Power systems are exposed to risks from various external events such as disasters and attacks. Hardening power lines prior to extreme events can effectively mitigate their impact on the system. For spatial power systems and events, a line hardening resource allocation model based on a tri-level optimization framework involving defender, event (or attacker), and operator roles for localized failures (or attacks) is proposed. The model accounts for defense costs associated with line length and calculates the failure probabilities of lines based on their distances from the failure center, employing Shannon’s information entropy theory to quantify this uncertainty. The tri-level model is transformed and solved using the Column-and-Constraint Generation (C&CG) method, thereby determining the optimal line hardening resource allocation strategy and identifying the corresponding worst-case scenario. The effectiveness of the proposed model is validated using the IEEE-24 bus system, and the results are analyzed. This model provides insights into enhancing the resilience of power systems against spatial extreme events.