Rainfall and coastal water level relative timing as drivers of compound flooding in Haikou, China
Wu, G.; Xu, H.; Sun, L.; Lu, C.; Fowler, H.J.; Willems, P.; Wang, J.; Liu, Q.; Sheng, Y. (2026). Rainfall and coastal water level relative timing as drivers of compound flooding in Haikou, China. Journal of Hydrology: Regional Studies 66: 103635. https://dx.doi.org/10.1016/j.ejrh.2026.103635
In: Journal of Hydrology: Regional Studies. Elsevier B.V.: Amsterdam. ISSN 2214-5818; e-ISSN 2214-5818, more
|  |
| Author keywords |
Compound flooding;Time-lag scenarios;Combined effect;Delft3D FM;Hydrodynamic modeling;storm-surge;extreme rainfall;risk;dependence;hazard;city |
| Authors | | Top |
|
- Lu, C.
- Fowler, H.J.
- Willems, P., more
|
- Wang, J.
- Liu, Q.
- Sheng, Y.
|
| Abstract |
Study Region: Urban area of Haikou, Hainan Province, China. Study focus: Compound flooding in typhoon-prone coastal cities arises from the interaction between coastal water level and rainfall, but the role of their relative timing remains underexplored. This study proposes a framework that captures the relative timing for improved compound flooding assessment. Using the D-Flow FM model, we simulate 49 time-lag scenarios to quantify how timing (zt) between coastal total water level and rainfall modulate inundation. For each scenario, we evaluate inundation volume and compute absolute (zM) and relative (zR) increments relative to the baseline. We delineate compound zones and decompose contributions from water level, rainfall, and their interaction. New hydrological insights for the region: The results indicate that time-lags strongly shape compound flooding and produce nonlinear responses. Inundation is consistently larger when rainfall peaks before the coastal water level peak. A window from -19 h to -4 h yields volumes exceeding the historical event, and the maximum inundation occurs at -13 h with zR = 9.44% and (zM = 8.53 million m3). Water level is the dominant driver, yet within compound zones, the interaction term exceeds rainfall alone. These findings highlight that taking zt= 0 h as the worstcase can underestimate risks, emphasizing that the relative timing of rainfall and coastal water levels is a critical factor for managing compound flooding in coastal cities. |
|
