The increasing interdependence between transportation networks (TNs) and distribution networks (DNs) presents significant challenges for the analysis and risk assessment of coupled networks. This paper presents a risk assessment framework of coupled TN and DN (TDN) considering charging power restriction and uncertainty factors, simulating operations through a coordinated optimization model. The model incorporates charging power restrictions for fast charging stations (FCSs). To enhance computation efficiency, nonlinear constraints are linearized, and an accuracy-aware adaptive piecewise linearization method is utilized. Random variables within the coupled TDN are employed as inputs to develop a surrogate model based on the Gaussian process regression method, complemented by suitable kernel functions designed for the scenarios with numerous discrete categorical variables. A global sensitivity analysis is conducted to identify the continuous and discrete uncertainty factors that significantly impact system performance. The proposed risk assessment framework serves as a valuable tool for evaluating the coupled TDN and facilitating the identification of continuous and discrete uncertainty factors affecting network operation while revealing risks associated with failures in the operation of coupled TDN.