Quantum-Enhanced Artificial Intelligence Models for Sustainable Agriculture and Climate-Resilient Crop Production

Abstract: This study investigates quantum-enhanced artificial intelligence (QEAI) models for climate-resilient crop production and sustainability-oriented decision support in agriculture. The framework integrates agroclimatic signals (rainfall, temperature, evapotranspiration, heatwave and dry-spell indicators, and vapor-pressure deficit) with sensor-derived crop-state variables (soil moisture, canopy temperature, and vegetation indices) to estimate yield and to identify robust management levers. Quantum enhancement is implemented through a quantum-ready optimization layer that encodes feature selection as a Quadratic Unconstrained Binary Optimization (QUBO) objective with explicit parsimony control. An applied empirical workflow is demonstrated using a realistic district–crop–year panel (2015–2024) calibrated to plausible ranges encountered in operational agroclimatic monitoring and on-farm sensing. The QEAI model achieves MAE=484.3 kg/ha and R²=0.661 on held-out data while using 5 features, compared with MAE=486.2 kg/ha and R²=0.647 for a baseline AI model using all 11 features. Scenario-based stress testing indicates systematic yield vulnerability under combined warming and precipitation shortfall, highlighting adaptation priorities for water management and microclimate buffering. Keywords: quantum-enhanced AI; sustainable agriculture; climate resilience; agroclimatic data; sensor data; soil moisture; canopy temperature; QUBO optimization.