Effect of the differences in spectral response of Mediterranean tree canopies on the estimation of evapotranspiration using vegetation index-based crop coefficients

The vegetation index (VI)-reference evapotranspiration (ETo) method incorporates the estimation of basal crop coefficients from spectral VIs into the FAO56 guidelines for computing crop evapotranspiration (ET). Previous research pointed to the possibility of the differential spectral response of some Mediterranean crops, specifically olive trees. To evaluate this hypothesis and the potential related effects on the VI-ETo method, this work studied the spectral response of four Mediterranean canopies under full vegetation coverage: three fruit trees (olive, orange and almond trees), and the holm oak trees of the dehesa ecosystem. Spectral measurements were taken on dense vegetation placed on a workbench and over dense treetops, avoiding in both cases the effect of soil background. The results showed that the soil adjusted vegetation index (SAVI) for full-cover olive trees was significantly lower than for other fruit trees (0.57 for olive trees vs. 0.71 for orange tree and 0.70 for almond tree). SAVI of olive vegetation measured on the workbench was lower than that measured over treetops, probably due to the effects of canopy architecture and shadowing. SAVI obtained on oak treetops (0.51) was even lower than that on olive treetops. This differential spectral response of olive and oak trees influenced the estimation of ET (and water stress). The validation using ET measurements obtained with the eddy covariance method in the olive orchards showed a reduction of root mean square deviation (RMSD) from 0.73 to 0.6 mm day(-1) when daily ET was estimated assuming SAVI(max) = 0.57 in comparison with a generic value for Mediterranean crops.