Flood Hazard Mapping And Emergency Evacuation Planning In Bangladesh

Bangladesh experiences frequent hydro-climatic disasters,such as flooding.Flood is a recurring phenomenon in the study area.This led to serious injury or death,infrastructure damages,socioeconomic instability,and biodiversity loss.Thus,regular monitoring and awareness of local coping mechanisms are indeed essential to address the most vulnerable communities to flood.In addition,these disasters are believed to be associated with land-use changes and climate variability.However,the identification of factors that lead to flooding is challenging.Pointing to the fact,this study is conducted to determine the causative factors affecting flood hazards in Bangladesh in four different scales as case studies,i.e.,national scale(20°34?N and 88°01?E to 26°38?N and 92°41?E),regional scale(24°30?N and 91°40?E),basin-scale(25°10?N and 91°30?E),and local scale(25°04?N and 91°25?E to 25°60?N and 91°29?E).Furthermore,this study discussed the influence of anthropogenic factors(land cover dynamics,population growth,and road infrastructure)on the relative changes of flooding areas.In addition,the current study proposed mathematical models to optimize the decisions on emergency relief materials(humanitarian aids),evacuating flood victims,and determine temporary site allocations in regards to flood hazard.Floods are the most devastating natural hazard in Bangladesh.The country experiences multi-type floods(i.e.,fluvial,flash,pluvial,and surge floods)every year.However,areas prone to multi-type floods have not yet been assessed at a national scale.At first,this study used locally weighted linear regression(LWLR),random subspace(RSS),reduced error pruning tree(REPTree),random forest(RF),and M5P model tree algorithms in a hybrid ensemble to assess multi-type flood probabilities at a national scale in Bangladesh.We used historical flood data(1988–2020),remote sensing images(e.g.,MODIS,Landsat 5–8,and Sentinel-1),and topography,hydrogeology,and environmental datasets to train and validate the proposed algorithms.The results were evaluated using statistical measures,i.e.,mean absolute error(MAE),root-mean-square error(RMSE),coefficient of determination(R~2),relative absolute error(RAE),root relative squared error(RRSE),and Taylor diagrams,based on the training and testing datasets.According to the results,the stacking ensemble machine learning LWLR-RF algorithm performed better than the other algorithms at predicting flood probabilities,with R~2=0.967–0.999,MAE=0.022–0.117,RMSE=0.029–0.148,RAE=4.48–23.38%,and RRSE=5.88–29.69%for the training and testing datasets.The resultant maps constructed using the LWLR-RF algorithm revealed that the proportions of different categories of flood probable areas in Bangladesh are as follows:low(1.50–5.72%),medium(1.04–8.90%),high(0.52–12.66%),and very high(0.90–13.77%).These findings can guide future flood risk management and sustainable land-use planning in the study area(Chapter 3).Second,this study developed models to identify optimal spatial distribution of emergency evacuation centers(EECs)such as schools,colleges,hospitals,and fire stations to improve flood emergency planning(regional scale)in the Sylhet region of northeastern Bangladesh.The use of location-allocation models(LAMs)for evacuation in regard to flood victims is essential to minimize disaster risk.In the first step,flood susceptibility maps(regional scale)were developed using machine learning models(MLMs),including Levenberg–Marquardt back propagation(LM-BP)neural network and decision trees(DT)and MCDM method.Performance of the MLMs and MCDM techniques were assessed considering the AUROC curve.Mathematical approaches in a GIS for four well-known LAM problems affecting emergency rescue time are proposed:maximal covering location problem(MCLP),the maximize attendance(MA),p-median problem(PMP),and the location set covering problem(LSCP).The results showed that existing EECs were not optimally distributed and that some areas were not adequately served by EECs(i.e.,not all demand points could be reached within a 60-minute travel time).We concluded that the proposed models can be used to improve planning of the distribution of EECs,and that application of the models could contribute to reducing human casualties,property losses,and improve emergency operation(Chapter 4).Third,this study mapped flood susceptibility(basin scale)in the northeast region of Bangladesh using Bayesian regularization back propagation(BRBP)neural network,classification®ression trees(CART),a statistical model(STM)using the evidence belief function(EBF),and their ensemble models(EMs)for three time periods(2000,2014,and 2017).The MLAs,STM,and EMs were assessed considering the area under the curve—receiver operating characteristic(AUC-ROC).Evaluation of the accuracy levels of the aforementioned algorithms revealed that EM4(BRBP-CART-EBF)outperformed(AUC>90%)standalone and other ensemble models for the three-time periods analyzed.Further,this study investigated the relationships among land cover change(LCC),population growth(PG),road density(RD),and relative change of flooding(RCF)areas for the period between 2000 and 2017.The results exhibited that areas with very high susceptibility to flooding increased by 19.72%between 2000 and 2017,while the PG rate increased by 51.68%over the same period.The Pearson correlation coefficient for RCF and RD was calculated to be 0.496.These findings highlighted a significant correlation between floods and causative factors.The study findings could be valuable to policymakers and resource managers as they can lead to improvements in flood management and reduction in flood damage and risks(Chapter 5).Fourth,a flood hazard map was prepared using the hydrodynamic model(HM)–FLO 2D coupled with the machine learning algorithm(MLA)–scaled conjugate gradient neural network(SCG-NN)at a local scale.The performance of the MLA was evaluated using a validation dataset and statistical measures such as the mean square error(MSE:0.080),root mean square error(RMSE:0.282),and coefficient of determination(R~2:0.830).According to the generated flood hazard map,most of the study area was classified as low(47.85%)or moderate(27.47%)hazardous zones,whereas only a small portion was delineated as high(20.64%)or very high(4.04%)hazardous zones.The accuracy of the hazard map(HM-MLA)versus ground truth was tested statistically and was found to be high.An investigation of local flood management strategies revealed that the current information system is not well prepared for emergencies,including the quantification of emergency relief necessities.Therefore,this study proposes a humanitarian aid information system(HAIS)to enhance the emergency support for flood victims.We conclude that the proposed HAIS will help humanitarian organizations to coordinate relief operations effectively in the worst-hit regions across the country(Chapter 6).Finally,this study concluded with several recommendations for the policymakers and relevant national,regional,and local authorities to design the control measures for potential floods in Bangladesh(Chapter 7).