Assessment of urban flood susceptibility using semi-supervised machine learning model

A comparative study on urban waterlogging susceptibility assessment based on multiple data-driven models

Accurate identification of urban waterlogging areas and assessing waterlogging susceptibility are crucial for preventing and controlling hazards. Data-driven models are utilized to forecast waterlogging areas by establishing intricate relationships between explanatory variables and waterlogging states. This approach tackles the constraints of mechanistic models, which are frequently complex and unable to incorporate socio-economic factors. Previous research predominantly employed single-type data-driven models to predict waterlogging locations and evaluation of their effectiveness. There is a scarcity of comprehensive performance comparisons and uncertainty analyses of different types of models, as well as a lack of interpretability analysis. The chosen study area was the central area of Beijing, which is prone to waterlogging. Given the high manpower, time, and economic costs associated with collecting waterlogging information, the waterlogging point distribution map released by the Beijing Water Affairs Bureau was selected as labeled samples. Twelve factors affecting waterlogging susceptibility were chosen as explanatory variables to construct Random Forest (RF), Support Vector Machine with Radial Basis Function (SVM-RBF), Particle Swarm Optimization-Weakly Labeled Support Vector Machine (PSO-WELLSVM), and Maximum Entropy (MaxEnt). The utilization of diverse single evaluation indicators (such as F-score, Kappa, AUC, etc.) to assess the model performance may yield conflicting results. The Distance between Indices of Simulation and Observation (DISO) was chosen as a comprehensive measure to assess the model's performance in predicting waterlogging points. PSO-WELLSVM exhibited the highest performance with a DISOtest value of 0.63, outperforming MaxEnt (0.78), which excelled in identifying areas highly susceptible to waterlogging, including extremely high susceptibility zones. The SVM-RBF and RF models demonstrated suboptimal performance and exhibited overfitting. The examination of waterlogging susceptibility distribution maps predicted by the four models revealed significant spatial differences due to variations in computational principles and input parameter complexities. The integration of four WSAMs based on logistic regression has been shown to significantly decrease the uncertainty of a single data-driven model and identify the most flood-prone areas. To improve the interpretability of the data model, a geographical detector was incorporated to demonstrate the explanatory capacity of 12 variables and the process of waterlogging. Building Density (BD) exhibits the highest explanatory power in relation to explain waterlogging susceptibility (Q value = 0.202), followed by Distance to Road, Frequency of Heavy Rainstorms (FHR), DEM, etc. The interaction between BD and FHR results in a nonlinear increase in the explanatory power of waterlogging susceptibility. The presence of waterlogging susceptibility risk in the research area can be attributed to the interactions of multiple factors.

Prediction of flood sensitivity based on Logistic Regression, eXtreme Gradient Boosting, and Random Forest modeling methods

Floods are one of the most destructive disasters that cause loss of life and property worldwide every year. In this study, the aim was to find the best-performing model in flood sensitivity assessment and analyze key characteristic factors, the spatial pattern of flood sensitivity was evaluated using three machine learning (ML) models: Logistic Regression (LR), eXtreme Gradient Boosting (XGBoost), and Random Forest (RF). Suqian City in Jiangsu Province was selected as the study area, and a random sample dataset of historical flood points was constructed. Fifteen different meteorological, hydrological, and geographical spatial variables were considered in the flood sensitivity assessment, 12 variables were selected based on the multi-collinearity study. Among the results of comparing the selected ML models, the RF method had the highest AUC value, accuracy, and comprehensive evaluation effect, and is a reliable and effective flood risk assessment model. As the main output of this study, the flood sensitivity map is divided into five categories, ranging from very low to very high sensitivity. Using the RF model (i.e., the highest accuracy of the model), the high-risk area covers about 44% of the study area, mainly concentrated in the central, eastern, and southern parts of the old city area.

Tool for fast assessment of stormwater flood volumes for urban catchment: A machine learning approach

Specific flood volume is an important criterion for evaluating the performance of sewer networks. Currently, mechanistic models - MCMs (e.g., SWMM) are usually used for its prediction, but they require the collection of detailed information about the characteristics of the catchment and sewer network, which can be difficult to obtain, and the process of model calibration is a complex task. This paper presents a methodology for developing simulators to predict specific flood volume using machine learning methods (DNN - Deep Neural Network, GAM - Generalized Additive Model). The results of Sobol index calculations using the GSA method were used to select the ML model as an alternative to the MCM model. It was shown that the DNN model can be used for flood prediction, for which high agreement was obtained between the results of GSA calculations for rainfall data, catchment and sewer network characteristics, and calibrated SWMM parameters describing land use and sewer retention. Regression relationships (polynomials and exponential functions) were determined between Sobol indices (retention depth of impervious area, correction factor of impervious area, Manning's roughness coefficient of sewers) and sewer network characteristics (unit density of sewers, retention factor - the downstream and upstream of retention ratio) obtaining R2 = 0. 55-0.78. The feasibility of predicting sewer network flooding and modernization with the DNN model using a limited range of input data compared to the SWMM was shown. The developed model can be applied to the management of urban catchments with limited access to data and at the stage of urban planning.

Forest fire surveillance systems: A review of deep learning methods

This review aims to critically examine the existing state-of-the-art forest fire detection systems that are based on deep learning methods. In general, forest fire incidences bring significant negative impact to the economy, environment, and society. One of the crucial mitigation actions that needs to be readied is an effective forest fire detection system that are able to automatically notify the relevant parties on the incidence of forest fire as early as possible. This review paper has examined in details 37 research articles that have implemented deep learning (DL) model for forest fire detection, which were published between January 2018 and 2023. In this paper, in depth analysis has been performed to identify the quantity and type of data that includes images and video datasets, as well as data augmentation methods and the deep model architecture. This paper is structured into five subsections, each of which focuses on a specific application of deep learning (DL) in the context of forest fire detection. These subsections include 1) classification, 2) detection, 3) detection and classification, 4) segmentation, and 5) segmentation and classification. To compare the model's performance, the methods were evaluated using comprehensive metrics like accuracy, mean average precision (mAP), F1-Score, mean pixel accuracy (MPA), etc. From the findings, of the usage of DL models for forest fire surveillance systems have yielded favourable outcomes, whereby the majority of studies managed to achieve accuracy rates that exceeds 90%. To further enhance the efficacy of these models, future research can explore the optimal fine-tuning of the hyper-parameters, integrate various satellite data, implement generative data augmentation techniques, and refine the DL model architecture. In conclusion, this paper highlights the potential of deep learning methods in enhancing forest fire detection that is crucial for forest fire management and mitigation.

Analyzing the impacts of topographic factors and land cover characteristics on waterlogging events in urban functional zones

Rapid urbanization and climate changes result in frequent occurrence of urban waterlogging disasters, which cause serious economic damage and pose a threat to residents' safety. Understanding the spatial characteristic and the key influencing factors of urban waterlogging has significant implications for mitigating waterlogging. In this study, the officially issued representative waterlogging points were obtained, as well as the topographic factors and land cover characteristics were selected to compare their impacts on the waterlogging event density in a highly urbanized area at urban functional zone (UFZ) scale, and to quantify the contributions of the key influencing factors on urban waterlogging events. Results showed the average density of urban waterlogging events in the study area is 9.2 points/km2, and 38.4 % of the waterlogging events are distributed in REZ. The distribution of waterlogging points in the study area revealed a significant multi-core and multilevel spatial aggregation pattern, and 12.1 % of the study area was high-density waterlogging area. In the total UFZs, the correlation coefficients of topographic indices with waterlogging density were relatively weaker than the other land cover characteristic metrics. The impervious surface ratio had significant contributions in all UFZ types. The larger ratio of impervious surface significantly increased the density of waterlogging events. The increase in the ratio of green space can effectively decrease the density of urban waterlogging. In the total UFZs, the top 3 key influencing factors of urban waterlogging were PR (35.9 %), COHESION (32.5 %) and DIVISION (11.8 %). The higher connectivity of landscape patches in REZ, INZ and COZ, as well as the increase of landscape dispersion or diversity in REZ, EGZ, INZ and GSZ can effectively reduce the occurrence of urban waterlogging. This study provides a better understanding of the formation mechanism of urban waterlogging disasters and potential implications for prioritized waterlogging mitigation strategies.

Assessment of urban flood susceptibility and role of urban green space (UGS) on flooding susceptibility using GIS-based probabilistic models

With rapid urbanization, the green space in urban areas is replaced with impervious built-up areas, which increases the frequency of urban floods. Kamrup Metropolitan District, Assam, is near the Brahmaputra and is highly prone to urban flooding. The present study aims to develop the urban flood susceptibility index (FSI) and to analyze the role of urban green space (UGS) as a nature-based solution (NBS) for urban flood susceptibility. Two types of flooded urban areas are observed using a two-stage cluster analysis. A GIS-based urban FSI is developed using logistic regression (LR), frequency ratio (FR), Shannon entropy (SE), certainty factor (CF), and weight of evidence (WoE) models, and variation of FSI is assessed for different UGS areas. According to the area under curve (AUC), the performance of all five models falls under the good to excellent class. The average UGS ratio for non-flooded is higher than for flooded areas, and with an increase in the area of UGS, the flooding probability decreases for all the models. The findings of the present study emphasize the importance of UGS and can be used for effective urban flood risk mitigation and management planning.

Control of climate and physiography on runoff response behavior through use of catchment classification and machine learning

Understanding of runoff response changes (RRC) is essential for water resource management decisions. However, there is a limited understanding of the effects of climate and landscape properties on RRC behavior. This study explored RRC behavior across controls and predictability in 1003 catchments in the contiguous United States (CONUS) using catchment classification and machine learning. Over 1000+ catchments are grouped into ten classes with similar hydrological behavior across CONUS. Indices quantifying RRC were constructed and then predicted within each class of the 10 classes and over the entire1000+ catchments using two machine learning models (random forest and CUBIST) based on 56 indicators of catchment attributes (CA) and 16 flow signatures (FS). This enabled the ranking of the important influential factors on RRC. We found that (i) CA/FS-based clusters followed the ecoregions over CONUS, and the impact of climate on RRC seemed to overlap with physiographic attributes; (ii) CUBIST outperforms the random forest model both within the cluster and over the whole domain, with a mean improvement of 39 % (depending on clusters) within clusters. Runoff sensitivity was better predicted than runoff changes; (iii) FS related to runoff ratio, average, and high flow are the most important for RRC, whereas climate (evaporation and aridity) is a secondary factor; and (iv) RRC patterns are substantial in the dominant factor space. High total changes and catchment characteristic-induced changes occurred mainly at 100°west longitude. The elasticity of climate and catchment characteristics was found to be high in spaces with high evaporation and low runoff ratios and low in spaces with low evaporation and high runoff ratios. Uncertainties existed in the number of catchments between clusters which was verified using a fuzzy clustering algorithm. We recommend that future research that clarifies the impact of uncertainty on hydrological or catchment behavior should be conducted.

Exploring effective ways to increase reliable positive samples for machine learning-based urban waterlogging susceptibility assessments

Machine learning (ML)-based urban waterlogging susceptibility studies suffer from class imbalance, as fewer positive samples are generally available than potential negative samples. Few studies have considered optimizing the results by improving the quality of training samples. To address this issue, we explored effective approaches to reliably increase the numbers of positive samples for such studies. The Synthetic Minority Over-Sampling Technique (SMOTE) and Optimized Seed Spread Algorithm (OSSA), representative of oversampling (synthesizing new samples based on the feature space) and physical (simulating potential inundated area based on the mechanisms of water flow) approaches, respectively, were employed to increase the number of positive samples. Waterlogging in Shenzhen was selected as a case study using eight selected spatial variables. An elaborate experiment was conducted to compare the quality of added samples based on the classifiers' performance and accuracy of waterlogging susceptibility maps (WSMs). The results indicated that (1) the performance of classifiers generated with SMOTE was worse than the original samples, while the use of OSSA improved the trained classifiers, and (2) the accuracy of WSMs was not improved with SMOTE but increased markedly with OSSA. These results may be driven by the diversity of information and features of the added samples. This study indicates the use of SMOTE fails to synthesize reliable samples when applied to waterlogging analysis in Shenzhen, whereas an effective solution for generating reliable positive samples is to use OSSA that simulates the potential submerged regions based on the mechanisms of disaster occurrence and spread.

System comprehensive risk assessment of urban rainstorm-induced flood-water pollution disasters

The urban rainstorm-induced flood-water pollution disaster is a kind of systematic risk, which may induce secondary disasters that can lead to more serious damage, so this paper first adopts the fuzzy comprehensive evaluation method to determine the flood risk by combining with the submergence depth derived from the risk field and other factors data, and then the grid environmental risk evaluation method, which is improved by increasing the induced possibility based on Bayesian theory, is used to evaluate the flood-induced water pollution risk, and the system comprehensive risk of rainstorm-induced flood-water pollution disasters is finally obtained by constructing risk level matrix, which can well depict the coupling superposition effect. Shenzhen City is selected as the study area, and the results showed that the area with high-risk of both flood and water pollution only accounts for about 0.14% of the total area, mainly distributed in the eastern junction of Longgang district and Pingshan district, where the rainstorms occur frequently and the enterprise risk sources are dense. The system comprehensive risk is mostly very low-low and very high-low, accounting for more than 76% of the total area. It is always necessary to pay attention not only to the areas with high risk level of both disasters, but also to the areas with high risk level of one disaster. The method proposed in this study can not only quantitatively reveal the formation of the induced risk, but also provide reference for early warning.

Flood sensitivity assessment of super cities

In the context of global urbanization, more and more people are attracted to these cities with superior geographical conditions and strategic positions, resulting in the emergence of world super cities. However, with the increasing of urban development, the underlying surface of the city has changed, the soil originally covered with vegetation has been substituted by hardened pavement such as asphalt and cement roads. Therefore, the infiltration capacity of urban rainwater is greatly limited, and waterlogging is becoming more and more serious. In addition, the suburbs of the main urban areas of super cities are usually villages and mountains, and frequent flash floods seriously threaten the life and property safety of people in there. Flood sensitivity assessment is an effective method to predict and mitigate flood disasters. Accordingly, this study aimed at identifying the areas vulnerable to flood by using Geographic Information System (GIS) and Remote Sensing (RS) and apply Logistic Regression (LR) model to create a flood sensitivity map of Beijing. 260 flood points in history and 12 predictors [elevation, slope, aspect, distance to rivers, Topographic Wetness Index (TWI), Stream Power Index (SPI), Sediment Transport Index (STI), curvature, plan curvature, Land Use/Land Cover (LULC), soil, and rainfall] were used in this study. Even more noteworthy is that most of the previous studies discussed flash flood and waterlogging separately. However, flash flood points and waterlogging points were included together in this study. We evaluated the sensitivity of flash flood and waterlogging as a whole and obtained different results from previous studies. In addition, most of the previous studies focused on a certain river basin or small towns as the study area. Beijing is the world's ninth largest super cities, which was unusual in previous studies and has important reference significance for the flood sensitivity analysis of other super cities. The flood inventory data were randomly subdivided into training (70%) and test (30%) sets for model construction and testing using the Area Under Curve (AUC), respectively. The results turn out that: (1) elevation, slope, rainfall, LULC, soil and TWI were highly important among these elements, and were the most influential variables in the assessment of flood sensitivity. (2) The AUC of the test dataset revealed a prediction rate of 81.0%. The AUC was greater than 0.8, indicating that the model assessment accuracy was high. (3) The proportion of high risk and extremely high risk areas was 27.44%, including 69.26% of the flood events in this study, indicating that the flood distribution in these areas was relatively dense and the susceptibility was high. Super cities have a high population density, and once flood disasters occur, the losses brought by them are immeasurable. Thus, flood sensitivity map can provide meaningful information for policy makers to enact appropriate policies to reduce future damage.

Urban flood risk differentiation under land use scenario simulation

The frequent urban floods have seriously affected the regional sustainable development in recent years. It is significant to understand the characteristics of urban flood risk and reasonably predict urban flood risk under different land use scenarios. This study used the random forest and multi-criteria decision analysis models to assess the spatiotemporal characteristics of flood risk in Zhengzhou City, China, from 2005 to 2020, and proposed a robust method coupling Bayesian network and patch-generating land use simulation models to predict future flood risk probability. We found that the flood risk in Zhengzhou City presented an upward trend from 2005 to 2020, and its spatial pattern was "high in the middle and low in the surrounding areas". In addition, land use patterns under the sustainable development scenario would be more conducive to reducing flood risk. Our results can provide theoretical support for scientifically optimizing land use to improve urban flood risk management.

Resilient landscape pattern for reducing coastal flood susceptibility

Evaluating flood susceptibility, identifying flood-prone areas, and planning reasonable landscape patterns are important measures in promoting sustainable urban development and flood mitigation. To this end, this study evaluated the flood susceptibility using a neural network model depending on a flood inundation map created from satellite data from 2010 to 2020, and explanatory factors for flood inundation selected by Geodetector and regularized random forest. Subsequently, the landscape pattern of the coastal city was quantified based on the land cover, and key landscape pattern metrics for flood susceptibility were selected at patch and class levels using statistical approaches. Eventually, urban spatial planning strategies for flood management were proposed based on the ecological significance of key metrics. Taking Xiamen as a case study, the flood susceptibility map showed that flood-prone areas in Xiamen are mainly distributed along river banks and coastlines. Key landscape pattern metrics for flood susceptibility selected by statistical approaches showed that patch-level metrics account for more explanatory power than class-level metrics, and the classes of the landscape would affect the role of patch-level metrics. Overall, the division index of the forest, the connectance index of water, the number of core areas and the fractal dimension index of urban, and the Euclidean nearest-neighbor distance of urban and water are significantly positively related to flood susceptibility, while the core area and the proximity index of urban, the similarity index, the core area index, and the edge contrast index of the forest, and the contiguity index of forest, grass, farmland, and shrub negatively related with flood susceptibility. Based on these findings, intensive urban planning and integrative Nature-based Solutions networks should be considered as strategies for enhancing coastal flood resilience.

A novel framework for addressing uncertainties in machine learning-based geospatial approaches for flood prediction

Globally, many studies on machine learning (ML)-based flood susceptibility modeling have been carried out in recent years. While majority of those models produce reasonably accurate flood predictions, the outcomes are subject to uncertainty since flood susceptibility models (FSMs) may produce varying spatial predictions. However, there have not been many attempts to address these uncertainties because identifying spatial agreement in flood projections is a complex process. This study presents a framework for reducing spatial disagreement among four standalone and hybridized ML-based FSMs: random forest (RF), k-nearest neighbor (KNN), multilayer perceptron (MLP), and hybridized genetic algorithm-gaussian radial basis function-support vector regression (GA-RBF-SVR). Besides, an optimized model was developed combining the outcomes of those four models. The southwest coastal region of Bangladesh was selected as the case area. A comparable percentage of flood potential area (approximately 60% of the total land areas) was produced by all ML-based models. Despite achieving high prediction accuracy, spatial discrepancy in the model outcomes was observed, with pixel-wise correlation coefficients across different models ranging from 0.62 to 0.91. The optimized model exhibited high prediction accuracy and improved spatial agreement by reducing the number of classification errors. The framework presented in this study might aid in the formulation of risk-based development plans and enhancement of current early warning systems.

Modelling flood susceptibility based on deep learning coupling with ensemble learning models

Modelling flood susceptibility is an indirect way to reduce the loss from flood disaster. Now, flood susceptibility modelling based on data driven model is state-of-the-art method such as ensemble learning and deep learning. However, the effect of deep learning coupling with ensemble learning models in flood susceptibility modelling is still unknown. Therefore, the aim of this paper is to propose three deep learning coupling with ensemble learning models by combining the deep learning (DL) with Filtered Classifier (FC), Rotation forest (RF) and Random Subspace (RSS) and explore the effect of coupling method for modelling flood susceptibility. The key step of this paper is as following: firstly, a Dingnan County which is lied in the Jiangxi Province of China is chosen as a case study, single flood event point and random sampling method was applied to generate the flood and non-flood data, respectively, then frequency ratio was utilized to analyze the relationship between each influencing factor and flood occurrence, based on the value of VIF, Spearman's correlation and One R classifier, the result show that there is no multicollinearity between each influencing factor, ten influencing factors have contribution to the flood occurrence and all of them are applied to construct the coupling model. Finally, the DL, FC-DL, RF-DL and RSS-DL were applied to produce flood susceptibility maps. Then, several statistical indexes such as area under the curve (AUC), Kappa index, accuracy (ACC), and F-measure were used to assess the accomplishment of these coupling models. For the train data, the FC-DL model acquired the highest AUC value (0.996), followed by RF-DL (0.944), RSS-DL (0.934), and DL (0.934). For the validation data, the result showed that all models have a good accomplishment (AUC>0.8). In a word, the deep learning coupling with ensemble learning models demonstrates the more reliable and excellent performance. Hence, the proposed new method will help the government for land use planning and can be applied in other area around the world.

Assessing the urban road waterlogging risk to propose relative mitigation measures

Road waterlogging has become a significant issue in developed cities due to the rapid urbanization in China. It is necessary to accurately identify the risk of waterlogging in urban roads and propose appropriate mitigation measures. This study considered urban waterlogging as a landscape ecological process. The road waterlogging risk was simulated and estimated using the Minimum Cumulative Resistance model under natural drainage conditions. The results indicate that: 1) The Minimum Cumulative Resistance model effectively assesses the waterlogging risk for each road segment. The roads in and around the central city have relatively higher waterlogging risks. The overall length of high-risk roads is 918.7 km, accounting for 31.3 % of the total. 2) There are 448 potential runoff paths and 448 inflow sites. The city's center and its north and south sides are the primary locations of the high-risk runoff paths and the inflow sites. 3) Road waterlogging is significantly more affected by the land-use types of High density residential and Industrial under rainfall intensities of a-year, 2-year, 3-year, and 5-year return periods. And the effects of various land-use types on waterlogging vary with the rainfall intensity. Using landscape ecology theory to analyze the risk of road waterlogging is a novel method to address urban waterlogging issues. This approach provides a more accurate approach to identifying the urban waterlogging risks and can be applied to developed cities suffering from waterlogging to help decision-makers devise the most effective mitigation measures.

Spatio-temporal assessment of land use land cover based on trajectories and cellular automata Markov modelling and its impact on land surface temperature of Lahore district Pakistan

This research aims to assess the urban growth and impact on land surface temperature (LST) of Lahore, the second biggest city in Pakistan. In this research, various geographical information system (GIS) and remote sensing (RS) techniques (maximum likelihood classification (MLC)) LST, and different normalized satellite indices have been implemented to analyse the spatio-temporal trends of Lahore city; by using Landsat for 1990, 2004, and 2018. The development of integrated use of RS and GIS and combined cellular automata-Markov models has provided new means of assessing changes in land use and land cover and has enabled the projection of trajectories into the future. Results indicate that the built-up area and bare area increased from 15,541 (27%) to 23,024 km² (40%) and 5756 km² (10%) to 13,814 km² (24%). Meanwhile, water area and vegetation were decreased from 2302 km² (4%) to 1151 km² (2%) and 33,961 km² (59%) to 19,571 km² (34%) respectively. Under this urbanization, the LST of the city was also got affected. In 1990, the mean LST of most of the area was between 14 and 28 ℃, which rose to 22-28 ℃ in 2004 and 34 to 36 ℃ in 2018. Because of the shift of vegetation and built-up land, the surface reflectance and roughness of each land use land cover (LULC) class are different. The analysis established a direct correlation between Normalized Difference Water Index (NDWI), Normalized Difference Vegetation Index (NDVI) with LST and an indirect correlation among Soil Adjusted Vegetation Index (SAVI), Normalized Difference Built-up Index (NDBI), and Built-up Index (BI) with LST. The results are important for the planning and development department since they may be used to guarantee the sustainable utilization of land resources for future urbanization expansion projects.

Spatial Prediction of Current and Future Flood Susceptibility: Examining the Implications of Changing Climates on Flood Susceptibility Using Machine Learning Models

The main aim of this study was to predict current and future flood susceptibility under three climate change scenarios of RCP2.6 (i.e., optimistic), RCP4.5 (i.e., business as usual), and RCP8.5 (i.e., pessimistic) employing four machine learning models, including Gradient Boosting Machine (GBM), Random Forest (RF), Multilayer Perceptron Neural Network (MLP-NN), and Naïve Bayes (NB). The study was conducted for two watersheds in Canada, namely Lower Nicola River, BC and Loup, QC. Three statistical metrics were used to validate the models: Receiver Operating Characteristic Curve, Figure of Merit, and F1-score. Findings indicated that the RF model had the highest accuracy in providing the flood susceptibility maps (FSMs). Moreover, the provided FSMs indicated that flooding is more likely to occur in the Lower Nicola River watershed than the Loup watershed. Following the RCP4.5 scenario, the area percentages of the flood susceptibility classes in the Loup watershed in 2050 and 2080 have changed by the following percentages from the year 2020 and 2050, respectively: Very Low = -1.68%, Low = -5.82%, Moderate = +6.19%, High = +0.71%, and Very High = +0.6% and Very Low = -1.61%, Low = +2.98%, Moderate = -3.49%, High = +1.29%, and Very High = +0.83%. Likewise, in the Lower Nicola River watershed, the changes between the years 2020 and 2050 and between the years 2050 and 2080 were: Very Low = -0.38%, Low = -0.81%, Moderate = -0.95%, High = +1.72%, and Very High = +0.42% and Very Low = -1.31%, Low = -1.35%, Moderate = -1.81%, High = +2.37%, and Very High = +2.1%, respectively. The impact of climate changes on future flood-prone places revealed that the regions designated as highly and very highly susceptible to flooding, grow in the forecasts for both watersheds. The main contribution of this study lies in the novel insights it provides concerning the flood susceptibility of watersheds in British Columbia and Quebec over time and under various climate change scenarios.

Hybrid-based Bayesian algorithm and hydrologic indices for flash flood vulnerability assessment in coastal regions: machine learning, risk prediction, and environmental impact

Natural hazards and severe weather events are a matter of serious threat to humans, economic activities, and the environment. Flash floods are one of the extremely devastating natural events around the world. Consequently, the prediction and precise assessment of flash flood-prone areas are mandatory for any flood mitigation strategy. In this study, a new hybrid approach of machine learning (ML) algorithm and hydrologic indices opted to detect impacted and highly vulnerable areas. The obtained models were trained and validated using a total of 189 locations from Wadi Ghoweiba and surrounding area (case study). Various controlling factors including varied datasets such as stream transport index (STI), stream power index (SPI), lithological units, topographic wetness index (TWI), slope angle, stream density (SD), curvature, and slope aspect (SA) were utilized via hyper-parameter optimization setting to enhance the performance of the proposed model prediction. The hybrid machine learning (HML) model, developed by combining naïve Bayes (NïB) approach and hydrologic indices, was successfully implemented and utilized to investigate flash flood risk, sediment accumulation, and erosion predictions in the studied site. The synthesized new hybrid model demonstrated a model accuracy of 90.8% compared to 87.7% of NïB model, confirming the superior performance of the obtained model. Furthermore, the proposed model can be successfully employed in large-scale prediction applications.

An Integrated Approach for Urban Pluvial Flood Risk Assessment at Catchment Level

With the rapid development of urbanization and global climate change, urban pluvial floods have occurred more frequently in urban areas. Despite of the increasing urban pluvial flood risk, there is still a lack of comprehensive understanding of the physical and social influencing factors on the process. To fill this knowledge gap, this paper proposes a novel approach to calculate the comprehensive urban pluvial flooding risk index (PFRI) and investigates the interplay impacts from different components at catchment level. To be more specific, PFRI is determined by two components, Exposure Index (EI) and Social Vulnerability Index (SoVI). EI is evaluated based on two indicators, the depression-based Topographic Control Index (TCI) and impervious area ratio. SoVI is measured based on a set of demographic and socio-economic indicators. Our results demonstrated the spatial heterogeneity of urban pluvial flood exposure and social vulnerability, as well as the composite flooding risk across the study area. Our catchment-based urban pluvial flooding risk assessment method can provide a comprehensive understanding of urban flooding and promote the formulation of effective flood mitigation strategies from the catchment perspective.

Machine Learning and Remote Sensing Application for Extreme Climate Evaluation: Example of Flood Susceptibility in the Hue Province, Central Vietnam Region

Floods are the most frequent natural hazard globally and incidences have been increasing in recent years as a result of human activity and global warming, making significant impacts on people’s livelihoods and wider socio-economic activities. In terms of the management of the environment and water resources, precise identification is required of areas susceptible to flooding to support planners in implementing effective prevention strategies. The objective of this study is to develop a novel hybrid approach based on Bald Eagle Search (BES), Support Vector Machine (SVM), Random Forest (RF), Bagging (BA) and Multi-Layer Perceptron (MLP) to generate a flood susceptibility map in Thua Thien Hue province, Vietnam. In total, 1621 flood points and 14 predictor variables were used in this study. These data were divided into 60% for model training, 20% for model validation and 20% for testing. In addition, various statistical indices were used to evaluate the performance of the model, such as Root Mean Square Error (RMSE), Receiver Operation Characteristics (ROC), and Mean Absolute Error (MAE). The results show that BES, for the first time, successfully improved the performance of individual models in building a flood susceptibility map in Thua Thien Hue, Vietnam, namely SVM, RF, BA and MLP, with high accuracy (AUC > 0.9). Among the models proposed, BA-BES was most effective with AUC = 0.998, followed by RF-BES (AUC = 0.998), MLP-BES (AUC = 0.998), and SVM-BES (AUC = 0.99). The findings of this research can support the decisions of local and regional authorities in Vietnam and other countries regarding the construction of appropriate strategies to reduce damage to property and human life, particularly in the context of climate change.

Climatic and Hydrological Factors Affecting the Assessment of Flood Hazards and Resilience Using Modified UNDRR Indicators: Ayutthaya, Thailand

This research aims to investigate the effect of climatic and hydrological factors on flood hazards and assess flood resilience in Ayutthaya, Thailand, using the 10 essentials for making cities resilient modified by the United Nations Office for Disaster Risk Reduction (UNDRR). Flood resilience assessment was performed based on a multi-criteria decision-making approach or the analytical hierarchy process (AHP) of pairwise comparison. The results indicate that runoff is considered the most influential factor in flood hazards, followed by land use, rainfall, and historical flood events, sequentially. Regarding the flood incident management concept, a questionnaire survey (n = 552) was conducted to understand the impacts of flood on local communities. The findings reveal that 50% of respondents had never received any flood information or participated in training sessions on flood preparedness. Most reported their concerns about the inadequate supply of drinking water during a flood. Spearman’s correlation coefficient shows positive correlations between flood disaster relief payments, preparedness training, access to flood hazard mapping, emergency health services, and their flood preparation actions. According to the modified UNDRR indicators, the top three highest AHP values in building community resilience to flood hazards in Ayutthaya are flood risk scenario identification, the effectiveness of emergency flood disaster response, integrated urban planning, and disaster risk reduction. The policy implications of this research include the need for national authorities to better understand the role cities can play a vital role in supporting both national and international climate resilience frameworks, especially Thailand’s National Disaster Management Plan, the Sendai Framework for Disaster Risk Reduction (SFDRR), and the global Sustainable Development Goals (SDGs).

Application of GIS and Machine Learning to Predict Flood Areas in Nigeria

Floods are one of the most devastating forces in nature. Several approaches for identifying flood-prone locations have been developed to reduce the overall harmful impacts on humans and the environment. However, due to the increased frequency of flooding and related disasters, coupled with the continuous changes in natural and social-economic conditions, it has become vital to predict areas with the highest probability of flooding to ensure effective measures to mitigate impending disasters. This study predicted the flood susceptible areas in Nigeria based on historical flood records from 1985~2020 and various conditioning factors. To evaluate the link between flood incidence and the fifteen (15) explanatory variables, which include climatic, topographic, land use and proximity information, the artificial neural network (ANN) and logistic regression (LR) models were trained and tested to develop a flood susceptibility map. The receiver operating characteristic curve (ROC) and area under the curve (AUC) were used to evaluate both model accuracies. The results show that both techniques can model and predict flood-prone areas. However, the ANN model produced a higher performance and prediction rate than the LR model, 76.4% and 62.5%, respectively. In addition, both models highlighted that those areas with the highest susceptibility to flood are the low-lying regions in the southern extremities and around water areas. From the study, we can establish that machine learning techniques can effectively map and predict flood-prone areas and serve as a tool for developing flood mitigation policies and plans.

Urban Flood-Risk Assessment: Integration of Decision-Making and Machine Learning

Urban flood-risk mapping is an important tool for the mitigation of flooding in view of continuing urbanization and climate change. However, many developing countries lack sufficiently detailed data to produce reliable risk maps with existing methods. Thus, improved methods are needed that can help managers and decision makers to combine existing data with more soft semi-subjective data, such as citizen observations of flood-prone and vulnerable areas in view of existing settlements. Thus, we present an innovative approach using the semi-subjective Analytic Hierarchy Process (AHP), which integrates both subjective and objective assessments, to help organize the problem framework. This approach involves measuring the consistency of decision makers’ judgments, generating pairwise comparisons for choosing a solution, and considering criteria and sub-criteria to evaluate possible options. An urban flood-risk map was created according to the vulnerabilities and hazards of different urban areas using classification and regression-tree models, and the map can serve both as a first stage in advancing flood-risk mitigation approaches and in allocating warning and forecasting systems. The findings show that machine-learning methods are efficient in urban flood zoning. Using the city Rasht in Iran, it is shown that distance to rivers, urban drainage density, and distance to vulnerable areas are the most significant parameters that influence flood hazards. Similarly, for urban flood vulnerability, population density, land use, dwelling quality, household income, distance to cultural heritage, and distance to medical centers and hospitals are the most important factors. The integrated technique for both objective and semi-subjective data as outlined in the present study shows credible results that can be obtained without complicated modeling and costly field surveys. The proposed method is especially helpful in areas with little data to describe and display flood hazards to managers and decision makers.

Machine Learning for Defining the Probability of Sentinel-1 Based Deformation Trend Changes Occurrence

The continuous monitoring of displacements occurring on the Earth surface by exploiting MTInSAR (Multi Temporal Interferometry SAR) Sentinel-1 data is a solid reality, as testified by the ongoing operational ground motion service in the Tuscany region (Central Italy). In this framework, anomalies of movement, i.e., accelerations or deceleration as seen by the time series of displacement of radar targets, are identified. In this work, a Machine Learning algorithm such as the Random Forest has been used to assess the probability of occurrence of the anomalies induced by slope instability and subsidence. About 20,000 anomalies (about 7000 and 13,000 for the slope instability and the subsidence, respectively) were collected between 2018 and 2020 and were used as input, while ten different variables were selected, five related to the morphological and geological setting of the study area and five to the radar characteristics of the data. The resulting maps may provide useful indications of where a sudden change of displacement trend may occur, analyzing the contribution of each factor. The cross-validation with the anomalies collected in a following timespan (2020–2021) and with official landslide and subsidence inventories provided by the regional authority has confirmed the reliability of the final maps. The adoption of a map for assessing the probability of the occurrence of MTInSAR anomalies may serve as an enhanced geohazard prevention measurement, to be periodically updated and refined in order to have the most precise knowledge possible of the territory.

Flood Risk Mapping by Remote Sensing Data and Random Forest Technique

Detecting effective parameters in flood occurrence is one of the most important issues that has drawn more attention in recent years. Remote Sensing (RS) and Geographical Information System (GIS) are two efficient ways to spatially predict Flood Risk Mapping (FRM). In this study, a web-based platform called the Google Earth Engine (GEE) (Google Company, Mountain View, CA, USA) was used to obtain flood risk indices for the Galikesh River basin, Northern Iran. With the aid of Landsat 8 satellite imagery and the Shuttle Radar Topography Mission (SRTM) Digital Elevation Model (DEM), 11 risk indices (Elevation (El), Slope (Sl), Slope Aspect (SA), Land Use (LU), Normalized Difference Vegetation Index (NDVI), Normalized Difference Water Index (NDWI), Topographic Wetness Index (TWI), River Distance (RD), Waterway and River Density (WRD), Soil Texture (ST]), and Maximum One-Day Precipitation (M1DP)) were provided. In the next step, all of these indices were imported into ArcMap 10.8 (Esri, West Redlands, CA, USA) software for index normalization and to better visualize the graphical output. Afterward, an intelligent learning machine (Random Forest (RF)), which is a robust data mining technique, was used to compute the importance degree of each index and to obtain the flood hazard map. According to the results, the indices of WRD, RD, M1DP, and El accounted for about 68.27 percent of the total flood risk. Among these indices, the WRD index containing about 23.8 percent of the total risk has the greatest impact on floods. According to FRM mapping, about 21 and 18 percent of the total areas stood at the higher and highest risk areas, respectively.

Ensemble Learning Technology for Coastal Flood Forecasting in Internet-of-Things-Enabled Smart City

Flooding is becoming a prominent issue in coastal cities, flood forecasting is the key to solving this problem. However, the lack and imbalance of research data and the insufficient performance of the model have led to the complexity and uncontrollability of flood forecasting. To forecast coastal floods accurately and reliably, the Internet of Things technology is used to collect data on floods and flood factors in smart cities. An ensemble learning method based on Bayesian model combination (BMC-EL) is designed to predict flood depth. First, flood intensity classification and K-fold cross-validation are introduced to generate multiple training subsets from the training set to realize uniform sampling and increase the diversity of subsets. Second, the backpropagation neural network (BPNN) and random forest (RF) are used as the base learners to build the prediction model and then import it into training subsets for training purposes. Finally, based on the prediction performance of the base learner in the validation sets, the Bayesian model combination strategy is formulated to integrate and output predicted values. We describe experiments conducted to forecast flood depth 1 h in advance that several machine learning models were trained and tested using real flood data taken from Macao, China. The models include linear regression, support vector machine, BPNN, RF and BMC-EL models. Results prove the accuracy and reliability of the BMC-EL method in flood forecasting for coastal cities.

Towards better flood risk management: Assessing flood risk and investigating the potential mechanism based on machine learning models

Integrating powerful machine learning models with flood risk assessment and determining the potential mechanism between risk and the driving factors are crucial for improving flood management. In this study, six machine learning models were utilized for flood risk assessment of the Pearl River Delta, in which the Gradient Boosting Decision Tree (GBDT), eXtreme Gradient Boosting (XGBoost), and Convolutional Neural Network (CNN) models were firstly applied in this field. Twelve indices were chosen and 2000 sample points were created for model training and testing. Hyperparameter optimization of the models was conducted to ensure fair comparisons. Due to uncertainty in the sample dataset, recorded inundation hot-spots were utilized to validate the rationality of the flood risk zoning maps. After determining the optimal model, the driving factors of different flood risk levels were investigated. Urban and rural areas and coastal and inland areas were also compared to determine the flood risk mechanism in different highest-risk areas. The results showed that the GBDT performed best and provided the most reasonable flood risk result among the six models. A comparison of the driving factors at different risk levels indicated that the disaster-inducing factor, disaster-breeding environment, and disaster-bearing body were not definitely becoming more serious as the flood risk increased. In the highest-risk areas, rural areas were featured by worse disaster-breeding environment than urban areas, and the disaster-inducing factors of coastal areas were more serious than those of inland areas. Moreover, the Digital Elevation Model (DEM), maximum 1-day precipitation (M1DP), and road density (RD) were the top three significant driving factors and contributed 52% to flood risk. This study not only expands the application of machine learning and deep learning methods for flood risk assessment, but also deepens our understanding of the potential mechanism of flood risk and provides insights into better flood risk management.

Investigating the Role of Green Infrastructure on Urban WaterLogging: Evidence from Metropolitan Coastal Cities

Urban green infrastructures (UGI) can effectively reduce surface runoff, thereby alleviating the pressure of urban waterlogging. Due to the shortage of land resources in metropolitan areas, it is necessary to understand how to utilize the limited UGI area to maximize the waterlogging mitigation function. Less attention, however, has been paid to investigating the threshold level of waterlogging mitigation capacity. Additionally, various studies mainly focused on the individual effects of UGI factors on waterlogging but neglected the interactive effects between these factors. To overcome this limitation, two waterlogging high-risk coastal cities—Guangzhou and Shenzhen, are selected to examine the effectiveness and stability of UGI in alleviating urban waterlogging. The results indicate that the impact of green infrastructure on urban waterlogging largely depends on its area and biophysical parameter. Healthier or denser vegetation (superior ecological environment) can more effectively intercept and store rainwater runoff. This suggests that while increasing the area of UGI, more attention should be paid to the biophysical parameter of vegetation. Hence, the mitigation effect of green infrastructure would be improved from the “size” and “health”. The interaction of composition and spatial configuration greatly enhances their individual effects on waterlogging. This result underscores the importance of the interactive enhancement effect between UGI composition and spatial configuration. Therefore, it is particularly important to optimize the UGI composition and spatial pattern under limited land resource conditions. Lastly, the effect of green infrastructure on waterlogging presents a threshold phenomenon. The excessive area proportions of UGI within the watershed unit or an oversized UGI patch may lead to a waste of its mitigation effect. Therefore, the area proportion of UGI and its mitigation effect should be considered comprehensively when planning UGI. It is recommended to control the proportion of green infrastructure at the watershed scale (24.4% and 72.1% for Guangzhou and Shenzhen) as well as the area of green infrastructure patches (1.9 ha and 2.8 ha for Guangzhou and Shenzhen) within the threshold level to maximize its mitigation effect. Given the growing concerns of global warming and continued rapid urbanization, these findings provide practical urban waterlogging prevention strategies toward practical implementations.

Flash Flood Susceptibility Assessment and Zonation Using an Integrating Analytic Hierarchy Process and Frequency Ratio Model for the Chitral District, Khyber Pakhtunkhwa, Pakistan

Pakistan is a flood-prone country and almost every year, it is hit by floods of varying magnitudes. This study was conducted to generate a flash flood map using analytical hierarchy process (AHP) and frequency ratio (FR) models in the ArcGIS 10.6 environment. Eight flash-flood-causing physical parameters were considered for this study. Five parameters were based on the digital elevation model (DEM), Advanced Land Observation Satellite (ALOS), and Sentinel-2 satellite, including distance from the river and drainage density slope, elevation, and land cover, respectively. Two other parameters were geology and soil, consisting of different rock and soil formations, respectively, where both layers were classified based on their resistance against water percolation. One parameter was rainfall. Rainfall observation data obtained from five meteorological stations exist close to the Chitral District, Pakistan. According to its significant importance in the occurrence of a flash flood, each criterion was allotted an estimated weight with the help of AHP and FR. In the end, all the parameters were integrated using weighted overlay analysis in which the influence value of the drainage density was given the highest value. This gave the output in terms of five flood risk zones: very high risk, high risk, moderate risk, low risk, and very low risk. According to the results, 1168 km2, that is, 8% of the total area, showed a very high risk of flood occurrence. Reshun, Mastuj, Booni, Colony, and some other villages were identified as high-risk zones of the study area, which have been drastically damaged many times by flash floods. This study is pioneering in its field and provides policy guidelines for risk managers, emergency and disaster response services, urban and infrastructure planners, hydrologists, and climate scientists.

Wildland Fire Susceptibility Mapping Using Support Vector Regression and Adaptive Neuro-Fuzzy Inference System-Based Whale Optimization Algorithm and Simulated Annealing

Fires are one of the most destructive forces in natural ecosystems. This study aims to develop and compare four hybrid models using two well-known machine learning models, support vector regression (SVR) and the adaptive neuro-fuzzy inference system (ANFIS), as well as two meta-heuristic models, the whale optimization algorithm (WOA) and simulated annealing (SA) to map wildland fires in Jerash Province, Jordan. For modeling, 109 fire locations were used along with 14 relevant factors, including elevation, slope, aspect, land use, normalized difference vegetation index (NDVI), rainfall, temperature, wind speed, solar radiation, soil texture, topographic wetness index (TWI), distance to drainage, and population density, as the variables affecting the fire occurrence. The area under the receiver operating characteristic (AUROC) was used to evaluate the accuracy of the models. The findings indicated that SVR-based hybrid models yielded a higher AUROC value (0.965 and 0.949) than the ANFIS-based hybrid models (0.904 and 0.894, respectively). Wildland fire susceptibility maps can play a major role in shaping firefighting tactics.

Explicit the urban waterlogging spatial variation and its driving factors: The stepwise cluster analysis model and hierarchical partitioning analysis approach

Urban waterlogging is a hydrological cycle problem that seriously affects people's life and property. Characterizing waterlogging variation and explicit its driving factors are conducive to prevent the damage of such disasters. Conventional methods, because of the high spatial heterogeneity and the non-stationary complex mechanism of urban waterlogging, are not able to fully capture the urban waterlogging spatial variation and identify the waterlogging susceptibility areas. A more robust method is recommended to quantify the variation trend of urban waterlogging. Previous studies have simulated the waterlogging variation in relatively small areas. However, the relationship between variables is often ignored, which cannot comprehensively reveal the dominant drivers affecting urban waterlogging. Therefore, a novel approach is proposed that combined stepwise cluster analysis model (SCAM) and hierarchical partitioning analysis (HPA) within a general framework and verifies the applicability through logistic regression, artificial neural network, and support vector machine. According to the dominant driving factors, different simulation scenarios are established to analyze waterlogging density variation. Results found that the SCAM provides accurate and detailed simulated results both in urban centers where waterlogging frequently occurs and urban fringe with few waterlogging events, which shows an excellent performance with a high classification accuracy and generalization capability. HPA detected that the impervious surface abundance (28.07%), vegetation abundance (20.80%), and cumulate precipitation (16.25%) are the dominant drivers of waterlogging. This result suggests that priority should be given to controlling these three factors to mitigate the risk of waterlogging. It is interesting to note that under different urbanization and rainfall scenarios, the urban waterlogging susceptibility has a considerable variation. The watershed spatial location and watershed characteristics are relevant aspects to be considered in identifying and assessing waterlogging susceptibility, which provides original insights that urban waterlogging mitigation strategies should be developed according to different local conditions and future scenarios.

Integrated Framework for Detecting the Areas Prone to Flooding Generated by Flash-Floods in Small River Catchments

In the present study, the susceptibility to flash-floods and flooding was studied across the Izvorul Dorului River basin in Romania. In the first phase, three ensemble models were used to determine the susceptibility to flash-floods. These models were generated by a combination of three statistical bivariate methods, namely frequency ratio (FR), weights of evidence (WOE), and statistical index (SI), with fuzzy analytical hierarchy process (FAHP). The result obtained from the application of the FAHP-WOE model had the best performance highlighted by an Area Under Curve—Receiver Operating Characteristics Curve (AUC-ROC) value of 0.837 for the training sample and another of 0.79 for the validation sample. Furthermore, the results offered by FAHP-WOE were weighted on the river network level using the flow accumulation method, through which the valleys with a medium, high, and very high torrential susceptibility were identified. Based on these valleys’ locations, the susceptibility to floods was estimated. Thus, in the first stage, a buffer zone of 200 m was delimited around the identified valleys along which the floods could occur. Once the buffer zone was established, ten flood conditioning factors were used to determine the flood susceptibility through the analytical hierarchy process model. Approximately 25% of the total delimited area had a high and very high flood susceptibility.

Key Disaster-Causing Factors Chains on Urban Flood Risk Based on Bayesian Network

Drivers of urban flood disaster risk may be related to many factors from nature and society. However, it is unclear how these factors affect each other and how they ultimately affect the risk. From the perspective of risk uncertainty, flood inundation risk is considered to be the probability of inundation consequences under the influence of various factors. In this paper, urban flood inundation risk assessment model is established based on Bayesian network, and then key disaster-causing factors chains are explored through influence strength analysis. Jingdezhen City is selected as study area, where the flood inundation probability is calculated, and the paths of these influential factors are found. The results show that the probability of inundation in most areas is low. Risk greater than 0.8 account for about 9%, and most of these areas are located in the middle and southern section of the city. The influencing factors interact with each other in the form of factor chain and, finally, affect the flood inundation. Rainfall directly affects inundation, while river is the key factor on inundation which is influenced by elevation and slope. In addition, in the chain of socio-economic factors, the population will determine the pipe density through affecting gross domestic product (GDP), and lead to the inundation. The approach proposed in this study can be used to find key disaster-causing factors chains, which not only quantitatively reveal the formation of risks but also provide reference for early warning.

Using Multi-Factor Analysis to Predict Urban Flood Depth Based on Naive Bayes

With global warming, the number of extreme weather events will increase. This scenario, combined with accelerating urbanization, increases the likelihood of urban flooding. Therefore, it is necessary to predict the characteristics of flooded areas caused by rainstorms, especially the flood depth. We applied the Naive Bayes theory to construct a model (NB model) to predict urban flood depth here in Zhengzhou. The model used 11 factors that affect the extent of flooding—rainfall, duration of rainfall, peak rainfall, the proportion of roads, woodlands, grasslands, water bodies and building, permeability, catchment area, and slope. The forecast depth of flooding from the NB model under different rainfall conditions was used to draw an urban inundation map by ArcGIS software. The results show that the probability and degree of urban flooding in Zhengzhou increases significantly after a return period of once every two years, and the flooded areas mainly occurred in older urban areas. The average root mean square error of prediction results was 0.062, which verifies the applicability and validity of our model in the depth prediction of urban floods. Our findings suggest the NB model as a feasible approach to predict urban flood depth.

Evaluation of Rainfall Erosivity Factor Estimation Using Machine and Deep Learning Models

Rainfall erosivity factor (R-factor) is one of the Universal Soil Loss Equation (USLE) input parameters that account for impacts of rainfall intensity in estimating soil loss. Although many studies have calculated the R-factor using various empirical methods or the USLE method, these methods are time-consuming and require specialized knowledge for the user. The purpose of this study is to develop machine learning models to predict the R-factor faster and more accurately than the previous methods. For this, this study calculated R-factor using 1-min interval rainfall data for improved accuracy of the target value. First, the monthly R-factors were calculated using the USLE calculation method to identify the characteristics of monthly rainfall-runoff induced erosion. In turn, machine learning models were developed to predict the R-factor using the monthly R-factors calculated at 50 sites in Korea as target values. The machine learning algorithms used for this study were Decision Tree, K-Nearest Neighbors, Multilayer Perceptron, Random Forest, Gradient Boosting, eXtreme Gradient Boost, and Deep Neural Network. As a result of the validation with 20% randomly selected data, the Deep Neural Network (DNN), among seven models, showed the greatest prediction accuracy results. The DNN developed in this study was tested for six sites in Korea to demonstrate trained model performance with Nash–Sutcliffe Efficiency (NSE) and the coefficient of determination (R2) of 0.87. This means that our findings show that DNN can be efficiently used to estimate monthly R-factor at the desired site with much less effort and time with total monthly precipitation, maximum daily precipitation, and maximum hourly precipitation data. It will be used not only to calculate soil erosion risk but also to establish soil conservation plans and identify areas at risk of soil disasters by calculating rainfall erosivity factors.

A GIS-Based Artificial Neural Network Model for Flood Susceptibility Assessment

This article presents a geographic information system (GIS)-based artificial neural network (GANN) model for flood susceptibility assessment of Keelung City, Taiwan. Various factors, including elevation, slope angle, slope aspect, flow accumulation, flow direction, topographic wetness index (TWI), drainage density, rainfall, and normalized difference vegetation index, were generated using a digital elevation model and LANDSAT 8 imagery. Historical flood data from 2015 to 2019, including 307 flood events, were adopted for a comparison of flood susceptibility. Using these factors, the GANN model, based on the back-propagation neural network (BPNN), was employed to provide flood susceptibility. The validation results indicate that a satisfactory result, with a correlation coefficient of 0.814, was obtained. A comparison of the GANN model with those from the SOBEK model was conducted. The comparative results demonstrated that the proposed method can provide good accuracy in predicting flood susceptibility. The results of flood susceptibility are categorized into five classes: Very low, low, moderate, high, and very high, with coverage areas of 60.5%, 27.4%, 8.6%, 2.5%, and 1%, respectively. The results demonstrate that nearly 3.5% of the study area, including the core district of the city and an exceedingly populated area including the financial center of the city, can be categorized as high to very high flood susceptibility zones.

Spatial Prediction of Future Flood Risk: An Approach to the Effects of Climate Change

Preparation of a flood probability map serves as the first step in a flood management program. This research develops a probability flood map for floods resulting from climate change in the future. Two models of Flexible Discrimination Analysis (FDA) and Artificial Neural Network (ANN) were used. Two optimistic (RCP2.6) and pessimistic (RCP8.5) climate change scenarios were considered for mapping future rainfall. Moreover, to produce probability flood occurrence maps, 263 locations of past flood events were used as dependent variables. The number of 13 factors conditioning floods was taken as independent variables in modeling. Of the total 263 flood locations, 80% (210 locations) and 20% (53 locations) were considered model training and validation. The Receiver Operating Characteristic (ROC) curve and other statistical criteria were used to validate the models. Based on assessments of the validated models, FDA, with a ROC-AUC = 0.918, standard error (SE = 0.038), and an accuracy of 0.86% compared to the ANN model with a ROC-AUC = 0.897, has the highest accuracy in preparing the flood probability map in the study area. The modeling results also showed that the factors of distance from the River, altitude, slope, and rainfall have the greatest impact on floods in the study area. Both models’ future flood susceptibility maps showed that the highest area is related to the very low class. The lowest area is related to the high class.

Flood Hazard Risk Mapping Using a Pseudo Supervised Random Forest

Devastating floods occur regularly around the world. Recently, machine learning models have been used for flood susceptibility mapping. However, even when these algorithms are provided with adequate ground truth training samples, they can fail to predict flood extends reliably. On the other hand, the height above nearest drainage (HAND) model can produce flood prediction maps with limited accuracy. The objective of this research is to produce an accurate and dynamic flood modeling technique to produce flood maps as a function of water level by combining the HAND model and machine learning. In this paper, the HAND model was utilized to generate a preliminary flood map; then, the predictions of the HAND model were used to produce pseudo training samples for a R.F. model. To improve the R.F. training stage, five of the most effective flood mapping conditioning factors are used, namely, Altitude, Slope, Aspect, Distance from River and Land use/cover map. In this approach, the R.F. model is trained to dynamically estimate the flood extent with the pseudo training points acquired from the HAND model. However, due to the limited accuracy of the HAND model, a random sample consensus (RANSAC) method was used to detect outliers. The accuracy of the proposed model for flood extent prediction, was tested on different flood events in the city of Fredericton, NB, Canada in 2014, 2016, 2018, 2019. Furthermore, to ensure that the proposed model can produce accurate flood maps in other areas as well, it was also tested on the 2019 flood in Gatineau, QC, Canada. Accuracy assessment metrics, such as overall accuracy, Cohen’s kappa coefficient, Matthews correlation coefficient, true positive rate (TPR), true negative rate (TNR), false positive rate (FPR) and false negative rate (FNR), were used to compare the predicted flood extent of the study areas, to the extent estimated by the HAND model and the extent imaged by Sentinel-2 and Landsat satellites. The results confirm that the proposed model can improve the flood extent prediction of the HAND model without using any ground truth training data.

Evaluation of the Impact of Rainfall Inputs on Urban Rainfall Models: A Systematic Review

Over the past several decades, urban flooding and other water-related disasters have become increasingly prominent and serious. Although the urban rain flood model’s benefits for urban flood simulation have been extensively documented, the impact of rainfall input to model simulation accuracy remains unclear. This systematic review aims to provide structured research on how rain inputs impact urban rain flood model’s simulation accuracy. The selected 48 peer-reviewed journal articles published between 2015 and 2019 on the Web of Science™ database were analyzed by key factors, including rainfall input type, calibration times and verification times. The results from meta-analysis reveal that when a traditional rain measurement was used as the rainfall input, model simulation accuracy was higher, i.e., the Nash–Sutcliffe efficiency coefficient (NSE) of traditional technology for rain measurement was higher than the 0.18 for the new technology rain measurement with respect to flow simulation. In addition, the single-field sub-flood calibration model was better than the multi-field sub-flood calibration model. NSE was higher than 0.14. The precision was better for the verification period; NSE of the calibration value showed a 0.07 higher verification value on average in flow simulation. These findings have certain significance for the development of future urban rain flood models and propose the development direction of the future urban rain flood model. Finally, in view of the rainfall input problem of the urban storm flood model, we propose the future development direction of the urban storm flood model.

Assessment of Landslide Susceptibility Combining Deep Learning with Semi-Supervised Learning in Jiaohe County, Jilin Province, China

Accurate and timely landslide susceptibility mapping (LSM) is essential to effectively reduce the risk of landslide. In recent years, deep learning has been successfully applied to landslide susceptibility assessment due to the strong ability of fitting. However, in actual applications, the number of labeled samples is usually not sufficient for the training component. In this paper, a deep neural network model based on semi-supervised learning (SSL-DNN) for landslide susceptibility is proposed, which makes full use of a large number of spatial information (unlabeled data) with limited labeled data in the region to train the mode. Taking Jiaohe County in Jilin Province, China as an example, the landslide inventory from 2000 to 2017 was collected and 12 metrological, geographical, and human explanatory factors were compiled. Meanwhile, supervised models such as deep neural network (DNN), support vector machine (SVM), and logistic regression (LR) were implemented for comparison. Then, the landslide susceptibility was plotted and a series of evaluation tools such as class accuracy, predictive rate curves (AUC), and information gain ratio (IGR) were calculated to compare the prediction of models and factors. Experimental results indicate that the proposed SSL-DNN model (AUC = 0.898) outperformed all the comparison models. Therefore, semi-supervised deep learning could be considered as a potential approach for LSM.

Depth prediction of urban flood under different rainfall return periods based on deep learning and data warehouse

With the global climate change and the rapid urbanization process, there is an increase in the risk of urban floods. Therefore, undertaking risk studies of urban floods, especially the depth prediction of urban flood is very important for urban flood control. In this study, an urban flood data warehouse was established with available structured and unstructured urban flood data. In this study, an urban flood data warehouse was established with available structured and unstructured urban flood data. Based on this, a regression model to predict the depth of urban flooded areas was constructed with deep learning algorithm, named Gradient Boosting Decision Tree (GBDT). The flood condition factors used in modeling were rainfall, rainfall duration, peak rainfall, evaporation, land use (the proportion of roads, woodlands, grasslands, water bodies and building), permeability, catchment area, and slope. Based on the rainfall data of different rainfall return periods, flood condition maps were produced using GIS. In addition, the feature importance of these conditioning factors was determined based on the regression model. The results demonstrated that the growth rate of the number and depth of the water accumulation points increased significantly after the rainfall return period of 'once in every two years' in Zhengzhou City, and the flooded areas mainly occurred in the old urban areas and parts of southern Zhengzhou. The relative error of prediction results was 11.52%, which verifies the applicability and validity of the method in the depth prediction of urban floods. The results of this study can provide a scientific basis for urban flood control and drainage.

GIS Based Hybrid Computational Approaches for Flash Flood Susceptibility Assessment

Flash floods are one of the most devastating natural hazards; they occur within a catchment (region) where the response time of the drainage basin is short. Identification of probable flash flood locations and development of accurate flash flood susceptibility maps are important for proper flash flood management of a region. With this objective, we proposed and compared several novel hybrid computational approaches of machine learning methods for flash flood susceptibility mapping, namely AdaBoostM1 based Credal Decision Tree (ABM-CDT); Bagging based Credal Decision Tree (Bag-CDT); Dagging based Credal Decision Tree (Dag-CDT); MultiBoostAB based Credal Decision Tree (MBAB-CDT), and single Credal Decision Tree (CDT). These models were applied at a catchment of Markazi state in Iran. About 320 past flash flood events and nine flash flood influencing factors, namely distance from rivers, aspect, elevation, slope, rainfall, distance from faults, soil, land use, and lithology were considered and analyzed for the development of flash flood susceptibility maps. Correlation based feature selection method was used to validate and select the important factors for modeling of flash floods. Based on this feature selection analysis, only eight factors (distance from rivers, aspect, elevation, slope, rainfall, soil, land use, and lithology) were selected for the modeling, where distance to rivers is the most important factor for modeling of flash flood in this area. Performance of the models was validated and compared by using several robust metrics such as statistical measures and Area Under the Receiver Operating Characteristic (AUC) curve. The results of this study suggested that ABM-CDT (AUC = 0.957) has the best predictive capability in terms of accuracy, followed by Dag-CDT (AUC = 0.947), MBAB-CDT (AUC = 0.933), Bag-CDT (AUC = 0.932), and CDT (0.900), respectively. The proposed methods presented in this study would help in the development of accurate flash flood susceptible maps of watershed areas not only in Iran but also other parts of the world.

Spatial Prediction of Landslide Susceptibility Based on GIS and Discriminant Functions

The areas where landslides occur frequently pose severe threats to the local population, which necessitates conducting regional landslide susceptibility mapping (LSM). In this study, four models including weight-of-evidence (WoE) and three WoE-based models, which were linear discriminant analysis (LDA), Fisher’s linear discriminant analysis (FLDA), and quadratic discriminant analysis (QDA), were used to obtain the LSM in the Nanchuan region of Chongqing, China. Firstly, a dataset was prepared from sixteen landslide causative factors, including eight topographic factors, three distance-related factors, and five environmental factors. A landslide inventory map including 298 landslide locations was also constructed and randomly divided with a ratio of 70:30 as training and validation data. Subsequently, the WoE method was used to estimate the relationship between landslides and the landslide causative factors, which assign a weight value to each class of causative factors. Finally, four models were applied using the training dataset, and the predictive performance of each model was compared using the validation datasets. The results showed that FLDA had a higher performance than the other three models according to the success rate curve (SRC) and prediction rate curve (PRC), illustrating that it could be considered a promising approach for landslide susceptibility mapping in the study area.

Identifying the Contributions of Multi-Source Data for Winter Wheat Yield Prediction in China

Wheat is a leading cereal grain throughout the world. Timely and reliable wheat yield prediction at a large scale is essential for the agricultural supply chain and global food security, especially in China as an important wheat producing and consuming country. The conventional approach using either climate or satellite data or both to build empirical and crop models has prevailed for decades. However, to what extent climate and satellite data can improve yield prediction is still unknown. In addition, socio-economic (SC) factors may also improve crop yield prediction, but their contributions need in-depth investigation, especially in regions with good irrigation conditions, sufficient fertilization, and pesticide application. Here, we performed the first attempt to predict wheat yield across China from 2001 to 2015 at the county-level by integrating multi-source data, including monthly climate data, satellite data (i.e., Vegetation indices (VIs)), and SC factors. The results show that incorporating all the datasets by using three machine learning methods (Ridge Regression (RR), Random Forest (RF), and Light Gradient Boosting (LightGBM)) can achieve the best performance in yield prediction (R2: 0.68~0.75), with the most individual contributions from climate (~0.53), followed by VIs (~0.45), and SC factors (~0.30). In addition, the combinations of VIs and climate data can capture inter-annual yield variability more effectively than other combinations (e.g., combinations of climate and SC, and combinations of VIs and SC), while combining SC with climate data can better capture spatial yield variability than others. Climate data can provide extra and unique information across the entire growing season, while the peak stage of VIs (Mar.~Apr.) do so. Furthermore, incorporating spatial information and soil proprieties into the benchmark models can improve wheat yield prediction by 0.06 and 0.12, respectively. The optimal wheat prediction can be achieved with approximately a two-month leading time before maturity. Our study develops timely and robust methods for winter wheat yield prediction at a large scale in China, which can be applied to other crops and regions.

Modeling flood susceptibility using data-driven approaches of naïve Bayes tree, alternating decision tree, and random forest methods

Floods are one of the most devastating types of disasters that cause loss of lives and property worldwide each year. This study aimed to evaluate and compare the prediction capability of the naïve Bayes tree (NBTree), alternating decision tree (ADTree), and random forest (RF) methods for the spatial prediction of flood occurrence in the Quannan area, China. A flood inventory map with 363 flood locations was produced and partitioned into training and validation datasets through random selection with a ratio of 70/30. The spatial flood database was constructed using thirteen flood explanatory factors. The probability certainty factor (PCF) method was used to analyze the correlation between the factors and flood occurrences. Consequently, three flood susceptibility maps were produced using the NBTree, ADTree, and RF methods. Finally, the area under the curve (AUC) and statistical measures were used to validate the flood susceptibility models. The results indicated that the RF method is an efficient and reliable model in flood susceptibility assessment, with the highest AUC values, positive predictive rate, negative predictive rate, sensitivity, specificity, and accuracy for the training (0.951, 0.892, 0.941, 0.945, 0.886, and 0.915, respectively) and validation (0.925, 0.851, 0.938, 0.945, 0.835, and 0.890, respectively) datasets.

Tackling environmental challenges in pollution controls using artificial intelligence: A review

This review presents the developments in artificial intelligence technologies for environmental pollution controls. A number of AI approaches, which start with the reliable mapping of nonlinear behavior between inputs and outputs in chemical and biological processes in terms of prediction models to the emerging optimization and control algorithms that study the pollutants removal processes and intelligent control systems, have been developed for environmental clean-ups. The characteristics, advantages and limitations of AI methods, including single and hybrid AI methods, were overviewed. Hybrid AI methods exhibited synergistic effects, but with computational heaviness. The up-to-date review summarizes i) Various artificial neural networks employed in wastewater degradation process for the prediction of removal efficiency of pollutants and the search of optimizing experimental conditions; ii) Evaluation of fuzzy logic used for intelligent control of aerobic stage of wastewater treatment process; iii) AI-aided soft-sensors for precisely on-line/off-line estimation of hard-to-measure parameters in wastewater treatment plants; iv) Single and hybrid AI methods applied to estimate pollutants concentrations and design monitoring and early-warning systems for both aquatic and atmospheric environments; v) AI modelings of short-term, mid-term and long-term solid waste generations, and various ANNs for solid waste recycling and reduction. Finally, the future challenges of AI-based models employed in the environmental fields are discussed and proposed.

Urban Flood Hazard Modeling Using Self-Organizing Map Neural Network

Floods are the most common natural disaster globally and lead to severe damage, especially in urban environments. This study evaluated the efficiency of a self-organizing map neural network (SOMN) algorithm for urban flood hazard mapping in the case of Amol city, Iran. First, a flood inventory database was prepared using field survey data covering 118 flooded points. A 70:30 data ratio was applied for training and validation purposes. Six factors (elevation, slope percent, distance from river, distance from channel, curve number, and precipitation) were selected as predictor variables. After building the model, the odds ratio skill score (ORSS), efficiency (E), true skill statistic (TSS), and the area under the receiver operating characteristic curve (AUC-ROC) were used as evaluation metrics to scrutinize the goodness-of-fit and predictive performance of the model. The results indicated that the SOMN model performed excellently in modeling flood hazard in both the training (AUC = 0.946, E = 0.849, TSS = 0.716, ORSS = 0.954) and validation (AUC = 0.924, E = 0.857, TSS = 0.714, ORSS = 0.945) steps. The model identified around 23% of the Amol city area as being in high or very high flood risk classes that need to be carefully managed. Overall, the results demonstrate that the SOMN model can be used for flood hazard mapping in urban environments and can provide valuable insights about flood risk management.

Flood Risk Assessment of Global Watersheds Based on Multiple Machine Learning Models

Machine learning algorithms are becoming more and more popular in natural disaster assessment. Although the technology has been tested in flood susceptibility analysis of several watersheds, research on global flood disaster risk assessment based on machine learning methods is still rare. Considering that the watershed is the basic unit of water management, the purpose of this study was to conduct a risk assessment of floods in the global fourth-level watersheds. Thirteen conditioning factors were selected, including: maximum daily precipitation, precipitation concentration degree, altitude, slope, relief degree of land surface, soil type, Manning coefficient, proportion of forest and shrubland, proportion of artificial surface, proportion of cropland, drainage density, population, and gross domestic product. Four machine learning algorithms were selected in this study: logistic regression, naive Bayes, AdaBoost, and random forest. The global susceptibility assessment model was constructed based on four machine learning algorithms, thirteen conditioning factors, and global flood inventories. The evaluation results of the model show that the random forest performed better in the test, and is an efficient and reliable tool in flood susceptibility assessment. Sensitivity analysis of the conditioning factors showed that precipitation concentration degree and Manning coefficient were the main factors affecting flood risk in the watersheds. The susceptibility map showed that fourth-level watersheds in the global high-risk area accounted for a large proportion of the total watersheds. With the increase of extreme hydrological events caused by climate change, global flood disasters are still one of the most threatening natural disasters. The global flood susceptibility map from this study can provide a reference for global flood management.

A Time-Integrated Index for Flood Risk to Resistance Capacity

The lack of open spaces and the intense land use occupation in flood plains makes floods in consolidated urban areas difficult to mitigate. In these areas, setting a standard pre-defined return period for projects can limit and even preclude flood mitigation actions. However, it is possible to propose flood control alternatives that are compatible with available spaces. Thus, determining how much the original risk is reduced and how significant the residual risk can be becomes the main target. In this context, a time-integrated index for risk to resistance capacity is proposed to address these questions. This index correlates the exposure of buildings and urban infrastructure to the hazard of a given flood and is then evaluated over a project horizon through a sequence of events. The proposed index is applied to the Canal do Mangue catchment, a highly urbanized watershed located in Rio de Janeiro. The results demonstrate the difficulty of designing flood mitigation measures in extremely occupied watersheds and the importance of evaluating residual risks associated with proposed projects. As an additional result, a scenario with concentrated measures is compared to another with distributed interventions, evidencing the greater coverage of the latter.