Development of a Decision Support System for Irrigation Management During the Second Paddy Planting Season Using Sentinel-2 Imagery in Lamongan Regency, Indonesia

This study aims to develop a web-based Decision Support System (DSS) for irrigation management during the second paddy planting season in Lamongan Regency, Indonesia, utilizing Sentinel-2 satellite imagery and the Google Earth Engine (GEE) platform. The DSS integrates remote sensing data with meteorological information to estimate irrigation water requirements according to the paddy growth phases. The research workflow consists of three main components: classifying paddy growth phases using the Random Forest method, calculating irrigation water needs, and developing a user-friendly web-based graphical user interface (GUI). Sentinel-2 Level-2A imagery was processed to derive the Normalized Difference Vegetation Index (NDVI) and the Modified Normalized Difference Water Index (MNDWI). The Random Forest model demonstrated high accuracy, achieving 99% for training and 95% for validation, highlighting its effectiveness in capturing paddy growth dynamics. The GUI provides users with access to information on paddy growth phases, effective rainfall, crop water requirements, and irrigation needs. Despite its effectiveness, the study acknowledges certain limitations, such as cloud cover affecting satellite imagery, potential inaccuracies in manual validation data, and the absence of water discharge data at irrigation gates in Lamongan Regency. Future improvements include incorporating multi-source data, such as Synthetic Aperture Radar (SAR) from Sentinel-1, to mitigate cloud cover issues and upgrading irrigation infrastructure in the region. This research contributes to enhanced irrigation management, optimized water use, and the promotion of sustainable agricultural practices in paddy cultivation.