction measures [ 14 , 16 ]. So far, various studies have been conducted to assess and map flood-prone areas in different regions of the world [ 9 , 17 , 18 ]. The study of Guo et al.[ 14 ] stated that the scope of conventional approaches for flood hazard mapping is usually narrow, due to a lack of sufficient data. For example, rainfall-runoff modeling methods, watermarks on buildings, models involving numerical simulations, etc., are not appropriate for comprehensive river and flood analysis [ 2 , 10 ]. The acquisition of adequate data for flood mapping using these methods and similar conventional techniques is expensive, time-consuming, and often not available at the watershed or regional level, especially in developing countries. Today, remote sensing and GIS are powerful tools and provide different data sources for hazard management, flood susceptibility, and its forecast [ 7 , 11 , 19 ].
Over the past few decades, numerous methods have been developed and used to investigate flood hazard and risk assessment. These methods include the analytical hierarchy process (AHP) [ 13 , 19 ], fuzzy logic and genetic algorithms [ 17 ], variable fuzzy theory [ 14 ], hydrological forecasting systems [ 20 , 21 ], random forest [ 22 ], artificial neural networks (ANNs) [ 18 , 22 ], adaptive neuro-fuzzy interface systems [ 23 ], logistic regression [ 24 ], weight of evidence [ 25 , 26 ], analytic network process (ANP) [ 27 ], statistical index [ 28 ], Shannon’s entropy [ 29 ], Copula-Based Bayesian Network [ 30 ], and frequency ratio models [ 1 , 25 , 31 ]. The ANN approach, which has been used for flood susceptibility mapping [ 18 , 32 ], tries to make an association between some input factors and an outcome. However, Tiwari and Chaterjee [ 33 ] reported that the length of the dataset can cause errors in the process of ANN modelling and also poor prediction. Das [ 12 ] applied AHP to map flood hazard zonation in the Vaitarna basin, Maharashtra, India. However the drawback of AHP lies in its dependence on expert opinion [ 34 ]. The most common statistical methods of logistic regression and frequency ratio (FR) can be considered as significant methods that use a simple and understandable perception [ 1 , 25 , 26 , 35 ]. Tehrany et al. [ 9 ] reported that logistic regression and FR models can generate acceptable flood risk maps, and the process of analysis is easily understandable. Among bivariate statistical models, the FR model is considered one of the most important method that is easy to apply and can produce acceptable risk analysis and mapping [ 9 , 26 , 35 , 36 ]. Accordingly, FR was selected from the set of bivariate statistical methods for this study. The results obtained from this model are easy to interpret. Although this model is infrequently used in flood hazard mapping, its superior performance has been proven in other fields of natural hazard such as landslides [ 34 , 37 – 40 ]. Furthermore, some studies show that bivariate statistical models sometimes have a higher accuracy than machine learning models, which require huge amounts of data as training for better accuracy [ 40 – 42 ]. FR is the bivariate statistical method that can consider the correlation between dependent factors (historical flood points) and independent factors (flood-causative factors) [ 1 , 25 , 43 ]. FR models have been successfully applied to flood susceptibility and vulnerability assessments in different flood prone regions of the world [ 1 , 25 , 26 ].
The Panjkora River Basin (PRB) is located in the eastern Hindu Kush region, Khyber Pakhtunkhwa province, Pakistan, which experiences flood events almost every year, generally during the monsoon seasons (June–September) [ 44 ]. Over the last decade, many disastrous floods have occurred in the region, which negatively affected human lives, property, agriculture, and other infrastructure [ 45 – 47 ]. The most devastating flood events have been recorded in the years 2005, 2010, 2014 and 2016. It has been reported by Rahman and Dawood [ 48 ] that climate change has intensified the spatiotemporal variability of rainfall, which poses serious threats to the local communities in the form of floods. In addition, the complex topography of the region coupled with the fragile socioeconomic condition of the local people triggers flood risk in the region [ 46 ]. So far, few studies have been conducted to assess flood hazards and map the flood-prone zones, especially in the middle and lower catchment of the PRB [ 46 , 47 ]. Therefore, the present study was designed to map the flood-prone areas in PRB and propose effective measures for flood risk reduction in the study region. The study is based on an integrated approach using ground-based observation, remote sensing, and relative frequency ratio (RFR) techniques. The current study is the first of its kind to map the flood-prone areas in the PRB using the RFR model.
2. Materials and methods
2.1 description of the study area.
The study area is located in the eastern Hindu Kush Khyber, Pakhtunkhwa province, Pakistan with the geographical extent of “34.33°–35.0° N latitudes and 71.0°–72.0° E longitudes” ( Fig 1 ). It covers the lower and middle catchments of the PRB, and comprises an area of 1,741 km 2 . A river runs through it northeast to southwest, joining up with tributaries and finally draining into the river Swat at Qalangi village [ 46 ]. Climatically, in winter, the temperature drops to -12 °C while in summer, the temperature rises to 35 °C. In monsoon seasons (June–September), the PRB receives more than 800 mm of rainfall [ 47 ]. In the study area, the soil structure varies from a clayey nature to loam and sandy loam. In most places, due to steep and delicate slopes, the ground is exposed and vulnerable to erosion. The fertile soils exist mostly on moderate slopes. Such areas are commonly used for agriculture.
- PPT PowerPoint slide
- PNG larger image
- TIFF original image
In recent years, the study area experienced disastrous floods in 2005, 2010, 2014, and 2016 with adverse impacts on people lives, property, agriculture, and infrastructure [ 46 , 47 ]. During the summer season, heavy rainfall causes floods in the region, and sometimes the extraordinary activity of the monsoon causes high surface run-off and peak discharge.
2.2 Flood inventory mapping
The database of past floods is important to the study of the relationship between different flood triggering factors and flood occurrence [ 18 , 49 ]. Moreover, the accuracy of the flood susceptibility mapping greatly relies on the accuracy of previous floods events [ 7 , 25 , 49 ]. In the present study, the flood inventory database was created after identifying 154 flood points using existing flood reports of the National Disaster Management Authority, Pakistan, Provincial Disaster Management Authority, Khyber Pakhtunkhwa, field surveys, and interpretation of satellite and Goggle earth images before and after the 2010 devastating flood in the target area. Based on the literature reviews, 70% of flooded locations (107 locations) were selected randomly as a training dataset to prepare the flood hazard map and 30% of the locations (47 locations) were used for validation of the results ( Fig 2 ) [ 7 , 26 , 50 ].
2.3 Identification of flood triggering and causal factors
To evaluate the flood vulnerability, it was necessary to investigate a series of flood triggering and causal factors and their relationship with flooding [ 51 , 52 ]. In past studies, different flood-controlling factors have been used [ 1 , 12 , 13 ]. There is no specific guideline for selecting flood-controlling factors that affect flood occurrence. The selection of flood-controlling factors is an important step for flood hazard mapping and depends on physical and natural characteristics of the study area and data availability [ 18 , 53 ]. The methodology adapted for this study is shown in Fig 3 . To prepare the flood susceptibility map for the PRB, various satellite images and ancillary datasets were acquired from government organizations and web sources: (i) Advanced Spaceborne Thermal Emission and Reflection Radiometer Digital Elevation Model (ASTER DEM) of 30 m spatial resolution; (ii) Landsat 8 (OLI) imagery (Date: 19-September-2018) are downloaded from USGS official website ( https://earthexplorer.usgs.gov ); and (iii) monthly rainfall data from 1980 to 2016 collected from the Regional Meteorological Center, Peshawar. In this study, we have identified and selected eight flood causative factors, namely, slope, elevation, curvature, TWI, land use and land cover (LULC), rainfall, NDVI, and drainage density to generate thematic layers for flood hazard mapping based on a literature review and local conditions [ 10 , 13 , 20 ]. Moreover, ArcGIS (10.2), SAGA GIS, and Erdas were used to generate the required thematic layers. The relationship of each factor with flooding is discussed below in Table 1 and illustrated in Figs 4 and 5 .
2.4 Relative frequency ratio model
Flood hazard assessment is an important technique in hydrological studies. In this study, an RFR model is used to map flood prone zones in the PRB. FR is a bivariate statistical analysis method, based on the spatial distribution (probability) dependent factor (flood location) and flood triggering and causal factors (i.e., slope, elevation, etc.) [ 25 , 42 ].
The bivariate probability of each independent flood triggering factor was determined by its relationship with flood occurrence [ 1 , 25 ]. The higher the bivariate probability (greater than 1) the stronger is the correlation between flood incidence and flood triggering factors, and the lower the probability (less than 1), the weaker the correlation [ 1 , 25 , 50 ].
The FR values were calculated using ( Eq 3 ) for all sub-classes of flood triggering factors based on their relationships with flood inventory, as shown in Table 2 .
In the next step, the FR was normalized in a range of probability values [0, 1] as relative frequency (RF) using Eq 4 .
3. Results and discussion
In this study, the flood susceptibility of the PRB has been assessed by using an integrated approach of the bivariate statistical method (FR) with geospatial techniques. FR was used to calculate the correlation between flood occurrence and flood triggering factors. Table 2 shows the relationship between different flood causative factors, sub-classes, and flood occurrence in the PRB. Eight flood-triggering factors, namely, elevation, slope, drainage density, LULC, curvature, NDVI, TWI, and rainfall were used in the study. There is a direct positive relationship between FR and flood probability.
Elevation is an important factor of flood occurrence, as water always flows from higher locations to low land areas [ 52 ]. The elevation class 577–913 m has the maximum RF value of 0.56, followed by 913–1146 m and 1146–1675 m with RF values of 0.15 and 0.12, respectively. The analysis reveals that almost 65% of past floods occurred in the first three classes of elevation. Elevations higher than 2436 m have the lowest RF value (0.00, see Table 2 ). These results are in agreement with previous studies, which found a low probability of flood occurrence at higher elevated regions and a high probability of flooding in lowland areas [ 54 , 57 ].
Slope regulates the incidence of flooding, as lowland areas in the rainy season have a strong connection with the flood state. It has been reported that a lower slope gradient has more chances of flooding and flood events [ 51 , 56 ]. The infiltration process is also partly controlled by the slope gradient. An increasing gradient decreases the process of infiltration but increases the surface runoff; as a result, in regions having a sudden descent gradient, an enormous extent of water becomes stagnant and causes flood conditions [ 61 ]. The results show that the two lower slope gradient classes, i.e., <6.8° and 6.8°–15.4° have the highest RF value of 0.68 and 0.15, respectively. In contrast, the slope gradient above 29.4° shows the lowest RF value of 0.02 ( Table 2 ). Approximately 68% of fast floods occurred in PRB areas having slope lower than 25°. Fig 4b indicated that the lower slope gradients are pointed on both sides of the river.
Drainage density is considered an essential element of flooding. The higher likelihood of flooding is strongly linked to higher drainage density as it points toward a greater surface runoff [ 54 ]. In this study, the drainage density has a direct relationship with flooding. The probability of flooding increases with an increase in drainage density and decreases with a decrease in drainage density. Drainage density was divided into five classes using the natural break method ( Fig 4c ). The class 1.82–2.75 km/km 2 and 0.034–0.75 km/km 2 have the highest and lowest probability of flooding with RF values of 0.58 and 0.2, respectively ( Table 2 ). High drainage density refers to high surface runoff, therefore, high flood probability exists in areas having high drainage density [ 43 , 54 ].
Land use patterns reveal the type of utilization of land by people and natural processes [ 7 , 12 ]. Urban areas increase runoff due extensive impervious soil and fallow farmland increases runoff where there is no vegetation cover to control and prevent the rapid flow of water to the soil surface. There is risk of flooding and soil erosion in those areas; therefore, they are the most vulnerable areas to flooding. For LULC, the maximum weights were allocated to water bodies (RF = 0.61), followed by built-up areas (0.15) and agriculture areas (0.13), while forest and snow cover are least vulnerable areas in the region with RF values of 0.00 and 0.3, respectively ( Table 2 ). Built-up areas located in proximity to rivers are most vulnerable to flooding due to their economic resources, infrastructure, and large population [ 7 , 12 , 25 ].
Similarly, curvature is also an important factor and represents the morphology of the topography [ 12 , 25 , 62 ]. The curvature map is classified into three classes. A positive value of curvature represents a convex surface, zero a flat surface, and a negative value a concave surface [ 7 , 54 ]. The results show that the highest RF was obtained for the flat surface at the rate of 0.61, while the lowest RF was obtained for the concave surface at 0.15 ( Table 2 ). It was observed that approximately 83% past flood had occurred in flat and convex shape slopes.
The NDVI is another important conditioning factor of flooding. The index values range from -1 to +1[ 7 ]. Khosravi et al. [ 25 ] stated that the negative values show water and the positive values show vegetation so, NDVI has negative relationship with flooding: higher NDVI values indicate lower probability of flood and lower NDVI values indicate higher flood probability. In this study, the NDVI values range from -0.15 to 0.53 and were classified into five classes using a natural break method ( Fig 5b ). For the class -0.15 to 0.16, the RF was highest 0.43 ( Table 2 ), which means that there is a high probability of flooding in the study region [ 43 ].
The TWI was classified into five classes: <5.85, 5.85–7.69, 7.69–10.37, 10.37–14.30, and 14.30–23.67 ( Fig 5c ). The RF values for the TWI classes of 14.30–23.67 and 10.37–14.30 were calculated as the highest at 0.38 and 0.37, respectively. Similarly, the RF value for the TWI class of <5.85 was lowest at 0.04 ( Table 2 ). TWI has a direct positive relationship with flooding [ 12 , 25 ]. The higher TWI class refers to higher chances of flooding in the watershed [ 10 ]. The results indicate that the higher TWI was found in the south, northeast, and middle of the study area (represented with a blue color in Fig 5c ), and a low TWI was mostly present in the north and in steep slopes.
Except for glaciers, rainfall is the only source of water in the study region. A sudden rainfall in an area can cause flash flood conditions in semi-arid regions [ 12 ]. A large number of previous studies have established a relationship between rainfall and flooding [ 17 , 52 , 54 ]. The PRB is characterized by semi-arid climatic conditions, where an enormous amount of rainfall occurs summer season due Asian monsoon system which causes flash flood [ 63 ]. The rainfall map was reclassified into five classes with natural breaks. The highest RF value (0.29) was observed for class >81.43 mm followed by class 76.03–78.63, 73.42–76.03, and 69.84–73.42 with RF values of 0.26, 0.21, and 0.14, respectively ( Table 2 ). The lowest RF value of 0.11 was observed for class 78.63–81.42 mm. It is interesting to note that the class 78.63–81.42 mm is the second highest rainfall region but the least vulnerable, because this region is characterized by high elevation, high slope gradient, and dense forest and floods occur in lowland area. Therefore an increase in rainfall has no impact on flooding [ 25 ].
After the preparation of all eight layers of flood triggering and causal factors and giving weights to each parameter using FR and RF, a final hazard map was obtained by summation of each factor PR (weight) and each class RF in a raster calculator ArcGIS 10.2 environment using Eq 6 . The flood hazard index (FHI) values of the study area are found to lie in the range from 8302 to 100311. The FSI values of the total area were divided in five subclasses using a natural break method: very low, low, medium, high, and very high and indicated in Fig 5 . The analysis illustrates that approximately 15% of the total area is in a very high and high flood hazard zone, 14% is in medium, 42% is in low, and 29% is in safe areas ( Table 3 ).
In the study region, the slope has the maximum contribution to flooding with a PR value of 3.98 closely followed by LULC and elevation with PR values of 3.88, 3.41, respectively. The curvature, NDVI, and TWI have a medium influence on flood occurrence with PR values of 2.79, 1.92, and 1.81, respectively, while the drainage density and rainfall are the least important factors with PR values of 1.32 and 1.00, respectively, in determining flood susceptibility in the study region ( Table 4 ). Fig 6 indicated that most of the very high and high risk areas are located near the banks of rivers Panjkora with low slope gradient, low elevation, flat curvature, higher TWI, and higher drainage density. From the final hazard map, it is clear that agriculture practices, commercial activities, or people living in high and very high flood susceptible zones are highly vulnerable to future flooding in the study region.
3.1 Validation of flood hazard map
The primary objective of hazard mapping is to demarcate the areas that are prone to flood hazards. There are many models used by researchers to analyze flood susceptibility, but it is essential to validate the results of the model used for flood hazard assessment [ 61 , 64 ]. The receiver operating characteristic (ROC) method is frequently used for the validation of prediction maps [ 9 , 53 ]. Moreover, the method is simple and produces clear and reliable results [ 25 , 65 ]. Many studies have used this method to validate results [ 1 , 26 ]. In this study, we used the ROC method to evaluate the success and prediction rate of the flood hazard map based on the previous flood incidents. To validate the model, we compared the existing flood data with the acquired flood probability map [ 64 , 66 ]. The results of the success rate were obtained using the training data set, and the prediction accuracy was calculated using the validation dataset that was not used in the training process [ 7 , 61 , 67 ]. The ROC curve for this study is shown in Fig 7 , with AUC values of success and prediction accuracy of 82.04% and 84.74%, respectively.
Flood susceptibility mapping is an important step for future flood management. In hydrological and flood management studies, flood susceptibility maps are widely used to determine flood-prone zones. The present study aimed to assess flood hazards and map the flood-prone zones in the PRB, eastern Hindu Kush region. For this purpose, the RFR method was integrated with remote sensing and geospatial techniques to assess and map the flood hazard-prone areas. In this study, we used eight conditioning factors including slope, elevation, TWI, LULC, NDVI, drainage density, curvature, and rainfall to develop flood susceptibility maps. Overall, 154 flood-inundated locations were identified based on the damage and needs assessment report of the 2010 flood, field survey, interpretation of Landsat-7 and google earth images. The flood points were randomly divided into a training data set and testing data set. We used 70% (107 flood locations) of the points for building the model, and the remaining 30% (47 flood locations) points were employed in the validation of the probability model.
The flood hazard area was divided into five subclasses of hazard zones: very high, high, medium, low, and very low. The study found that approximately 15% of the total area is highly prone to flood hazard, 14% is moderately susceptible, 42% is low, and approximately 29% is very low. Furthermore, the study indicates that the high flood-prone areas are situated in the mid, southern, and western portions of the study area, as these areas are near the river with a low slope gradient, flat curvature, low elevation, high TWI value, and high drainage density. The ROC curve was used to measure the efficiency of the model and evaluate the results. The validation results showed good prediction efficiency with AUC values of success rate at 82.04% and of prediction rate at 84.74% of the flood susceptibility map. Therefore, the flood susceptibility map generated in this study can be considered an important tool to incorporate in flood risk management plans for disaster managers, decision-makers, and engineers. Based on the findings of this study, the concerned authorities can adopt appropriate mitigation and preparedness measures to minimize the impacts of prevailing and future floods.
We would like to thank Regional Meteorological Center, Peshawar for providing us rainfall data and United States Geological survey (USGS) for Landsat- 8 and ASTER DEM images. The authors greatly appreciate the reviewers and editors for their critical comments that greatly helped in improving the quality of this manuscript.
- View Article
- Google Scholar
Water Resources Management and Sustainability pp 213–220 Cite as
Flood Modelling Using HEC-RAS for Purna River, Navsari District, Gujarat, India
- Darshan J. Mehta 6 &
- Yennam Varun Kumar 6
- First Online: 02 February 2022
Part of the Advances in Geographical and Environmental Sciences book series (AGES)
The small and medium river banks are seriously flooded due to high-intensity of rainfall in monsoon seasons. The floods threatened human safety, life and property. This paper presents a process of 1D steady flow analysis used Hydrologic Engineering Center River Analysis System software. The one-dimensional modelling is applied on research Purna River, this river is one of the non-perennial rivers in Gujarat. This work contains flood model in which station, elevation of each cross-section was assessed at a particular section of the study reach. Steady flow analysis and hydraulic design analysis were carried out and after providing slope and discharge (flood event) at particular cross section software will compute the water surface elevation, depth of water and velocity of the water. The result from the research analysis could be used by flood management authorities to mensurate the flood at various cross-section of the study region.
- Purna River
- Steady flow analysis
This is a preview of subscription content, log in via an institution .
- Available as PDF
- Read on any device
- Instant download
- Own it forever
- Available as EPUB and PDF
- Compact, lightweight edition
- Dispatched in 3 to 5 business days
- Free shipping worldwide - see info
- Durable hardcover edition
Tax calculation will be finalised at checkout
Purchases are for personal use only
Agrawal R, Regulwar DG (2016) Flood analysis of Dhudhana river in upper Godavari basin using HEC-RAS. Int J Eng Res 1:188–191
Ahmad HF, Alam A, Bhat MS, Ahmad S (2016) One dimensional steady flow analysis using HECRAS–a case of River Jhelum, Jammu and Kashmir. Eur Sci J 12:340–350
Demir V, Kisi O (2016) Flood hazard mapping by using geographic information system and hydraulic model: Mert River, Samsun, Turkey. Adv Meteorol
Ingale H, Shetkar RV (2017) Flood analysis of Wainganga River by using HEC-RAS model. Int J Sci Eng Technol 6(7):211–215
Khattak MS, Anwar F, Saeed TU, Sharif M, Sheraz K, Ahmed A (2016) Floodplain mapping using HEC-RAS and ArcGIS: a case study of Kabul River. Arab J Sci Eng 41(4):1375–1390
CrossRef Google Scholar
Kumara YV, Mehtab DJ (2020) Water productivity enhancement through controlling the flood inundation of the surrounding region of Navsari Purna river, India. Water Prod J. https://doi.org/10.2203/WPJ.2021.264752.1024
Mehta DJ, Yadav SM (2020) Hydrodynamic simulation of River Ambica for riverbed assessment: a case study of Navsari Region. In: Advances in water resources engineering and management. Springer, Singapore, pp 127–140
Mehta DJ, Ramani MM, Joshi MM (2013a) Application of 1-D HEC-RAS model in design of channels. Methodology 1(7):4–62
Mehta D, Yadav SM, Waikhom S (2013b) Geomorphic channel design and analysis using HEC-RAS hydraulic design functions. Paripex Int J Glob Res Anal 2(4):90–93
Mehta DJ, Yadav SM, Waikhom SI (2017) HEC-RAS flow analysis in the River Tapi. In: Proceedings of the 37th IAHR world congress, Kuala Lumpur, Malaysia, vol 16
Mehta D, Yadav SM, Waikhom S, Prajapati K (2020) Stable channel design of Tapi River using HEC-RAS for Surat region, vol 91. Springer, Cham, pp 25–36
Mehta DJ, Eslamian S, Prajapati K (2021) Flood modelling for a data-scare semi-arid region using 1-D hydrodynamic model: a case study of Navsari Region. Model Earth Syst Environ 1–11
Parhi PK, Sankhua RN, Roy GP (2012) Calibration of channel roughness for Mahanadi River, (India) using HEC-RAS model. J Water Resour Prot 4(10):847–850
Patel KB, Sanjay Y (2019) One dimensional unsteady flow analysis using HEC-RAS modelling approach for flood in Navsari City. In: Proceedings of recent advances in interdisciplinary trends in engineering & applications (RAITEA)
Timbadiya PV, Patel PL, Porey PD (2011) Calibration of HEC-RAS model on prediction of flood for lower Tapi River, India. J Water Resour Prot 3(11):805
Timbadiya PV, Patel PL, Porey PD (2014) One-dimensional hydrodynamic modelling of flooding and stage hydrographs in the lower Tapi River in India. Curr Sci 708–716
Authors are thankful to the Navsari irrigation department for providing data of Purna River and Principal, Dr. S. & S. S. Ghandhy Government Engineering College, Surat for granting permission to carry out research work.
Authors and affiliations.
Dr. S & S. S. Ghandhy Government Engineering College, Surat, Gujarat, India
Darshan J. Mehta & Yennam Varun Kumar
You can also search for this author in PubMed Google Scholar
Editors and affiliations.
Department of Geography, Delhi School of Economics, University of Delhi, New Delhi, Delhi, India
Krishi Vigyan Kendra, Korba, Chhattisgarh, India
Gaurav Kant Nigam
Department of Environmental and Water Resources Engineering, Chhattisgarh Swami Vivekanand Technical University, Bhilai, Chhattisgarh, India
Manish Kumar Sinha
Dept of Geography, Aditi Mahavidyalya, University of Delhi, Delhi, Delhi, India
Rights and permissions
Reprints and permissions
© 2022 The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
About this chapter
Cite this chapter.
Mehta, D.J., Kumar, Y.V. (2022). Flood Modelling Using HEC-RAS for Purna River, Navsari District, Gujarat, India. In: Kumar, P., Nigam, G.K., Sinha, M.K., Singh, A. (eds) Water Resources Management and Sustainability. Advances in Geographical and Environmental Sciences. Springer, Singapore. https://doi.org/10.1007/978-981-16-6573-8_11
DOI : https://doi.org/10.1007/978-981-16-6573-8_11
Published : 02 February 2022
Publisher Name : Springer, Singapore
Print ISBN : 978-981-16-6572-1
Online ISBN : 978-981-16-6573-8
eBook Packages : Earth and Environmental Science Earth and Environmental Science (R0)
Share this chapter
Anyone you share the following link with will be able to read this content:
Sorry, a shareable link is not currently available for this article.
Provided by the Springer Nature SharedIt content-sharing initiative
- Publish with us
Policies and ethics
- Find a journal
- Track your research