Spatial estimation of actual evapotranspiration over irrigated turfgrass using sUAS thermal and multispectral imagery and TSEB model

Authors: MEZA, K - Utah State University; TORRES, A - Utah State University; HIPPS, L - Utah State University; Kustas, William - Bill; GAO, R - Utah State University; CHRISTIANSE, L - Utah State University; KOPP, K - Utah State University; NIETO, H - Institute Of Agricultural Sciences; BURCHARD-LEVINE, V - The Institute Of Agricultural Sciences; MARTIN, M - The Institute Of Agricultural Sciences; COOPMANS, C - Utah State University; GOWING, I - Utah State University

Submitted to: Irrigation Science
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 10/31/2023
Publication Date: 12/6/2023
Citation: Meza, K., Torres, A., Hipps, L.E., Kustas, W.P., Gao, R., Christianse, L., Kopp, K., Nieto, H., Burchard-Levine, V., Martin, M., Coopmans, C., Gowing, I. 2023. Spatial estimation of actual evapotranspiration over irrigated turfgrass using sUAS thermal and multispectral imagery and TSEB model. Irrigation Science. https://doi.org/10.1007/s00271-023-00899-y.
DOI: https://doi.org/10.1007/s00271-023-00899-y

Interpretive Summary: Urban green surfaces are a significant part of urban landscapes and include golf courses, parks, athletic fields, home lawns, and other urban green spaces. In arid regions, urban turfgrass is negatively impacting water resources due to increasing water scarcity caused by human demands and limited water supplies. To better manage urban turfgrass water use or actual evapotranspiration (ETa) for landscape trees and turfgrass in arid regions needs to be quantified. This was achieved using high spatial resolution multispectral and thermal imagery data acquired from small Unmanned Aircraft Systems (sUAS) to quantify ETa using the two-source energy balance (TSEB) model. Comparison of TSEB model output of ETa to observations supports the capability of the TSEB model using sUAS imagery to estimate reliable spatial and temporal variation of daily ETa for an urban turfgrass surface. This is critical information for landscape irrigation management, particularly in arid environments, where routine ETa monitoring is not feasible with any other technology.

Technical Abstract: Green urban areas are increasingly affected by water scarcity and climate change. The combination of warmer temperatures and increasing drought poses substantial challenges for water management of urban landscapes in the western U.S. A key component for water management, actual evapotranspiration (ETa) for landscape trees and turfgrass in arid regions is poorly documented as most rigorous evapotranspiration (ET) studies have focused on natural or agricultural areas. ET is a complex and non-linear process, and especially difficult to measure and estimate in urban landscapes due to the large spatial variability in land cover/land use. Therefore, to understand water consumption processes in these landscapes, efforts using standard measurement techniques, such as the eddy covariance (EC) method as well as ET remote sensing-based modeling are necessary. While previous studies have evaluated the performance of the remote sensing-based two-source energy balance (TSEB) in natural and agricultural landscapes, the validation of this model in urban turfgrass remains unattended. In this study, EC flux measurements and hourly flux footprint models were used to validate the energy fluxes from the TSEB model in green urban areas at the golf course near Roy, Utah, USA. High-spatial resolution multispectral and thermal imagery data at 5.4 cm were acquired from small Unmanned Aircraft Systems (sUAS) to model hourly ETa. A protocol to measure and estimate leaf area index (LAI) in turfgrass was developed using an empirical relationship between spectral vegetation indices (SVI) and observed LAI, which was used as an input variable within the TSEB model. Additionally, factors such as sUAS flight time, shadows, and thermal band calibration were assessed for the creation of TSEB model inputs. The TSEB model was executed for five datasets collected in 2021 and 2022, and its performance was compared against EC measurements. Ultimately, an extrapolation technique based on incident solar radiation was used to compute daily ETa from the hourly remotely-sensed UAS ET. A daily flux footprint and measured ETa were used to validate the daily extrapolation technique. Results showed that the average of corrected daily ETa values in summer ranged from about 4.6 mm to 5.9 mm in 2021 and 2022. The Near Infrared (NIR) and Red Edge-based SVI derived from sUAS imagery were strongly related to LAI in turfgrass, with the highest R2 (0.76-0.84) and the lowest RMSE (0.5-0.6). The TSEB’s latent and sensible heat flux retrievals were accurate with an RMSE 50.5 W m-2 and 34.6 W m-2 respectively compared to EC closed energy balance. The expected RMSE of the upscaled TSEB daily ET estimates across the turfgrass is below 0.6 mm day-1. This study highlights the ability of the TSEB model using sUAS imagery to estimate the spatial variation of daily actual ET for an urban turfgrass surface, which is useful for landscape irrigation management in drought conditions.