Skip to main content
ARS Home » Northeast Area » Beltsville, Maryland (BARC) » Beltsville Agricultural Research Center » Hydrology and Remote Sensing Laboratory » Research » Publications at this Location » Publication #389391

Research Project: Integrating Remote Sensing, Measurements and Modeling for Multi-Scale Assessment of Water Availability, Use, and Quality in Agroecosystems

Location: Hydrology and Remote Sensing Laboratory

Title: Downscaling UAV land surface temperature using a coupled wavelet-machine learning-optimization algorithm and its impact on evapotranspiration

item ABOUTALEBI, M. - E & J Gallo Winery
item TORRES, A. - Utah State University
item MCKEE, M. - Utah State University
item Kustas, William - Bill
item NIETO, H. - University Of Alcala
item ALSINA, M. - E & J Gallo Winery
item White, William - Alex
item Prueger, John
item McKee, Lynn
item Alfieri, Joseph
item HIPPS, L.E. - Utah State University
item COOPMANS, C. - Utah State University
item SANCHEZ, L. - E & J Gallo Winery
item DOKOOZLIAN, N. - E & J Gallo Winery

Submitted to: Irrigation Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 4/22/2022
Publication Date: 6/27/2022
Citation: Aboutalebi, M., Torres, A., McKee, M., Kustas, W.P., Nieto, H., Alsina, M., White, W.A., Prueger, J.H., McKee, L.G., Alfieri, J.G., Hipps, L., Coopmans, C., Sanchez, L., Dokoozlian, N. 2022. Downscaling UAV land surface temperature using a coupled wavelet-machine learning-optimization algorithm and its impact on evapotranspiration. Irrigation Science. 40:553-574.

Interpretive Summary: Monitoring evapotranspiration (ET) or crop water use is essential for improving irrigation water use efficiency and detecting crop stress. Land surface temperature (LST) measured by satellites and unmanned aerial vehicles (UAVs) have shown great potential in mapping ET and plant stress using the Two-Source Energy Balance (TSEB) model. Higher resolution LST data has been shown to produce more accurate ET over vineyards with the TSEB model. This study uses a coupled wavelet, machine learning, and optimization algorithm for downscaling LST imagery from 60 cm to 15 cm to more accurately discriminate temperatures. Application of downscaled LST values into the TSEB model resulted in more accurate estimates of ET when compared to flux tower measurements. This downscaling technique has potential to provide more accurate high resolution ET maps that can be used to validate courser resolution satellite ET products over vineyards and other complex crop canopies.

Technical Abstract: Monitoring evapotranspiration (ET) is possible through land surface temperature (LST) measured by satellites and unmanned aerial vehicles (UAV). The assumption that the higher resolution of LST may improve the performance of remote sensing ET models was verified in a recently published article showing that higher resolution LST led to increased performance of the Two-source Energy Balance Model (TSEB) one of the well-known ET models. However, because of technology limitations, the spatial resolutions of satellites and UAVs in thermal wavelengths are coarser than those in optical and near-infrared (NIR) bands. Therefore, developing thermal sharpening techniques and assessing their impacts on ET models performance are imperative. Although previous studies have developed and evaluated downscaling LST methods for satellite imagery, implementation of those methods on UAV imagery is limited. In this study, a coupled wavelet, machine learning, and optimization algorithm was implemented for downscaling UAV thermal imagery from 60 cm to UAV optical imagery (15 cm) because 60 cm pixel resolution still incorporate mixed temperatures from the soil, vine canopy, active cover crop and shaded regions. A 2D discrete wavelet transform (2-D DWT) was employed for the decomposition of inputs to 60 cm and inverse transformation of low thermal resolution to higher resolution. Four machine-learning based algorithms (Decision Tree Regression (DTR), Ensemble Decision Tree (DTER), Support Vector Machine (SVM), and Gaussian process regression (GPR)) along with four linear regression-based models (linear, interactions linear, robust linear and stepwise linear) are used as the potential fitting models, and a grid search algorithm is used for auto tuning parameters of the machine learning algorithms. Additionally, a novel sampling technique was designed to provide more representative samples for training steps in the regressing models. Four sets of high-resolution images were provided by the Utah State University AggieAir sUAS Program as part of the ARS-USDA GRAPEX Project collected since 2014 over multiple vineyards located in California. After applying the proposed downscaling algorithm, a separation method was used for estimation of canopy and soil temperatures from the original and sharpened thermal imagery. Ultimately, the TSEB model was executed for these pairs of temperature components, and its performance compared to eddy covariance measurements. Results demonstrated that the proposed sampling algorithm can significantly accelerate the computation time for the UAV temperature sharpening efforts. Among all the fitting models, GPR, SVM and DTER were the most accurate in terms of R-square. The correlation between NDVI and radiometric temperature (Tr) was significantly improved when the downscaled Tr (DTr) was used in the NDVI-Tr domain for the separation procedure. Compared to additional IRT sensors temperatures, Tc and particularly Ts derived from the DTr were closer to the observed measurements. After feeding the TSEB model with DTr products, results demonstrated that estimations of soil heat flux (G) were significantly improved, while large LE differences were reduced.