Netzer, Y.; Lenski, I.; Michael, Y.

Vegetation indices (VI) derived from a new generation of high-spatial resolution satellites, namely Planet and Sentinel-2, were tested as proxies for stem water potential (Ψ-stem) in commercial vineyards. Multivariable linear regression models were developed from the Planet and Sentinel-2 data and in-situ Ψ-stem measurements in 82 vineyards in Israel, providing Ψ-stem estimates every 1-2 weeks. With multivariable regression VIs in the VIS-NIR region (Planet) and moisture VIs in the NIR-SWIR region (Sentinel-2) were correlated with in-situ Ψ-stem measurements. In general, Sentinel-based models performed better (higher correlations and higher ability to capture temporal and spatial variability) than Planet-based models. The trade-off between spectral, spatial and temporal resolutions of these two satellites are further discussed. © Wageningen Academic Publishers 2019