Describing the spatio-temporal variability of vines and soil by satellite-based spectral indices: A case study in Apulia (South Italy)

A time series of Landsat 8 OLI (L8 OLI) multispectral images acquired between May 2013 and February 2016 were used to investigate vigour, vine and soil water content in a vineyard of Moscato Reale (syn. Moscato Bianco) sited in the Castel del Monte DOCG area. Normalized Difference Vegetation Index (NDVI) and Normalized Difference Water Index (NDWI) were calculated and compared with vine midday stem water potential (Psi(MDstem)) and soil volume water content (VWC), to calibrate estimation models. Estimation models were calibrated using already existing ground observation datasets from previous ordinary vineyard management operations: Psi(MDstem) was measured at two different locations in vineyard at 6 different dates in summer 2014; VWC was continuously measured from June to October 2014 and from January to September 2015. Results showed that: a) vine stem water potential can be locally estimated with an accuracy ranging from +/- 0.046 (high vigour vines) to +/- 0.127 (low vigour vines) MPa; b) soil volume water content can be locally estimated with an accuracy of about +/- 1.7%. Medium resolution satellite imagery proved, therefore, to be effective, at vineyard level, to describe vigour, vine and soil water status and their seasonality. This is an important issue to focus on since, as Landsat 8 images are free, the entire process is economic enough to be consistent with cost and incoming of the farming system.