Innovative Data Integration for Resilient Irrigation Planning and Food Security: A Remote Sensing and AI-Based Approach
Conference
10th International Conference on Agricultural Statistics
Format: CPS - ICAS 2026 75
Keywords: agricultural statistics, digital, food security, generative ai, geospatial, policy, remote sensing, resilience
Abstract
Ensuring resilient agri-food systems requires integrated environmental monitoring, digital innovation, and policy-relevant statistics. In East Java—Indonesia’s largest rice-producing province with over 9.27 million tons of rice production in 2024—drought remains a major challenge, particularly for the 488,000 hectares of rainfed rice fields. During the 2024 dry season, more than 31,000 hectares were affected by drought, including nearly 8,000 hectares with complete crop failure. This study develops an integrated framework combining remote sensing, geospatial analytics, Input–Output (I/O) modeling, and Generative AI through the development of AgriPolicy AI for evidence-based irrigation planning. A Water Deficit Risk Score was constructed using six satellite-derived indicators: precipitation, land surface temperature, soil moisture, evapotranspiration, NDVI, and Solar-Induced Fluorescence (SIF). Using entropy gain weighting at 500-meter resolution, the analysis captures spatial and seasonal drought patterns across Malang, Lumajang, and Jember. Results show that drought risk intensifies during the dry planting season, especially in upland and rain-shadow areas. By integrating drought risk, water availability, and irrigation proximity, the study calculates a Priority Irrigation Rehabilitation Score to identify intervention areas. Economic simulations further indicate that irrigation rehabilitation can increase agricultural output, household income, and rural employment. The framework demonstrates how remote sensing, official statistics, and AI can support resilient irrigation planning and agricultural policymaking in climate-vulnerable regions.