Mapping evapotranspiration with high-resolution aircraft imagery over vineyards using one- and two-source modeling schemes

Abstract

Abstract. Thermal and multispectral remote sensing data from low-altitude aircraft can provide high spatial resolution necessary for sub-field ( ≤ 10 m) and plant canopy (≤ 1 m) scale evapotranspiration (ET) monitoring. In this study, highresolution (sub-meter-scale) thermal infrared and multispectral shortwave data from aircraft are used to map ET over vineyards in central California with the two-source energy balance (TSEB) model and with a simple model having operational immediate capabilities called DATTUTDUT (Deriving Atmosphere Turbulent Transport Useful To Dummies Using Temperature). The latter uses contextual information within the image to scale between radiometric land surface temperature (T R ) values representing hydrologic limits of potential ET and a non-evaporative surface. Imagery from 5 days throughout the growing season is used for mapping ET at the sub-field scale. The performance of the two models is evaluated using tower-based measurements of sensible (H ) and latent heat (LE) flux or ET. The comparison indicates that TSEB was able to derive reasonable ET estimates under varying conditions, likely due to the physically based treatment of the energy and the surface temperature partitioning between the soil/cover crop inter-row and vine canopy elements. On the other hand, DATTUTDUT performance was somewhat degraded presumably because the simple scaling scheme does not consider differences in the two sources (vine and inter-row) of heat and temperature contributions or the effect of surface roughness on the efficiency of heat exchange. Maps of the evaporative fraction (EF = LE/(H + LE)) from the two models had similar spatial patterns but different magnitudes in some areas within the fields on certain days. Large EF discrepancies between the models were found on 2 of the 5 days (DOY 162 and 219) when there were significant differences with the tower-based ET measurements, particularly using the DATTUTDUT model. These differences in EF between the models translate to significant variations in daily water use estimates for these 2 days for the vineyards. Model sensitivity analysis demonstrated the high degree of sensitivity of the TSEB model to the accuracy of the T R data, while the DATTUTDUT model was insensitive to systematic errors in T R as is the case with contextual-based models. However, it is shown that the study domain and spatial resolution will significantly influence the ET estimation from the DATTUTDUT model. Future work is planned for developing a hybrid approach that leverages the strengths of both modeling schemes and is simple enough to be used operationally with high-resolution imagery.

Estimation of evapotranspiration of temperate grassland based on high-resolution thermal and visible range imagery from unmanned aerial systems

“…Component net radiation was calculated using the approach described in Xia et al (2016) and Song et al (2016), which is based on measurements of incoming radiation components, the component temperatures T s and T c , and information on surface properties such as albedo, transmittance through the canopy and emissivity. Soil heat flux was estimated as a function of soil net radiation following the approach proposed by Santanello and Friedl (2003), which accounts for the phase shift between the two quantities during daytime.…”

“…Thus Zipper and Loheide (2014) argued that contextual models are not applicable at field scales since the vegetation cover in agricultural landscapes is fairly homogeneous and thus the identification of the extreme limits is complicated. Although contextual models have been extensively tested over a wide range of landscapes using moderate-resolution remotely sensed imagery, their applicability at the subfield scale using high-resolution imagery has not been evaluated to the same extent (Xia et al 2016). …”

“…The rationale behind many EB models is that the latent heat flux is the residual term of the SEB once all other components are either measured or estimated. Net radiation is typically estimated based on measured incoming radiative components, LST, and information on surface properties such as surface albedo and emissivity (Boegh, Schelde, and Soegaard 2000; Su 2002; Xia et al 2016). The soil heat flux is often assumed to be a fraction of net radiation (Santanello and Friedl 2003; Murray and Verhoef 2007; Norman, Kustas, and Humes 1995; Liebethal and Foken 2007).…”

The Grape Remote Sensing Atmospheric Profile and Evapotranspiration Experiment

“…In 2013, a manned aircraft collected imagery at nominally 0.1-m pixel resolution in the visible and near-infrared and 0.5 m in the thermal infrared for three IOPs. A detailed description of the processing and analysis of the data is provided in Ting et al (2016).…”

“…Unfortunately, these methods are not sufficiently robust, particularly for highly structured canopies such as vineyards and tree orchards. They cannot accurately separate crops and the combined interrow soil and cover crop water use, and the crop coefficients are not easily adjustable for stressed conditions (e.g., Ting et al 2016). As a result, significant errors in the timing and amount of irrigation relative to crop water needs have led to an overprescription of irrigation applications.…”