Immediately after a fault occurs, the rate of change of frequency (RoCoF) at different nodes in the power grid varies, which is influenced not only by the fault location but also by the distribution of inertia across the system and the topology and parameters of the transmission networks. Consequently, even when evaluated based on the system’s center of inertia (COI), the post-fault RoCoF may still exceed safe thresholds. To address the challenge, this paper proposes an optimal unit commitment (UC) method considering the spatial distribution characteristics of inertia and RoCoF. First, a spatial distribution model of inertia-RoCoF is developed to analyze the variations of injected power and voltage phase angles at different nodes across a power system immediately after a contingency. Then, security constraints of nodal RoCoF are formulated and integrated into the optimal UC problem, considering contingencies. Finally, simulation results of the WSCC9-bus system and the Southeast Australia power system demonstrate that the proposed UC strategy offers significant advantages over traditional ones in maintaining the security of nodal RoCoF.