This study introduces advanced super-resolution architectures, GISSRGAN and UGISSRGAN, designed to enhance satellite imagery resolution by integrating auxiliary geospatial data. The proposed models address challenges in texture reconstruction in remote sensing applications by introducing novel modifications to the ESRGAN framework, including Masked Residual-in-Residual Dense Blocks (MRRDBs) and a U-Net-based GIS Data Feature Block (UGISDFB). Experimental results demonstrate the effectiveness of these models in improving perceptual quality and reconstruction accuracy, particularly for complex foliage structures. GISSRGAN and UGISSRGAN leverage collateral data from datasets such as Landsat 8–9 and surface temperature measurements, achieving superior results compared to conventional methods. The study highlights the role of auxiliary data in refining network attention to surface-specific features, enabling improved environmental monitoring and natural resource management. Future work will explore further optimization and integration of diverse geospatial datasets to enhance model performance and scalability.