This study develops a stochastic inexact mixed-integer fractional optimization (SIMFO) model to enhance solid waste management (SWM) systems in isolated island communities. The model integrates inexact optimization, chance-constrained optimization, linear fractional programming, and mixed integer linear programming. It aims to maximize waste flow diversion from landfills, minimize system costs, and adhere to environmental emission caps. According to the analysis of a case study in British Columbia, Canada, by optimizing waste flows, implementing a reasonable facility expansion plan, and fully involving transfer stations, the island SWM system is expected to achieve a waste diversion rate of more than 75 %. In the future scenario, the daily amount of waste transported outside the island is reduced from 62 tonnes to zero, thereby reducing costs and environmental burdens. Compared with the present scenario (62.02 billion CO2e) and optimization programming focusing on cost reduction (69.29 billion CO2e), the upper limit result of the five-year cycle under the future scenario is 54.92 billion CO2e, representing reductions of 12 % and 21 %, respectively. The proposed SIMFO framework addresses uncertainties, optimizes facility capacity, and supports dynamic decision-making processes. This research offers a robust tool for policymakers, promoting sustainable SWM practices and long-term environmental stewardship in isolated island regions.