Optimizing environmental performance of rice production in semi-arid Iran: An Life Cycle Assessment (LCA)-Multi-Objective Particle Swarm Optimization (MOPSO) approach

Rice production in water-scarce regions faces severe environmental and resource efficiency challenges that threaten long-term sustainability. This study introduces a novel integrated framework combining Life Cycle Assessment (LCA) with Multi-Objective Particle Swarm Optimization (MOPSO) to simultaneously minimize environmental impacts and optimize input use without compromising yield. A cradle-to-mill gate LCA was performed in a semi-arid region of Iran using the CML 2 -baseline method across ten environmental categories. Results showed that electricity (34 %), diesel fuel (16 %) and irrigation equipment (14 %) were the dominant contributors to overall environmental burdens, particularly to global warming potential and fossil resource depletion. Through MOPSO-based optimization, total environmental damage was reduced by 48 %, while rice yield increased substantially (from 6299 to 10,422 kg ha⁻¹). Although the energy ratio declined from 1.13 to 0.47, the integrated model effectively demonstrated trade-offs between productivity and sustainability. These findings provide a quantitative decision-support framework for policymakers and environmental managers to promote technology-driven, resource-efficient, and climate-resilient rice farming systems in arid and semi-arid regions.