Evaluation of a Two-Source Model to Estimate Vineyard Evapotranspiration Using High-Resolution Thermal Images Acquired by an Unmanned Aerial Vehicle (UAV)

Introduction

Traditionally, vineyard evapotranspiration (ETv = ETo*Kc)) is computed using the reference evapotranspiration (ETo) and crop coefficients (Kc). However, in most cases, the values of Kc used are empirical and do not adapt to local vineyard conditions. In this regard, several researchers have indicated that the spatial variability of ETv and Kc can be evaluated using thermal infrared (TIR) sensors placed on satellite and unmanned aerial vehicle (UAV) platforms. However, the practical application of satellite platforms for site-specific irrigation management is limited by the frequency of the satellite's overpass, cloudiness, and resolution of the image (each pixel covers 30 m x 30 m). As an alternative, the UAV platform equipped with high-resolution TIR cameras has been suggested as a tool to evaluate the intra-vineyard spatial variability of ETv and Kc. Thus, the objective of this study was to evaluate the Shuttleworth and Wallace (SW) model to estimate ETv using UAV-TIR images and meteorological data. In addition, net radiation (Rni) and soil heat fluxes (Gi) were calibrated to estimate the available energy (A) from the soil surface and complete canopy.

MATERIALS AND METHODS

This study was carried out during the 2018/19 and 2019/20 growing seasons, over a vineyard (cv Cabernet Sauvignon) located in the Pencahue valley (Maule Region, Chile), which has a Mediterranean-semiarid climate. The vineyard was established in 2015 and grafted on 110 Richter rootstock with a spacing of 1 m x 2 m. Vines at the experimental site were drip-irrigated (2 drippers plant^-1 with a 2 L h^-1 flow rate) and trained on a vertical shoot position with a canopy height of 1.85 m.

Midday stem water potential (MSWP) measurements were made using a pressure chamber to evaluate vine water status. Simultaneously, a LI-COR gas analyzer was used to measure stomatal resistance (rst) on two leaves that were directly exposed to the sun and were located on the mid-section of the vine canopy. ETv was measured using an eddy covariance system (EC) system installed at 2.2 m above the soil surface. In this case, latent (LEi) and heat (Hi) fluxes were measured using a fast response open-path infrared gas analyzer and a 3-dimensional sonic anemometer, respectively. Raw data of Hi and LEi were post-processed considering corrections of density and sonic temperature.

Images were collected using UAV equipped with a TIR camera (FLIR TAU-2). Additionally, the UAV had an automatic pilot (PIXHAWK) used to generate a flight route. The flight height was configured at 30 m above ground level obtaining imagery at a spatial resolution of 6 cm x 6 cm. After each flight, the images were downloaded to a personal computer and processed through the AGISOFT Metashape software (Agisoft LLC, St. Petersburg, Russia) to generate a thermal Mosaic for each field campaign.

Details of the SW model are indicated by Fuentes-Peñailillo et al., (2018), Ortega-Farias, et al. (2010, 2016, 2021), Riveros-Burgos et al., (2021), and Shuttleworth and Wallace (1985). For the model validation, a comparison between the observed and estimated values of Rni, Gi, and ETv was carried out using root mean square error (RMSE) and mean absolute error (MAE). Also, the index of agreement (d) and the ratio of the observed to estimated values (b) were computed.

Results

Throughout the study period, the MSWP varied between -0.3 and -1.0 MPa indicating that vines were under no water stress with the stomatal resistance ranging between 105-405 s/m. In addition, the ratio of (H+LE) to (R_n-G) at 30 min time intervals was 0.91, indicating that the turbulent fluxes were underestimated by approximately 9%.

Fair comparisons between the observed and estimated values of Rn, Gi, and LEi at the time of the UAV overpass were obtained, as shown in Table 1. This table indicates that the SW model underestimated ETv by approximately 8% of the EC values with RMSE and MAE values of 0.55 and 0.50 mm/day, respectively. The submodels were able to compute LEi, Rni, and Gi with errors less than 15 % relative to ground-true measurements, while RMSE and MAE varied between 37-65 W m^–2. Main disagreements were associated with days when vines were under moderate water stress (MSWP < -1.5 MPa) and had the highest values of canopy resistance.

.

Table 1. Validation of sub-models that compute latent heat flux (LE), net radiation (Rni), soil heat flux (G), and vine evapotranspiration (ETv) over a commercial vineyard using a UAV

Variable	MAE	RMSE	d	B
Rni (W/m2)	37	43	0.80	0.95
Gi (W/m2)	42	47	0.55	0.85
LEi (W/m2)	55	65	0.76	0.91
ETv (mm/day)	0.50	0.55	0.80	0.92

MAE = mean absolute error; RMSE = root mean square error; b = ratio of observed to computed values

Literature cited

Ortega-Farias, S., Poblete-Echeverría, C., and Brisson, N. (2010). Parameterization of a two-layer model for estimating vineyard evapotranspiration using meteorological measurements. Agric. For. Meteorol. 150, 276–286.

Ortega-Farías, S., Ortega-Salazar, S., Poblete, T., Kilic, A., Allen, R., Poblete-Echeverría, C., Ahumada-Orellana, L., Zuñiga, M., and Sepúlveda, D. (2016). Estimation of Energy Balance Components over a Drip-Irrigated Olive Orchard Using Thermal and Multispectral Cameras Placed on a Helicopter-Based Unmanned Aerial Vehicle (UAV). Remote Sens. 8, 638.

Ortega-Farias, S, Esteban-Condori, W., Riveros-Burgos, C., Fuentes-Peñailillo, F., and Bardeen, M. 2021. Evaluation of a two-source patch model to estimate vineyard energy balance using high-resolution thermal images acquired by an unmanned aerial vehicle (UAV). Agricultural and Forest Meteorology, 304–305: 108433. https://doi.org/10.1016/j.agrformet.2021.108433.

Riveros-Burgos, C., Ortega-Farías, S., Morales-Salinas, L. Fuentes‑Peñailillo, F., and Tian, F. 2021. Assessment of the clumped model to estimate olive orchard evapotranspiration using meteorological data and UAV-based thermal infrared imagery. Irrig Sci 39: 63–80. https://doi.org/10.1007/s00271-020-00716-w

Shuttleworth, W.J. and Wallace, J.S. 1985. Evaporation from sparse crops—An energy combination theory. Q. J. R. Meteorol. Soc. 1985, 111, 839–855.