Assessing spatial suitability and climate resilience of agroforestry systems in the Indian Himalaya of Uttarakhand using remote sensing and artificial neural networks

Agroforestry systems are widely recognized for their capacity to enhance ecological sustainability, livelihood security, and climate resilience in mountain environments. However, their spatial suitability and long-term resilience under changing climatic conditions remain poorly quantified in the Indian Himalaya. This study assesses the spatial suitability and climate resilience of agroforestry systems in Uttarakhand using an integrated framework of remote sensing, geographic information systems (GIS), and artificial neural networks (ANNs). Landsat 8 OLI imagery (30 m resolution) was classified using a supervised Gaussian Maximum Likelihood Classifier (MLC), while land-use categories were defined based on a modified Anderson Level I/II classification scheme to represent the heterogeneous Himalayan landscape. Field surveys conducted across representative agroecological zones documented the composition of tree and crop species within prevailing agroforestry systems. A multi-criteria land suitability analysis, guided by FAO (Food and Agriculture Organization of the United Nations) principles, was implemented using key biophysical variables including altitude, slope, aspect, Normalized Difference Vegetation Index (NDVI), soil properties, temperature, and precipitation. Future agroforestry patterns were simulated using an ANN model under RCP 4.5 climate scenarios. The model demonstrated strong predictive performance, indicating reliable simulation of agroforestry distribution. Results show that mid-altitudinal zones offer the highest suitability for agroforestry expansion, whereas small and fragmented systems exhibit reduced climate resilience under future projections. The findings provide a spatially explicit framework to support climate-resilient agroforestry planning and policy formulation in the Himalayan region.