| Disaster geomorphology refers to the geomorphic effects and phenomena to endanger human's life and infrastructure.Bank collapse is a common disaster geomorphology in global large rivers.It may threaten the safety of levee,cause land loss of waterfront,and gradually destroy the stability of river course,resulting in irreparable consequences in river regulation projects.Bank stability and its protection have always been a hot issue in the field of river geomorphology and hydraulic engineering.With the application of multibeam underwater detection technology,the influence of subaqueous scarps,scour pits and dunes on the stability of river banks has also been concerned.Therefore,the bank collapse,scour pits,scarps and dunes are collectively referred to as subaqueous disaster geomorphology.It is a scientific challenge to what extent the landform will develop into a disaster geomorphology.Manual interpretation remains the main approach for subaqueous geomorphology recognition,which limits the efficiency of subaqueous disaster geomorphology analysis.This article will carry out the research on automatic recognition of subaqueous disaster geomorphology,improve the speed of disaster early warning,and provide the basis for decision makers and planners to formulate corresponding risk prevention and control measures,which has great practical significance.In China,subaqueous disaster geomorphology is particularly common and prominent in the middle and lower reaches of the Yangtze River.Only in 1998,the length of bank collapse in the middle and lower reaches of the Yangtze River reached35.7 % of the total length of the shoreline.Bank stability and its protection has always been a hot issue in the field of river control engineering.Based on the assessment of the development trend of subaqueous disaster geomorphology in the middle and lower reaches of the Yangtze River,supported by advanced intelligent observation technology,this study innovatively introduces the terrestrial morphology element classification technology to automatically identify subaqueous disaster geomorphology,including bank collapse,scarp,scour pit and dunes.According to the spatial structure of morphology elements,the types of dunes are further identified to correspond to different disaster risk.The research results are as follows:(1)The quantitative recognition method of subaqueous disaster geomorphology is proposed.(a)By introducing the terrestrial morphology element classification method:Geomorphons,the morphological features of subaqueous disaster geomorphology are classified into 10 basic elements,including slope,ridge,valley,flat,spur,hollow,shoulder,footslope,peak and pit.Geomorphons method is improved by increasing the flatness of 0° to automatic recognition of bank collapse scar,scour pit and dunes.Specifically,when the flatness is 10°,the coverage area of slope element has a slope ratio of 1/2-1/3,close to the maximum stable slope ratio,which can be used to identify the subaqueous scarps potential to failure;the flat element coverage area is relatively gentle,which is used to identify the riverbed;the footslope element is distributed between the slope element and the flat element,which can be used to identify the subaqueous bank toe,regarded as the boundary between the riverbed and the scarp.When the flatness is 0°,the water depth of pit element is greater than the surrounding water depth,which can be used to identify the scour pit;the hollow element is located under the arc collapse,which is used to identify the subaqueous bank scar caused by bank collapse.When the flatness is less than 3°,the ridge and valley elements are located at the crest and trough of the dunes,respectively,which can be used to identify the subaqueous dunes.(b)Based on a large number of collected multibeam bathymetric data,it is demonstrated that when the minimum skip radius is set to 3 m,the interference of multibeam bathymetric point cloud noise on the recognition of subaqueous disaster geomorphology can be effectively suppressed without reducing the accuracy,thus significantly reducing the post-processing workload of field data acquisition,This is one of the key basic parameters to improve the timeliness of underwater disaster detection and recognition.(2)On the basis of the proposed quantitative recognition method of subaqueous disaster geomorphology,the tile similarity estimation method of spatial structure of morphology elements is introduced to realize the algorithm of subaqueous dunes location and type recognition.The algorithm is suitable for small-medium and largemedium scale,and catenary-bead dunes,and has good robustness under the multibeam measurement condition of insufficient lateral overlap.(a)The algorithm can be divided into three parts: tile segmentation,tile feature extraction and similarity calculation.The morphology element groups of connected grids are counted as tile features.Based on the Wave-Hedges similarity function,all tile features in the study area are compared with the tile features of the target type of dune,and the similarity index is calculated.(b)Compared with the Wood criteria and self organization map(SOM)technology,the improved Gomorphones method enriches the diversified description of the bed surface morphological features,and makes it clear that the improved Gomorphones method is the optimal subaqueous morphology element classification method,which can provide a concise morphology element classification map without area loss,It is suggested to creat the input layer for subaqueous dunes location and type recognition algorithm.(3)Based on the recognition method of subaqueous disaster geomorphology,the development of subaqueous disaster geomorphology in the bend section of Lower Jingjiang River in the middle reaches of the Yangtze River and the typical section with bank collapse from Zhenjiang to Yangzhong in the lower reaches of the Yangtze River are analyzed,and the subaqueous disaster geomorphology zoning map is drawn.(a)When the thalweg is close to the shore and the water depth is more than 11 m,it tends to form larger scour pits.Arc collapse tends to make the bank scarp move back and shorten,so as to improve the stability of the bank slope or reduce the scale of potential collapse.(b)The subaqueous disaster geomorphology in river system have superposition relation,corresponding to different roughness or texture.Relatively large flatness can filter relatively small roughness or texture and separate the scarp from the riverbed.To extract bank scars,scour pits and dunes from surface roughness and texture,a small flatness is used.(c)If the underwater bank slope is damaged with scar signs,the real-time monitoring of bank deformation should be carried out although the bank collapse has not detected above the water surface.To sum up,this study introduces the terrestrial landform recognition method,and improves it to adapt to subaqueous disaster geomorphology recognition.Two key flatness values of bank collapse,scarp,scour pit and dune are given to establish the recognition scheme between morphology elements and subaqueous disaster geomorphology;the minimum skip radius is given to filter the disturbance of outliers in bathymetric data;the subaqueous disaster geomorphology recognition and bank collapse risk assessment of Lower Jingjiang River in the middle reach Yangtze River and Zhenjiang-Yangzhong reach in the lower Yangtze River are completed;Based on the spatial structure of morphology elements,the technical process of identifying dune types from bedform features is proposed.The underwater disaster geomorphology recognition system established in this study can effectively locate the subaqueous scars caused by collapse,unstable scarps,erosion pits at the foot of slopes and dunes,which will provide guidance for decision-makers and planners to formulate targeted disaster risk prevention and control measures. |
PDF Full Text Request