识别河岸侵蚀的风险是确保早期预警和预防或控制包括长江在内的河流集水区崩岸的一项重要挑战。本文引入一种基于地貌要素的算法，通过调整多波束回波探测数据的平坦度来提取河岸侵蚀信息。该算法绘制了10 个水下地貌形态要素，包括坡、坡脚、水平面、脊、顶、谷、凹陷、凸起、坑和坡肩。利用21 个平坦度值构建水下河岸侵蚀特征的识别策略。结果表明：当平坦度为10°时，作为侵蚀载体的岸坡陡坎被坡要素覆盖。平坦度为0°时，冲刷坑和河岸破损由河岸附近的坑要素和岸坡中的凹陷表示。河道水下沙波是加速河岸侵蚀的重要因素，尤其是靠近河岸坡脚的沙波；沙波的临界平坦度为3°。分析了水下地貌形
态要素的分布，并用于绘制河岸侵蚀库存图。分析结果表明，近岸区水深较大，易形成较大的冲刷坑和较长的河岸陡坎。窝崩往往发生在较长的岸坡处，以缩短其长度。经讨论陆地、海洋和河流环境中平坦度值的不同设置，得出的结论是，多样化的平坦度值能够显著识别河流水下地貌形态。因此，本研究为基于平坦度的河流水下地貌形态要素识别提供参考，增强了从大量多波束测深数据定位水下岸坡失稳迹象与风险的能力。

Recognizing the risk of fluvial bank erosion is an important challenge to ensure the early warning and prevention or control of bank collapse in river catchments, including in the Yangtze River. This study introduces a geomorphons-based algorithm to extract river bank erosion information by adjusting the flatness from multibeam echo-sounding data. The algorithm maps 10 subaqueous morphological elements, including the slope, footslope, flat, ridge, peak, valley, pit, spur, hollow, and shoulder. Twenty-one flatness values were used to build an interpretation strategy for the subaqueous features of riverbank erosion. The results show that the bank scarp, which is the erosion carrier, is covered by slope cells when the flatness is 10°. The scour pits and bank scars are indicated by pit cells near the bank and hollow cells in the bank slope at a flatness of 0°. Fluvial subaqueous dunes are considered an important factor accelerating bank erosion, particularly those near the bank toe; the critical flatness of the dunes was evaluated as 3°. The distribution of subaqueous morphological elements was analyzed and used to map the bank erosion inventory. The analysis results revealed that the near-bank zone, with a relatively large water depth, is prone to form large scour pits and a long bank scarp. Arc collapse tends to occur at the long bank scarp to shorten its length. The varied assignment of flatness values among terrestrial, marine, and fluvial environments is discussed, concluding that diversified flatness values significantly enable fluvial subaqueous morphology recognition. Consequently, this study provides a reference for the flatness-based recognition of fluvial morphological elements and enhances the targeting of subaqueous signs and risks of bank failure with a range of multibeam bathymetric data.