With the introduction of many new types of remotely sensed imagery in recent years, opportunities to compare and integrate these data abound. In this paper, we present a hue-saturation-value (HSV) transformation approach for the fusion of surface roughness and shallow subsurface information from Airborne Synthetic Aperture Radar (AIRSAR) data with surface spectral reflectivity from Landsat 7 Enhanced Thematic Mapper Plus (ETM+) data. The fused image is informative and brings out new features that are not evident in the original images, and it also helps to identify many features that are not clear in the original images. In this paper, we also present a combined principal components analysis (CPCA) approach to effectively remove the banding noise that degrades the radar image quality. We examine the differences in surface roughness, texture, and penetration ability at different radar frequencies, the differences between radar and optical (ETM+) images, and the utility of radar images in mapping morphologically defined structures such as fault scarps. The earthquake-related active faults in the Franklin Mountains and Valentine region in West Texas are presented as examples of successful applications of data fusion.