Accurate estimates of sensible (H) and latent (LE) heat fluxes and actual evapotranspiration (ET) are required for monitoring vegetation growth and improved agricultural water management. A large aperture scintillometer (LAS) was used to provide these estimates with the objective of quantifying the effects of surface heterogeneity due to soil moisture and vegetation growth variability. The study was conducted over drip-irrigated vineyards located in a semi-arid region in Albacete, Spain during summer 2007. Surface heterogeneity was characterized by integrating eddy covariance (EC) observations of H, LE and ET; land surface temperature (LST) and normalized difference vegetation index (NDVI) data from Landsat and MODIS sensors; LST from an infrared thermometer (IRT); a data fusion model; and a two-source surface energy balance model. The EC observations showed 16% lack of closure during unstable atmospheric conditions and was corrected using the residual method. The comparison between the LAS and EC measurements of H, LE, and ET showed root mean square difference (RMSD) of 25 W m−2, 19 W m−2, and 0.41 mm day−1, respectively. LAS overestimated H and underestimated both LE and ET by 24 W m−2, 34 W m−2, and 0.36 mm day−1, respectively. The effects of soil moisture on LAS measurement of H was evaluated using the Bowen ratio, β. Discrepancies between HLAS and HEC were higher at β ≤ 0.5 but improved at 1 ≥ β > 0.5 and β > 1.0 with R2 of 0.76, 0.78, and 0.82, respectively. Variable vineyard growth affected LAS performance as its footprints saw lower NDVILAS compared to that of the EC (NDVIEC) by ~0.022. Surface heterogeneity increased during wetter periods, as characterized by the LST–NDVI space and temperature vegetation dryness index (TVDI). As TVDI increased (decreased) during drier (wetter) conditions, the discrepancies between HLAS and HEC, as well as LELAS and LEEC Re decreased (increased). Thresholds of TVDI of 0.3, 0.25, and 0.5 were identified, above which better agreements between LAS and EC estimates of H, LE, and ET, respectively, were obtained. These findings highlight the effectiveness and ability of LAS in monitoring vegetation growth over heterogonous areas with variable soil moisture, its potential use in supporting irrigation scheduling and agricultural water management over large regions.
A numerical framework to advance agricultural water management under hydrological stress conditions in a data scarce environment