Combining target sampling with within field route-optimization to optimise on field yield estimation in viticulture

Abstract

HAL is a multi-disciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L'archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d'enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Is the optimal strategy to decide on sampling route always the same from field to field using the same sampling method to estimate yield?

“…Recent papers (Meyers et al, 2020;Oger et al, 2020) proposed new sampling approaches which simultaneously takes into account the stochastic nature of the spatial variable and the sampling time to produce optimal sampling routes in viticulture. The approach proposed by Meyers et al (2020) provides optimal samples in terms of NDVI values taken by the selected pixels.…”

“…In this approach, the choice of the measurement sites within a pixel (90 m²) is left at the practitioner discretion. The Oger et al (2020) approach combines stochastic methods with constraint optimization to minimize the practitioner's travel time. The approach ensures the representativeness of the measurement in the attribute space (values) by considering the distribution of a high spatial resolution auxiliary data and the optimality of the path to go from one observation site to another.…”

“…The approach ensures the representativeness of the measurement in the attribute space (values) by considering the distribution of a high spatial resolution auxiliary data and the optimality of the path to go from one observation site to another. The Oger et al (2020) approach presents the advantage to directly propose to the practitioner sample sites distributed over an optimal sampling route and does not limit sampling routes to successive pixels (squares of 90 m²) where the practitioner still has to randomly select sampling sites. Assuming that the yield is fully known and a value is available for each within-field site, the Oger et al (2020) method is interesting because it allows one to reverse its application to verify what the optimal sampling route according to the characteristics of the field would be.…”

“…The Oger et al (2020) approach presents the advantage to directly propose to the practitioner sample sites distributed over an optimal sampling route and does not limit sampling routes to successive pixels (squares of 90 m²) where the practitioner still has to randomly select sampling sites. Assuming that the yield is fully known and a value is available for each within-field site, the Oger et al (2020) method is interesting because it allows one to reverse its application to verify what the optimal sampling route according to the characteristics of the field would be. The application of such an approach to fields with known yield values is interesting because it allows: (I) to check whether empirical sampling protocols like the two-row round trip are relevant and if they apply properly to all the fields, (II) to explore whether original and non-trivial routes can emerge from the application of this algorithm and (III) to see whether specific field characteristics promote particular sampling procedures.…”

“…The application of such an approach to fields with known yield values is interesting because it allows: (I) to check whether empirical sampling protocols like the two-row round trip are relevant and if they apply properly to all the fields, (II) to explore whether original and non-trivial routes can emerge from the application of this algorithm and (III) to see whether specific field characteristics promote particular sampling procedures. The objective of this study is, therefore, to apply the methodology of Oger et al (2020) on theoretical yield data with known features to study how factors like the spatial organisation of the yield, the length of the rows, the erratic variance, etc., may affect the optimal sampling route and the error of the yield estimation. This work remains theoretical because it requires prior knowledge of the yield at any point in the field (which is not the case in reality).…”