This paper presents a three-stage stochastic optimization model for the planning of networked microgrids (NMG) that incorporates considerations for seismic activity. Utilizing a combination of Monte Carlo Simulation (MCS) and K-Means clustering, the model effectively generates and reduces earthquake scenarios. An peak-ground acceleration based earthquake propagation model and utilizes fragility curves of critical infrastructures within power systems to develop a comprehensive component failure model are investigated. Monte Carlo simulations are conducted to determine the failure rates of utility transformers, distribution lines, and Distributed Energy Resources (DERs). The planning model is structured in three stages, each targeting specific aspects of microgrid planning. The first stage aims to minimize initial investment costs, the second stage focuses on reducing operational costs under various fault scenarios, and the third stage strives to minimize the unserved load during earthquake events. The model encompasses several investment and operational objectives, including the reduction of capital costs, CO2 emissions, load curtailment costs, and operational costs of DERs, while also considering power generation and flow constraints. The efficacy of the model is demonstrated through its application to a simplified distribution network in Tanjung KOTA on Lombok Island, a region prone to frequent earthquakes. The simulation results show that the proposed NMG not only achieves 100% electrification and significantly reduces CO2 emissions but also enhances the resilience of the power system. This contributes to facilitating a green transition, ensuring robust performance even under seismic challenges.