<p>In response to the complex risks that disasters caused by heavy rains bring to rammed earth ancient city site. This study proposes a heavy rain risk monitoring method based on digital twin technology, and uses the risk assessment model as an important twin model for intelligent prediction, aiming to improve the efficiency of risk identification, meet the needs of high-quality decision-making, and optimize the disaster management process. First, a user-friendly digital twin monitoring platform architecture is defined to include four layers: data layer, entity layer, model layer and function layer, which work together to enhance monitoring and prediction capabilities. Then, combining the dangers of disaster-causing factors and the vulnerability of the ancient city site itself, a risk assessment model was established to predict potential risks and threats in heavy rains. Finally, the effectiveness of this method was verified through practical application in the ancient city of Puzhou. This research not only provides a comprehensive management solution for ancient city site protection and disaster prevention, but also opens up a new path for digitally driven flood disaster risk assessment and prediction.</p>