Alternate Wetting and Drying (AWD) is a water-saving irrigation practice that requires reliable monitoring of rice-field inundation. Although L-band Synthetic Aperture Radar (SAR) has shown promising performance for this purpose, existing studies largely emphasize aggregate classification accuracy, while the combined effects of crop, environmental, and observation conditions on individual field-date reliability remain insufficiently explored. This study proposes a reliability-aware AWD monitoring framework for rice paddies in Nueva Ecija, Philippines, using ALOS-2 PALSAR-2 SAR, supplementary optical observations, and ground measurements. Ultimately, the framework aims to provide initial field-level inundation classifications with calibrated reliability estimates from satellite observations, enabling users to determine when satellite-only monitoring is sufficient and when additional ground observations are needed to improve or verify classification outcomes and the reliability of their estimates. As an initial investigation, a pilot study assessed whether 30-m Harmonized Landsat–Sentinel-2 (HLS) observations reflect fine-scale rice-field characteristics. UAV-based Structure-from-Motion (SfM) photogrammetric point clouds from two rice-farm areas in Ryugasaki, Japan, across three dates in 2025 were compared with temporally matched HLS data. Statistical analyses tested associations between within-pixel surface-elevation variability and HLS spectral observations while accounting for rice-pixel purity and temporal variation. Significant associations were found only under specific farm and observation-date conditions, with sensitivity analyses supporting these findings. The results suggest that the sensitivity of HLS observations to fine-scale surface variability is condition-dependent, highlighting the need to evaluate their suitability as supplementary indicators under different field and observation conditions. Planned work will extend the pilot using LiDAR-derived structural measurements and develop and validate SAR-based inundation classifications and reliability estimates against ground observations in the Philippine study area.