Developing Innovation in Sustainability, Resilience, and Adaptation (ISRA) metrics for agricultural innovations in western Kenya

Climate variability and increasing environmental pressures boost the need for reliable metrics to assess agricultural innovations that enhance sustainability, resilience, and adaptation in smallholder farming systems. This study develops and applies the Innovation in Sustainability, Resilience, and Adaptation (ISRA) metrics to support evidence-based evaluation of climate-smart agricultural interventions in Western Kenya. Using a systematic review and scoring framework, we identified and assessed potential metrics derived from remote sensing, modelling, and socio-economic indicators to determine their relevance, scalability, and alignment with climate-smart agricultural goals. The analysis focused on three innovation areas: targeted intercropping for resilience, climate-smart irrigation management, and soil health improvement, selected through stakeholder consultations and technical feasibility assessments. For intercropping, satellite-based crop classification methods utilising Sentinel imagery and machine learning models were employed to distinguish between monocrop and intercropped maize systems, while drought monitoring involved Vegetation Indices such as the Vegetation Condition Index (VCI). Results show moderate classification accuracy and indicate that intercropped systems consistently maintain higher vegetation health during drought periods compared to monocrop systems. Concerning irrigation, a composite drought-monitoring framework integrating vegetation, temperature, precipitation, and soil moisture indicators was developed to support climate-smart irrigation advisories. For soil health, spatial nutrient gap analysis and soil fertility monitoring were proposed to guide targeted nutrient management interventions. Across these innovations, yield estimation emerged as a cross-cutting metric linking agronomic performance to resilience and adaptation outcomes. The findings highlight the potential of integrating Earth observation (EO) data, machine learning techniques, and socio-economic indicators to produce scalable metrics that support climate-resilient agricultural decision-making. The ISRA framework provides a practical approach to monitoring and evaluating agricultural innovations across levels, from farm management to regional policy planning, thereby strengthening evidence-based strategies for climate-smart agriculture in smallholder systems.