Shanghai Urban Flood Simulation And Risk Response Based On Improved DEM

Flood disaster has become a serious environmental problem threatening coastal areas.With current global urbanization,a large number of people have moved into cities,resulting in a high concentration of property and resources,which will cause a lot of economic losses and casualties once floods occur.Therefore,accurate prediction of flood disasters is of great importance to maintain the order of human production and life and reduce the impact of disasters on human life.The current flood risk analysis has entered the era of high-precision numerical simulation,and a large number of hydrodynamic models can effectively simulate the diffusion process of water flow on the ground to achieve relatively accurate and convenient flood prediction.Digital Elevation Model（DEM）plays a critical role in hydrologic and hydraulic modelling.Flood inundation mapping is highly dependent on the accuracy of DEMs.There are many open-access DEM products available for scientific and non-commercial use,and there are large vertical differences between open-access DEM products due to the use of different observation satellites and algorithms.This problem is particularly acute in coastal cities with flat topography.Thus,it is necessary to assess the differences by input DEMs in flood simulation and to reduce anomalous errors of DEMs.This thesis applies urban flood simulation in Huangpu River by using the LISFLOOD-FP hydrodynamic model and six open-access DEMs（i.e.,SRTM,GDEM,AW3 D,MERIT,NASADEM and Coastal DEM）,and discusses the differences in flood inundation mapping.Then,the DEMs are modified by using two statistical methods-Dixon criterion and direct mean method to generate an improved DEM-MDEM（modified DEM）,and the results are compared with the original DEMs.Finally,through the good performance DEM superposition of Huangpu River embankment data,comprehensive analysis of the current Huangpu River fortification level and potential risks,and according to the results.The main results are as follows:（1）There are significant differences in flood inundation results for the six original DEM sets under the same boundary conditions,and the extreme case of choosing different DEMs under the same conditions in this study leads to more than a 1200-fold difference in inundation area predictions.This suggests that using only a single DEM dataset may lead to huge biased flooding results,which is not sufficient for exposure analysis and flood management responses that require high confidence levels.Researchers and experts associated with flood simulation should pay attention to the uncertainty of results due to DEM quality.（2）The accuracy of MDEM corrected by Dixon criterion for predicting inundation extent is improved,in addition to reducing errors in the extreme water depths.This result indicates that MDEM provides some performance improvements in the accuracy of flood simulations and contributes to the quality of topographic data.A freely available accurate and highresolution DEM is still needed to support robust flood mapping.Flood-related stakeholders should pay attention to the uncertainty caused by DEM quality.（3）The comparison of several DEMs with outstanding performance in the study area reveals that the current level of protection of the Huangpu River is generally good,with 50-year and 100-year flood events hardly posing a threat to the middle reaches of the Huangpu River and the downtown area of Shanghai.The upstream to the middle of the Huangpu River is still potentially threatening,but the design expectations are still met considering that the upstream to the middle reaches were not developed extensively at the time of the previous planning and therefore the protection standard is relatively lower than that of the city center.However,combined with Shanghai’s 14 th Five-Year Plan and the 2035 Vision,the strategic development plan for Qingpu New City and Fengxian New City,it is important to raise the standard of protection for the upstream section of the Huangpu River from Qingpu to Minhang and Fengxian,and even for the whole Huangpu River,to protect the social and economic development of Shanghai,resist extreme disasters,and improve the happiness of residents’ life.This study focuses on the potential impact of topographic data uncertainty in the numerical simulation of flooding and proposes a method to improve the accuracy of topographic data,which is investigated by the flooding event of Huangpu River in Shanghai.The study expands the current thinking for DEM performance enhancement and combines the new elevation data to carry out an example study in Shanghai,which provides a certain reference for Shanghai flood control and flood reduction.