Climate change disrupts the hydrological cycle of river basins, leading to irregularities in rainfall patterns that can cause devastating floods. A two-dimensional hydrodynamic numerical model, GPU (Graphical Processing Unit) Accelerated Surface Water Flow and Transport was employed to simulate the process of flood progression in a small watershed during periods of heavy rainfall. The objective of this study was to examine the flood characteristics of these basins. Hydraulic properties of typical cross-sections were calculated for rainfall return period of 50, 100, 500, and 1000 years, and a thorough assessment of flood vulnerability in the watershed was conducted. The results suggest that the rate of increase in the highest flood level is highest at the 500 year return period for all sections, reaching a value of 51.38%. However, during the span of the 1000 year return period, the growth rate undergoes a decrease to 7.61%. During extreme rainfall, the S1 cross-section exhibits the greatest water depth among the five selected cross-sections. The maximum water depth, ranging from 1.41 to 1.93 m for return periods between 50 and 1000 years, is 2.03-2.76 times that of the S3 cross-section. The magnitude of floods in the watershed remains relatively consistent despite a larger return period. To maximize the socio-economic benefits, it is recommended to determine the dimensions of new road and railway bridge culverts in this small watershed according to the flood peak that happens once every 500 years. The research findings serve as vital resources for examining the reaction of small watersheds to heavy rainfall and evaluating flood control technologies.