Abstract: Rapid and effective identification of damages and their locations in high-speed turnout switch rail bases holds significant importance for ensuring railway transportation safety. First, ultrasonic guided wave methods were employed to study switch rail damage detection, starting from guided wave mode shapes to solve for switch rail guided wave modes and dispersion curves, thereby determining excitation frequencies. Then, a 3D simulation model was established to verify modal analysis correctness, with different-sized damages simulated to ensure the solved modes´ detection capability. Finally, a high-low temperature test platform was built, using a self-developed handheld ultrasonic guided wave flaw detector to collect echo data, analyzing the influence of environmental temperature changes on ultrasonic guided waves and developing compensation algorithms, followed by simulated field tests. Results showed that after temperature compensation, damage localization errors significantly decreased, confirming that the proposed ultrasonic guided wave technology could meet the long-distance, rapid inspection requirements for turnout switch rail base detection.