On July 4, 2025, unprecedented flash flooding in Texas Hill Country led to the deaths of at least 134 people, with 101 still missing. The extreme rainfall caused rapid water level rises and significant infrastructure damage. This study uses the Adaptive Hydraulics (AdH) model in its non-hydrostatic flow mode to simulate the flood's behavior.

Using high-resolution terrain data, radar rainfall observations, and stream gauge records, we are calibrating a two-dimensional AdH model for selected counties, including Kerr, Kendall, Travis, Burnet, and the Guadalupe River. We will compare hydrostatic and non-hydrostatic model runs to assess the significance of vertical accelerations on flood wave dynamics, flow velocity, and inundation extent.

Preliminary results indicate that the non-hydrostatic model more accurately captures critical flood features, such as steep wavefronts and peak depths in narrow valleys. This study emphasizes the importance of full vertical momentum dynamics in modeling flash floods and contributes to improving early warning systems and infrastructure risk assessments, enhancing flood resilience against climate change impacts.