Reliable Tree-level Evapotranspiration Estimation of Pomegranate Trees Using Lysimeter and UAV Multispectral Imagery

Abstract

The accurate estimation and mapping of evapotranspiration (ET) are essential for crop water management. As one of the traditional ET estimation methods, crop coefficient (Kc) has been commonly used. Many studies indicated a linear regression relationship between the Kc curve and the vegetation index curve. The linear regression model is usually developed between the Kc and the normalized difference vegetation index (NDVI) derived from satellite imagery. The satellite images can provide temporally and spatially distributed measurements. However, multispectral satellite imagery’s spatial resolution is in the range of meters, which is often not enough for crops with clumped canopy structures, such as trees and vines. Little ET estimation has been studied based on the single-tree level. Thus, the purpose of this study was to develop a reliable tree-level ET estimation method using UAV high-resolution multispectral images. Compared with satellite imagery, the spatial resolution of UAV images can be as high as centimeter-level. A field study was conducted to investigate pomegranate trees at the USDA-ARS (US Department of Agriculture, Agricultural Research Service) San Joaquin Valley Agricultural Sciences Center in Parlier, California, USA. The NDVI map was derived from UAV imagery. The Kc values were calculated based on the actual ET from a weighing lysimeter and reference ET from the weather station. The authors then established a linear regression model between the NDVI and Kc to estimate the actual daily ET. Results showed that the linear regression model could estimate tree-level ET with an R2 and mean absolute error (MAE) of 0.9143 and 0.39 mm/day, respectively.