Hyperspectral UV-Vis Reflectance Imaging Using UAVs for Leaf Area Index Remote Sensing

Although ultraviolet (UV) reflectance is linked to various environmental factors, it remains underutilized in remote sensing applications. This study explores the potential of UV-visible (UV-vis) reflectance for vegetation monitoring using unmanned aerial vehicles (UAVs). A UAV-mounted spectrometer was employed to collect point reflectance data across the study area, which was then georeferenced and interpolated to produce continuous reflectance images. The leaf area index (LAI) was used to illustrate the effectiveness of UV reflectance in vegetation monitoring. Our findings indicate strong agreement between UAV-derived reflectance images and Sentinel-2 data. Validation revealed that incorporating UV reflectance into LAI models alongside visible reflectance resulted in an R² improvement of up to 29.2% and an RMSE reduction of up to 18.9%, compared to models using only visible reflectance. This study demonstrates that UV reflectance measurements in the 320–400 nm range are feasible with UAV-based remote sensing and that hyperspectral UV-vis reflectance imaging offers significant value for vegetation monitoring. Additionally, the results suggest that refining our measurement system or performing experiments in a different environment could enable reflectance measurements at wavelengths as low as 290 nm.