Validating UAS-based NDVI Data with Satellite Landsat Imagery for Bald Eagle Habitat Prediction in the Del Rio Springs Ecosystem

Vegetation health is commonly assessed using the Normalized Difference Vegetation Index (NDVI), which can be derived from multispectral sensors operating at different spatial resolutions. Validating NDVI products across sensor platforms is essential to determine their reliability for environmental monitoring and habitat assessment. This research compares NDVI derived from moderate-resolution satellite imagery and high-resolution unmanned aircraft system (UAS) imagery collected over the same study area. Landsat imagery, provided through the joint USGS–NASA mission, was used to represent satellite-based vegetation patterns, while high-resolution multispectral data were acquired using a MicaSense sensor mounted on a UAS to capture fine-scale vegetation detail. NDVI values from both datasets were evaluated for spatial correspondence using the coefficient of determination (R²) to quantify the proportion of satellite-derived NDVI variability explained by UAS-based observations. To further explore spatial vegetation structure at finer scales, a Getis-Ord Gi* hotspot analysis was applied to the MicaSense NDVI data to identify statistically significant vegetation clusters associated with potential bald eagle nesting and foraging habitats. This approach demonstrates the value of high-resolution UAS imagery for detecting localized vegetation patterns not fully resolved by satellite data. Future work will incorporate atmospherically corrected Level-2 Landsat products and coordinate additional UAS data collections with Landsat overpass times to minimize illumination and viewing inconsistencies. Multi-season UAS imagery will also be acquired to assess temporal vegetation stress and change. These enhancements will improve cross-platform NDVI validation and strengthen the application of multisensor geospatial data for habitat prediction, landscape analysis, and long-term ecological monitoring.