Extreme flood events often lead to heavy casualties, with flood risk to humans varying with the flow conditions and the body attributes. Therefore, it is important to propose an appropriate criterion for the stability of a human body in floodwaters in the form of an incipient velocity. In this study, two formulae for the incipient velocity of a human body for sliding and toppling instability were derived, based on a mechanics-based analysis, and with both formulae accounting for the effect of body buoyancy and the influence of a non-uniform upstream velocity profile acting on the flooded human body. More than 50 tests were conducted in a flume to obtain the conditions of water depth and velocity at instability for a model human body, with the experimental data being used to calibrate two parameters in the derived formulae. Finally, the proposed formulae were validated in detail against existing experimental data for real human subjects, with different stability thresholds being obtained for children and adults in terms of assessing their stability related to floodwaters.

Significant channel adjustments have occurred in the Jingjiang Reach of the Middle Yangtze River, because of the operation of the Three Gorges Project (TGP). The Jingjiang Reach is selected as the study area, covering the Upper Jingjiang Reach (UJR) and Lower Jingjiang Reach (LJR). The reach-scale bankfull channel dimensions in the study reach were calculated annually from 2002 to 2013 by means of a reach-averaged approach and surveyed post-flood profiles at 171 sections. We find from the calculated results that: the reach-scale bankfull widths changed slightly in the UJR and LJR, with the corresponding depths increasing by 1.6 m and 1.0 m; the channel adjustments occurred mainly with respect to bankfull depth because of the construction of large-scale bank revetment works, although there were significant bank erosion processes in local regions without the bank protection engineering. The reach-scale bankfull dimensions in the UJR and LJR generally responded to the previous five-year average fluvial erosion intensity during flood seasons, with higher correlations being obtained for the depth and cross-sectional area. It is concluded that these dynamic adjustments of the channel geometry are a direct result of recent human activities such as the TGP operation.