Network partition of large-scale multimodal traffic networks is critical for coordinated multimodal traffic management. Existing methods addressed different modes separately and thus overlooked their interactions. This study presents a joint spectral partitioning method for road and tramway networks tram disturbances. The network is modeled as a two-layer graph, of which road and tramway segments are represented as vertices and intersections as edges. This allows the capturing of the abstract structure of different layers through edge weights, which evaluate similarities between vertices based on distinct features of road and tramway traffic evolutions. The network is decomposed based on the joint spectrum accommodating both layers, which is similar to the road traffic layer spectrum and enables to represent the tramway layer. The proposed method is verified by numerical experiments on a synthetic grid network with diverse road traffic distributions and a real network of Dresden city. Results indicate the proposed method is effective in preserving the homogeneity of both road and tramway traffic operations within sub-networks, while reducing the complexity of tram rescheduling.