Abstract
A well-planned layout for origin-based cold storage is crucial for minimizing post-harvest losses, reducing costs, improving logistics efficiency, and mitigating environmental impacts. To address gaps in existing research on multi-technology coordination and trade-offs among siting objectives, this paper proposes an integrated multi-method framework that combines Geographic Information Systems (GIS), multi-criteria decision making (MCDM), and an improved Harris Hawks Optimization (IHHO) algorithm for facility siting optimization. We develop an evaluation system that incorporates logistics infrastructure, natural conditions, and agricultural development and use a GIS-MCDM model with spatial constraints to delineate highly suitable areas. K-medoids spatial clustering and an economies-of-scale cost model are then used, with IHHO determining the optimal number, locations, and capacities of facilities. A case study in Helan County, China, indicated that highly suitable zones are concentrated in the south, accounting for approximately 1.25% of the study area; nine candidate regions were identified, and six optimal sites were selected. Scenario analysis revealed that higher fixed construction costs favor larger facilities, while growing demand supports centralized, high-capacity cold stores rather than dispersed, smaller ones. Overall, the proposed framework provides a systematic tool for scientific planning and suitability assessment of on-farm cold-chain infrastructure, with the potential to enhance logistics efficiency, reduce postharvest losses, and promote sustainable agricultural development. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1038/s41598-026-40766-2.