UAV-Based Hyperspectral Ultraviolet-Visible Interpolated Reflectance Images for Remote Sensing of Leaf Area Index

Abstract

Despite its relation to a number of environmental parameters, ultraviolet (UV) reflectance is rarely used in remote sensing. In this study, we investigate the applicability of UV-vis reflectance for vegetation monitoring with unmanned aerial vehicles (UAV). We measure point reflectance over the study area using a UAV-borne spectrometer, project the points onto the Earth's surface, and interpolate them to obtain continuous reflectance images. We use the leaf area index (LAI) to demonstrate the applicability of UV reflectance for vegetation monitoring. Our results show that the UAV reflectance images match the Sentinel-2 reflectance. Our validation shows that the inclusion of UV reflectance to the visible reflectance in LAI models leads to the $r^{2}$ increase of up to 29.2% and RMSE decrease of up to 18.9% in comparison to the LAI models using visible reflectance only. We have shown that measurement of UV reflectance is feasible in the 320–400 nm range using UAV remote sensing and that hyperspectral UV-vis reflectance imaging is useful for vegetation monitoring. Moreover, the obtained results lead us to believe that improvement of our measurement system, or conducting the experiments in a different location should make it possible to measure the reflectance at a wavelength of 290 nm. Finally, we discuss other potential applications of UV in remote sensing.

Keywords

• Interpolation
• hyperspectral imaging
• ultraviolet sources
• vegetation
• vegetation mapping