This study presents an automated and robust framework for multi-year processing of Landsat satellite imagery to generate seamless mosaics and derive key spectral indices for environmental monitoring. The approach dynamically selects applicable Landsat sensors and image collections based on acquisition year, differentiating between Landsat TM (pre-2012) and Landsat OLI (post-2012) datasets. The methodology prioritizes image selection near a target date in the dry season (February 15), applying a cloud cover threshold of 10% to minimize atmospheric interference and temporal inconsistency among path/row tiles. Using Google Earth Engine, the framework filters and scales surface reflectance bands, then identifies unique Landsat path/row tiles within a designated study area. For each tile, the optimal image with the lowest cloud cover nearest to the target date is selected to ensure high-quality mosaics. The mosaicked reflectance images are clipped to the study area boundary, and a corresponding cloud quality assurance mosaic is generated for validation. Subsequently, several spectral indices, including NDVI, SAVI, MSAVI, NDWI, and MNDWI, are calculated from the mosaics using band-specific expressions suitable for each sensor type. The system exports these indices alongside all surface reflectance and quality assurance bands with metadata-rich filenames that include satellite type, path/row identifiers, acquisition date, and cloud cover, enhancing traceability and reproducibility. Demonstrated on selected years (1988, 2009, and 2025), this automated workflow provides a scalable, efficient means for environmental change detection, vegetation health monitoring, and surface water mapping over multi-decadal timescales. The approach facilitates consistent, high-quality Landsat data utilization for long-term ecological studies and supports decision-making in climate impact assessment and land management strategies.