Mapping vegetation biophysical and biochemical properties using unmanned aerial vehicles-acquired imagery

Abstract

ABSTRACTRemote-sensing imagery with different spatial resolutions has been widely applied to retrieve vegetation properties in various ecosystems. However, the spatial resolutions of most space-borne or air-borne images (e.g. 2 m) are not fine enough to investigate canopy-level (e.g. 0.5 m) properties, especially in heterogeneous ecosystems (e.g. grasslands, wetlands). Unmanned aerial vehicles (UAVs) as an emerging remote-sensing platform offer a unique ability to acquire imagery at very high spatial resolution (centimetres). This study utilizes an UAV to investigate vegetation properties in a heterogeneous grassland, focusing on estimating LAI and chlorophyll content and building digital elevation model (DEM) to support analysing spatial variations of these vegetation properties. The acquired images were mosaicked and geometrically and radiometrically corrected. LAI and chlorophyll content were then estimated using vegetation indices and textural variables. Results show LAI and chlorophyll content can be quantitatively retrieved (e.g. LAI from 0 to 8, chlorophyll content from 0 to 180 µg cm−2), with higher accuracy for imagery acquired in the middle growing season. Complex terrain in the study area highly influences the sprout and growth of vegetation. Methods discussed in this article can also be applied to study other ecosystems, such as agricultural fields and wetlands.