Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Doctorat → États-Unis → 2021 → Estimation of High-Resolution Evapotranspiration in Heterogeneous Environments Using Drone-Based Remote Sensing

Utah State University (2021)

Estimation of High-Resolution Evapotranspiration in Heterogeneous Environments Using Drone-Based Remote Sensing

Nassar, Ayman M. M. 

Titre : Estimation of High-Resolution Evapotranspiration in Heterogeneous Environments Using Drone-Based Remote Sensing

Auteur : Nassar, Ayman M. M. 

Université de soutenance : Utah State University

Grade : Doctor of Philosophy (PhD) 2021

Evapotranspiration (ET) is a key element of hydrological cycle analysis, irrigation demand, and for better allocation of water resources in the ecosystem. For successful water resources management activities, precise estimate of ET is necessary. Although several attempts have been made to achieve that, variation in temporal and spatial scales constitutes a major challenge, particularly in heterogeneous canopy environments such as vineyards, orchards, and natural areas. The advent of remote sensing information from different platforms, particularly the small unmanned aerial systems (sUAS) technology with lightweight sensors allows users to capture high-resolution data faster than traditional methods, described as “flexible in timing”. In this study, the Two Source Energy Balance Model (TSEB) along with high-resolution data from sUAS were used to bridge the gap in ET issues related to spatial and temporal scales. Over homogeneous vegetation surfaces, relatively low spatial resolution information derived from Landsat (e.g., 30 m) might be appropriate for ET estimate, which can capture differences between fields. However, in agricultural landscapes with presence of vegetation rows and interrows, the homogeneity is less likely to be met and the ideal conditions may be difficult to identify. For most agricultural settings, row spacing can vary within a field (vineyards and orchards), making the agricultural landscape less homogenous. This leads to a key question related to how the contextual spatial domain/model grid size could influence the estimation of surface fluxes in canopy environments such as vineyards. Furthermore, temporal upscaling of instantaneous ET at daily or longer time scales is of great practical importance in managing water resources. While remote sensing-based ET models are promising tools to estimate instantaneous ET, additional models are needed to scale up the estimated or modeled instantaneous ET to daily values. Reliable and precise daily ET (ETd) estimation is essential for growers and water resources managers to understand the diurnal and seasonal variation in ET. In response to this issue, different existing extrapolation/upscaling daily ET (ETd) models were assessed using eddy covariance (EC) a nd sUAS measurements. On the other hand, ET estimation over semi-arid naturally vegetated regions becomes an issue due to high heterogeneity in such environments where vegetation tends to be randomly distributed over the land surface. This reflects the conditions of natural vegetation in river corridors. While significant efforts were made to estimate ET at agricultural landscapes, accurate spatial information of ET over riparian ecosystems is still challenging due to various species associated with variable amounts of bare soil and surface water. To achieve this, the TSEB model with high-resolution remote sensing data from sUAS were used to characterize the spatial heterogeneity and calculate the ET over a natural environment that features arid climate and various vegetation types at the San Rafael River corridor


Version intégrale (9,3 Mb)

Page publiée le 4 janvier 2022