On the relationship between some production parameters and a vegetation index in viticulture

The use and timing of many agronomical practices such as the scheduling of irrigation and harvesting are dependent on accurate vineyard sampling of qualitative and productive parameters. Crop forecasting also depends on the representativeness of vineyard samples during the whole phenological period. This manuscript summarizes the last two years of precision viticulture in Sicily (Italy); agronomic campaigns were carried out in 2012 and 2013 within the "Tenute Rapitalà" and "Donnafugata" farms. Normalized Difference Vegetation Index derived from satellite images (RapidEye) acquired at berry set, pre-veraison and ripening phenological stages (occurred at June, July and August respectively) have been related to production parameters (sugar and anthocyanins contents) at harvesting of a selected red autochthonous cultivar (Nero D'Avola). The research aims to assess how robust are prediction models based on simple linear regression analysis, in particular: 1) whether there is a suitable period for acquiring the remote sensing image to evaluate these parameters at harvesting, when their knowledge is required; 2) if these relationships are consistent between years or need to be re-calibrated; 3) the models transferability to other vineyard of the same cultivar. © 2013 SPIE.