Modeling the agricultural suitability of key crops under climate change scenarios in british columbia using spatial machine learning

This study evaluates the projected future suitability of agricultural land for field-grown crops in the Canadian province of British Columbia (BC), using the Intergovernmental Panel on Climate Change’s (IPCC) Shared Socioeconomic Pathways (SSP) scenario framework, namely SSP2, SSP3, and SSP5. Seven crops were selected for this study based on their economic and dietary importance to the province and the potential for growing these crops indoors using controlled environment agriculture methods: cabbage, cauliflower, lettuce, strawberry, kale, broccoli, and celery. Machine learning models were developed and employed to spatially assess current modelled crop suitability and project potential future suitability through 2100. Results indicate that cabbage, cauliflower, and lettuce are projected to experience northward expansion of areas with relatively high modelled suitability. Strawberry, kale, and broccoli modeling exhibit expansions of cultivation ranges under moderate warming scenarios, while high-warming scenarios are projected to result in losses of areas with relatively high modelled suitability in the latter half of the 21st Celery modelled suitability is projected to decline consistently across all scenarios. The modeling framework developed in this study offers a replicable and adaptable method for combining geospatial, climatic, and economic datasets to generate spatially explicit agricultural predictions over large geographic regions. Further application of the framework could enhance agricultural planning and policymaking across Canada and in other regions to make progress toward sustainable and resilient food systems in the face of changing climate.