<p>Urban morphology regulates surface flows and influences pluvial flooding. However, conventional models and studies lack a mechanistic, system-level explanation of how macroscopic building configuration governs water accumulation, due to fragmented representations of spatial structure and an inability to isolate morphological causality. To address these limitations, we examined the impact of urban morphology using three morphological parameters adapted from statistical physics and porous media theory: porosity (<i>P</i>), order parameter (<i>φ</i>), and feature distance (<i>d</i>). We used reverse Monte Carlo (RMC) simulation to generate urban morphology ensembles with controlled parameters. Each synthetic layout was evaluated with a high-precision hydrodynamic model, and we interpreted the resulting differences in water accumulation through three hydraulic mechanisms: anisotropy, hydraulic connectivity, and vorticity. The results reveal a porosity-related trade-off: its direct capacity to reduce water accumulation is strongly counteracted by an indirect, connectivity-mediated effect that increases accumulation, resulting in a weak net reduction. The order parameter (<i>φ</i>) primarily exacerbates water accumulation by intensifying rotational flow motion. We also identified nonlinear interactions: high <i>φ</i> combined with large <i>d</i> produces a synergistic increase in accumulation, whereas high <i>P</i> can buffer the adverse effect of high <i>φ</i>. This research demonstrates that urban water accumulation patterns emerge from complex, nonlinear interactions between building morphological parameters, and it provides a physics-grounded basis for designing flood-resilient urban layouts.</p>