Computational fluid dynamics to guide the search for victims of drowning in urban rivers

Computational fluid dynamics to guide the search for victims of drowning in urban rivers

© @ULiège

Worldwide, unintentional drowning results in nearly 400,000 fatalities annually, and drowning is one of the leading causes of death among young people. Climate change, by increasing the frequency and intensity of hydro-meteorological extremes, tends to elevate the risk of drowning. Extreme floods, for instance, lead to a significant number of drowning incidents. Similarly, heatwaves encourage more unauthorized bathing in urban rivers, contributing to a rise in accidental drownings. Additionally, urban planning policies that promote the use of riverbanks for leisure activities (cooler temperatures, scenic views, walking and cycling paths) make it timely to address the issue of drowning in urban rivers.

The causes of drowning in urban rivers are diverse, including swimming, traffic accidents, drug or alcohol use, and suicide. Compared to swimming pools or seaside locations, the chances of surviving a river drowning are two to three times lower. Reducing drowning fatalities requires a comprehensive approach, combining epidemiological studies, prevention efforts (such as education, raising risk awareness, surveillance or bathing bans), preparation for search and rescue, as well as advancements in resuscitation and medical treatment.

My research focuses on search operations, both for rescue and body recovery. Under certain conditions, resuscitation can be possible within a time frame of up to two hours. However, speeding up body recovery remains crucial beyond this period. Quick recovery allows victims' relatives to begin their mourning process sooner, facilitates police investigations by preserving the body, and helps search teams optimize resources and limit risks to first responders like divers.

Search operations in urban rivers are particularly challenging due to high flow velocity, low visibility, and the presence of subsurface vortices caused by obstacles like bridge piers and ship movement. Technologies like sonar are less effective in urban rivers due to debris on the riverbed, which complicates body detection. Currently, search operations in urban rivers are not guided by models that accurately simulate river flow.

The objective of my PhD research is to develop a mathematical and probabilistic model to predict the motion of a drowned body in an urban river using computational fluid dynamics. This model generates dynamic probability maps for locating a body, starting from the estimated drowning site and time, as well as specific victim characteristics (such as mass and clothing) and river conditions. These maps allow for more targeted search operations, aiming to reduce search times significantly. The model is based on general principles rather than the site-specific approaches used in current practice.

The project involves formulating new hypotheses about how a human body drifts in a river, derived from principles of physics and forensic knowledge. These hypotheses are translated into a mathematical model, with lab experiments helping to determine the necessary parameters. The model incorporates uncertainties (such as body decomposition and deformation) and is validated against real-life cases and field observations. Finally, strategies for applying the model in real-world situations are explored in collaboration with key stakeholders such as firefighters, civil protection teams, and police, alongside international academic partners.