Estimation of vineyard water requirements using satellite-based surface energy balance

Abstract

Since the 90's as a consequence of the Global Warming and the “La Niña” (ENSO) phenomena, the Chilean wine industry has been forced to increase irrigation systems investments, adopting new strategies of irrigation scheduling to improve the vineyard water use efficiency without affecting the grapes yield and quality. To develop an adequate irrigation strategy, the first step has been to estimate the vineyard water uptake or actual evapotranspiration ETa).Traditionally, vineyard ETa has been estimated by multiplying the reference evapotranspiration (ETo) by a single crop coefficient (Kc). Unfortunately, a major uncertainty in this approach is that many Kc values reported in literature are empirically determined from point-based measurements, and they are unable to describe the spatial variability of vineyard Kc, for each phenological stage. Recent advancements in using satellite remote sensing to determine ETa over space and time have made it possible to assess the variation in crop ETa at low-cost. One of the most cited remote sensing-based algorithms is METRIC (Mapping EvapoTranspiration at high Resolution with Internalized Calibration) which is a onedimensional surface energy based residual model that has been extensively customized for application in full covered crops. For sparse crops such vineyards, as far as we know the application of METRIC to estimate ETa and Kc is still unexplored. In this way, this thesis was developed with the aim of critically study the applicability of METRIC over a drip-irrigated Merlot vineyard. To meet this, a study was carried out during the 2006-2007, 2007-2008 and 2008-2009 growing seasons. METRIC-based estimations were compared against ground measurements of vegetation indexes, surface energy balance components and vineyard ETa. Results indicated that METRIC’s semi-empirical sub-models for estimate vegetation indexes should be calibrated before its direct application to vineyards (Chapter 2). On the other hand, in the evaluation of METRIC to estimate the different components of the vineyard surface energy balance, results indicated that the compensation between the instantaneous fluxes of net radiation, soil heat fluxes and latent heath fluxes produced acceptable estimations of latent heat luxes (error less than 10%) to extrapolate them to daily (24 h) evapotranspiration (ETa) (Chapter 3). Finally for the main phenological stages of the Merlot vineyard, METRIC overestimated the Kc by about 10% in relation to ground measurements. However, those errors did not significantly affect the overall performance of METRIC during the study period into the estimation of daily (24 h) ETa and Kc (Chapter 4). Results exposed from Chapters two to four suggest that it is possible to confirm that the METRIC model can be used for obtain relatively good estimations of the vineyard ETa during the complete growing season.