The variable spatial-temporal process of ice disaster weather presents a significant challenge for developing preventive optimal dispatching schemes in novel power system, necessitating the incorporation of its uncertain influence into day-ahead preventive dispatching. In view of the potential for transmission line failure and the intermittent nature of renewable power output during ice disaster weather conditions, this paper proposes a distributionally robust preventive unit commitment method that incorporates uncertainties in both the failure rate of transmission lines and the prediction deviation of wind power, photovoltaic power as well as hydropower. The failure rate model of transmission lines under ice disaster weather is constructed by considering the spatial-temporal physical processes involved. Considering the uncertainty of transmission line failure rate under ice disaster weather, an ambiguity set for transmission line failure status and an uncertainty set for prediction deviations of wind-solar-hydro are constructed.A two-stage distributionally robust optimal dispatching model is constructed based on the ambiguity set, aiming to optimize the unit start-up/shut-down mode and line maintenance scheduling mode. And a column-and-constraint generation (C&CG) algorithm is employed to solve the optimal dispatching model considering the worst transmission line failure rate and wind-solar-water prediction deviation.The IEEE RTS-24 node system is used to verify the effectiveness of the proposed method of the proposed preventive dispatching model and solution method.