Adaptive Momentum-Backpropagation Algorithm for Flood Prediction and Management in the Internet of Things

Flooding is a hazardous natural calamity that causes significant damage to lives and infrastructure in the real world.Therefore,timely and accurate decision-making is essential for mitigating flood-related damages.The traditional flood prediction techniques often encounter challenges in accuracy,timeliness,complexity in handling dynamic flood patterns and leading to substandard flood management strategies.To address these challenges,there is a need for advanced machine learning models that can effectively analyze Internet of Things(IoT)-generated flood data and provide timely and accurate flood predictions.This paper proposes a novel approach-the Adaptive Momentum and Backpropagation(AM-BP)algorithm-for flood prediction and management in IoT networks.The AM-BP model combines the advantages of an adaptive momentum technique with the backpropagation algorithm to enhance flood prediction accuracy and efficiency.Real-world flood data is used for validation,demonstrating the superior performance of the AM-BP algorithm compared to traditional methods.In addition,multilayer high-end computing architecture(MLCA)is used to handle weather data such as rainfall,river water level,soil moisture,etc.The AM-BP’s real-time abilities enable proactive flood management,facilitating timely responses and effective disaster mitigation.Furthermore,the AM-BP algorithm can analyze large and complex datasets,integrating environmental and climatic factors for more accurate flood prediction.The evaluation result shows that the AM-BP algorithm outperforms traditional approaches with an accuracy rate of 96%,96.4%F1-Measure,97%Precision,and 95.9%Recall.The proposed AM-BP model presents a promising solution for flood prediction and management in IoT networks,contributing to more resilient and efficient flood control strategies,and ensuring the safety and well-being of communities at risk of flooding.

Prediction of the flooding area of the northeastern Caspian Sea from satellite images

Studying the dynamic changes in the coastline of the northeastern Caspian Sea is significant since the level of the Caspian is unstable,and the coastline change can cause enormous damage to the ecology,economy,and population of the coastal part of Kazakhstan.In this work,we use remote sensing and Geographic Information System(GIS)technologies to study the changes in the coastline of the northeastern Caspian Sea and predict the extent of flooding with increasing water levels.The proposed methodology for creating dynamic maps can be used to monitor the coastline and forecast the extent of flooding in the area.As a result of this work,the main factors affecting changes in the coastline were identified.After analyzing the water level data from 1988 to 2019,it was revealed that the rise in water level was observed from 1980 to 1995.The maximum sea level rise was recorded at-26.04 m.After that,the sea level began to fall,and between 1996 and 2009,there were no significant changes;the water level fluctuated with an average of-27.18 m.Then,a map of the water level dynamics in the Caspian Sea from 1988 to 2019 was compiled.According to the dynamics map,water level rise and significant coastal retreat were revealed,especially in the northern part of the Caspian Sea and the northern and southern parts of Sora Kaydak.The method for predicting the estimated flooding area was described.As a result,based on a single map,the flooding area of the northeast coast was predicted.A comparative analysis of Landsat and SRTM data is presented.

Flood prediction and assessment of vulnerability risk in the southern coasts of the Caspian Sea

The southern part of the Caspian Sea shoreline in Iran with a length of 813 km has different topographic conditions.Owing to sea fluctuation,these zones have various dimensions in different times.During the last years,the Caspian Sea experienced enormous destructive rises.The historical information and tidal gauge measurements showed different ranges of sea rise from
30 m to
22 m from the mean sea level.On the other hand,the probable flooding zone is related to slope gradient of coasts.To help the determination of the probable flooding area owing to sea level rises,the coastal zones can be modelled using geographic information system(GIS)environment as vulnerability risk rates.These rates would be useful for making decisions in coastal management programs.This study examined different scenarios of sea rise to determine hazard-flooding rates in the coastal cities of the Mazandaran province and classified them based on vulnerability risk rates.The 1:2000 scale topographic maps of the coastal zones were prepared to extract topographic information and construct the coastal digital elevation model.With the presumption of half-metre sea rise scenarios,the digital elevation models classified eight scenarios from
26 to
22 m.The flooding areas in each scenario computed for 11 cities respectively.The vulnerability risk rate in each rise scenario was computed by dividing the flooded area of each scenario to city area.The results showed that in the first four scenarios,from
26 to
24 m,the Behshahr,Joibar,Neka and Babolsar cites would be more vulnerable than other cites.Moreover,for the second four scenarios from
24 to
22 m sea level rise scenario,only the coastal area of Chalous city would be vulnerable.It was also observed that the coastal region of Behshahr would be critical in total scenarios.Further studies would be necessary to complete this assessment by considering social-economic and land use information to estimate the exact hazardous and vulnerable zones.

On the calibration of the parameters governing the PWRI distributed hydrological model for flood prediction

The Public Works Research Institute Distributed Hydrological(PWRI-DH)for flood modeling is a combination of the tank model and the kinematic wave method.In the PWRI-DH model,fitting the required parameters plays a fundamental role.The developers of the PWRI-DH model have introduced the capability of obtaining parameters automatically using the baseline parameters;however,the results are not always the expected results because they depend on several factors and must be calibrated manually.The last issue has limited the interest of researchers regarding in the usage of the PWRI-DH model.In this paper,we present a methodology to obtain the parameters required for the PWRI-DH model that enables to focusing only on the key parameters.First,a parametric study is performed by identifying the influence of each parameter in the discharge.From this study,we found that only four parameters play a fundamental role in the flood modeling using the PWRI-DH model.Five flood events in the Upper Aikawa River basin are used to calibrate the model.The results showed that the proposed methodology is suitable and improve the efficient on the flood simulation of Aikawa River and similar rivers,when using the PWRI-DH model.

On the Flood Disasters in the Lower Jingjiang Reaches: Grey Prediction Model and Application

In the light of the historical substantial data (covering a 70-year period) collected in the Lower Jingjiang segment and aided by topological grey method, here we attempt to characterize the occurrence and future trend of flood calamities in the study area. Our findings indicate that overall the high-frequent flood disasters with middle to lower damage prevail at present. A series of dramatic flood waves will appear in the years of 2016, 2022, 2030 and 2042, particularly a destructive flood will occur between 2041 and 2045 in the Lower Jingjiang reaches. Typical of sensitive response to flood hazards in close association with its special geographical location, the lower Jingjiang segment hereby can reflect the development trend of floods in the middle Yangtze reaches. According to the results, a good fitness was revealed between the prediction and practical values. This actually hints that the topological grey method is an effective mathematical means of resolving problems containing uncertainty and indetermination, thus providing valuable information for the flood prediction in the middle Yangtze catchment.

Analysis of Causes and Seasonal Prediction of the Severe Floods in Yangtze／Huaihe Basins during Summer 1991

The present paper shows that a seasonal prediction for the large scale flooding and waterlogging of the mid-lower Yangtze/ Huaihe River basins in the summer of 1991 made successfully in early April 1991.The seasonal forecasting method and some predictors are also introduced and analyzed herein. Because the extra extent of the abnormally early onset of the plum rain period in 1991 was unexpected,great efforts have been made to find out the causes of this abnormality. These causes are mainly associated with the large scale warming of SST surrounding the south and east part of Asia during the preceding winter,while the ENSO-like pattern of SSTA occurred in the North Pacific.In addition,the possible influence of strong solar proton events is analyzed.In order to improve the seasonal pre4iction the usage of the predicted SOl in following spring/summer is also introduced.The author believes thatthe regional climate anomaly can be correctly predicted for one season ahead only on the basis of physical understanding of the interactions of many preceding factors.

Rapid Prediction Model for Urban Floods Based on a Light Gradient Boosting Machine Approach and Hydrological–Hydraulic Model

Global climate change and sea level rise have led to increased losses from flooding.Accurate prediction of floods is essential to mitigating flood losses in coastal cities.Physically based models cannot satisfy the demand for real-time prediction for urban flooding due to their computational complexity.In this study,we proposed a hybrid modeling approach for rapid prediction of urban floods,coupling the physically based model with the light gradient boosting machine(LightGBM)model.A hydrological–hydraulic model was used to provide sufficient data for the LightGBM model based on the personal computer storm water management model(PCSWMM).The variables related to rainfall,tide level,and the location of flood points were used as the input for the LightGBM model.To improve the prediction accuracy,the hyperparameters of the LightGBM model are optimized by grid search algorithm and K-fold cross-validation.Taking Haidian Island,Hainan Province,China as a case study,the optimum values of the learning rate,number of estimators,and number of leaves of the LightGBM model are 0.11,450,and 12,respectively.The Nash-Sutcliffe efficiency coefficient(NSE)of the LightGBM model on the test set is 0.9896,indicating that the LightGBM model has reliable predictions and outperforms random forest(RF),extreme gradient boosting(XGBoost),and k-nearest neighbor(KNN).From the LightGBM model,the variables related to tide level were analyzed as the dominant variables for predicting the inundation depth based on the Gini index in the study area.The proposed LightGBM model provides a scientific reference for flood control in coastal cities considering its superior performance and computational efficiency.

Flood Hazard Mapping of Lower Indus Basin Using Multi-Criteria Analysis

Flooding has been one of the recurring occurred natural disasters that induce detrimental impacts on humans, property and environment. Frequent floods is a severe issue and a complex natural phenomenon in Pakistan with respect to population affected, environmental degradations, and socio-economic and property damages. The Super Flood, which hit Sindh in 2010, has turned out to be a wakeup call and has underlined the overwhelming challenge of natural calamities, as 2010 flood and the preceding flood in 2011 caused a huge loss to life, property and land use. These floods resulted in disruption of power, telecommunication, and water utilities in many districts of Pakistan, including 22 districts of Sindh. These floods call for risk assessment and hazard mapping of Lower Indus Basin flowing in the Sindh Province as such areas were also inundated in 2010 flood, which were not flooded in the past in this manner. This primary focus of this paper is the use of Multi-criteria Evaluation (MCE) methods in integration with the Geographical Information System (GIS) for the analysis of areas prone to flood. This research demonstrated how GIS tools can be used to produce map of flood vulnerable areas using MCE techniques. Slope, Aspect, Curvature, Soil, and Distance from Drainage, Land use, Precipitation, Flow Direction, and Flow Accumulation are taken as the causative factors for flooding in Lower Indus Basin. Analytical Hierarchy Process-AHP was used for the calculation of weights of all these factors. Finally, a flood hazard Map of Lower Indus Basin was generated which delineates the flood prone areas in the Sindh province along Indus River Basin that could be inundated by potential flooding in future. It is aimed that flood hazard mapping and risk assessment using open source geographic information system can serve as a handy tool for the development of land-use strategies so as to decrease the impact from flooding.

THE FLOODS AND DROUGHTS OF THE LOWER YANGTZE VALLEY AND THEIR PREDICTIONS

During Ching dynasty from 1644 to 1911, an interval of 268 years, there occurred in the lower Yangtze valley 197 floods and 156 droughts. The most serious droughts came in 1785, 1814, and 1856; and the most disastrous floods in 1680,

A Framework on Fast Mapping of Urban Flood Based on a Multi-Objective Random Forest Model

Fast and accurate prediction of urban flood is of considerable practical importance to mitigate the effects of frequent flood disasters in advance.To improve urban flood prediction efficiency and accuracy,we proposed a framework for fast mapping of urban flood:a coupled model based on physical mechanisms was first constructed,a rainfall-inundation database was generated,and a hybrid flood mapping model was finally proposed using the multi-objective random forest(MORF)method.The results show that the coupled model had good reliability in modelling urban flood,and 48 rainfall-inundation scenarios were then specified.The proposed hybrid MORF model in the framework also demonstrated good performance in predicting inundated depth under the observed and scenario rainfall events.The spatial inundated depths predicted by the MORF model were close to those of the coupled model,with differences typically less than 0.1 m and an average correlation coefficient reaching 0.951.The MORF model,however,achieved a computational speed of 200 times faster than the coupled model.The overall prediction performance of the MORF model was also better than that of the k-nearest neighbor model.Our research provides a novel approach to rapid urban flood mapping and flood early warning.

Merging GIS and Machine Learning Techniques: A Paper Review

GIS (Geographic Information Systems) data showcase locations of earth observations or features, their associated attributes and spatial relationships that exist between such observations. Analysis of GIS data varies widely and may include some modeling and predictions which are usually computing-intensive and complicated, especially, when large datasets are involved. With advancement in computing technologies, techniques such as Machine learning (ML) are being suggested as a potential game changer in the analysis of GIS data because of their comparative speed, accuracy, automation, and repeatability. Perhaps, the greatest benefit of using both GIS and ML is the ability to transfer results from one database to another. GIS and ML tools have been used extensively in medicine, urban development, and environmental modeling such as landslide susceptibility prediction (LSP). There is also the problem of data loss during conversion between GIS systems in medicine, while in geotechnical areas such as erosion and flood prediction, lack of data and variability in soil has limited the use of GIS and ML techniques. This paper gives an overview of the current ML methods that have been incorporated into the spatial analysis of data obtained from GIS tools for LSP, health, and urban development. The use of Supervised Machine Learning (SML) algorithms such as decision trees, SVM, KNN, and perceptron including Unsupervised Machine Learning algorithms such as k-means, elbow algorithms, and hierarchal algorithm have been discussed. Their benefits, as well as their shortcomings as studied by several researchers have been elucidated in this review. Finally, this review also discusses future optimization techniques.

STUDY ON MIXED MODEL OF NEURAL NETWORK FOR FARMLAND FLOOD/DROUGHT PREDICTION

The paper concerns a flood/drought prediction model involving the continuation of time series of a predictand and the physical factors influencing the change of predictand.Attempt is made to construct the model by the neural network scheme for the nonlinear mapping relation based on multi-input and single output.The model is found of steadily higher predictive accuracy by testing the output from one and multiple stepwise predictions against observations and comparing the results to those from a traditional statistical model.

MODELING AND PREDICTION CONCERNING TIME SERIES OF FLOOD/DROUGHT RUNS USING THE SELF-EXCITING THRESHOLD AUTOREGRESSIVE MODEL

When linear regressive models such as AR or ARMA model are used for fitting and predicting climatic time series,results are often not sufficiently good because nonlinear variations in the time series.In this paper, a nonlinear self-exciting threshold autoregressive(SETAR)model is applied to modeling and predicting the time series of flood/drought runs in Beijing,which were derived from the graded historical flood/drought records in the last 511 years(1470—1980).The results show that the modeling and predicting with the SETAR model are much better than that of the AR model.The latter can predict the flood/drought runs with a length only less than two years,while the formal can predict more than three-year length runs.This may be due to the fact that the SETAR model can renew the model according to the run-turning points in the process of predic- tion,though the time series is nonstationary.