Abstract: Typhoons are among the most destructive natural hazards threatening road infrastructure in coastal regions, and assessing typhoon-induced road damage is essential for effective disaster risk reduction． Taking Hainan Island as a case study, this paper models the impacts of typhoon-induced damage on road infrastructure． Focusing on Typhoons Kalmaegi and Sarika, we employed linear, quadratic, power, exponential, and logistic functions to model the relationship between rainfall hazard indicators (including maximum daily rainfall, average daily rainfall, and total event rainfall) and historical loss data, aiming to identify the most representative indicators for constructing a loss assessment model． To further validate the model's reliability, three additional typhoon events (Rammasun, Parma, and Dianmu) are used． The results indicate that maximum daily rainfall is the most critical factor influencing road flood damage, exhibiting the highest correlation with road losses (correlation coefficient is 0.17)． Among the tested models, the logistic function provides the best fit for simulating road losses, with prediction errors ranging from 6.77% to 50.22% across the three validation typhoon cases． Overall, this study provides a representative case for regional road loss assessment under tropical cyclones and offers scientific support for enhancing the resilience of road infrastructure to typhoon disasters in Hainan Province．