Assessing the Efficacy of Unmanned Aerial Vehicles (UAVs) in Monitoring Crop Evapotranspiration within a Heterogeneous Soil.

Evapotranspiration (ET) estimates have been made across entire management fields using airborne and satellite remote sensing methods. However, airborne-based ET is not cost-effective, and satellite-based ET provides insufficient spatial/temporal information needed for proper irrigation management. Problems with remotely sensed ET are furtherly compounded in the presence of soil spatial heterogeneity. Unlike airborne and satellite-based ET, Unmanned Aerial Vehicles (UAV) can serve as a cost-effective method towards estimating ET and provide adequate spatial and temporal detail. However, UAV approaches towards estimating ET are rare, particularly within the presence of soil spatial variability. The overall goal of this project was to assess the capability of UAVs in modeling cotton-based ET across a soil heterogeneous landscape. Within a 27-ha field in Central Texas, low-altitude (120-m) UAV surveys were conducted throughout the cotton growing season over two soil types. UAVs were equipped with thermal and multispectral cameras to derive canopy temperature and NDVI, respectively. UAV data were supplemented by ground-truth measurements such as Leaf Area Index (LAI), plant height and plant/soil ground temperature. Both remote sensing and ground-truth parameters were used to model and validate UAV-based ET using a Two-Source Energy Balance (TSEB) model. Within each soil type, UAV latent heat (LE) fluxes (i.e. ET) and other energy balance fluxes (Rn, G, H) were compared with eddy covariance tower fluxes. UAV-based LE fluxes were also compared with manned airborne and satellite (Landsat)-based LE fluxes collected at or near the time of the UAV surveys.