In remote sensing, validation exercises are essential to ensure the quality of the products originated from satellite Earth observations. To assess the measurement uncertainty derived from satellite products, several ground field data from different ecosystems must be available for use. In the same order of importance, it is necessary to define data sampling and up-scaling methodologies to allow a suitable comparison between the ground data and the pixel size of the product. This paper shows the applied methodology used in the FP7 ImagineS project (Implementing Multi-scale Agricultural Indicators Exploiting Sentinels) to validate 10-days global LAI, FAPAR and vegetation cover products at 1km spatial resolution using in-situ data. These global products are derived from PROBA-V observations in the Copernicus Global Land Service. In particular, this case study shows the results of the field-campaign carried out in January of 2015 in the agricultural area of Chimbarongo, Chile. The methodology to scale the ground data and to cre­ate ground-based maps using FASat-C Chilean satellite imagery with a 5,8 m spatial resolution using multivariate least squares regression is shown. Finally, the same methodology was used with a 30 m spatial resolution Landsat-8 image to analyze the effect of the field-data input on the ground-truth maps used to validate the results. Our results show the reliability on the presented methodology and the consistency of the method with regard to the input data. Better results and lower RMSE errors were obtained using FASat-C data. The comparison with satellite products at 1 km shows a good agreement with Copernicus Global Land products derived from PROBA-V observations, and systematic negative bias for the MODIS products.