Impact of the number of dates and their sampling on a NDVI time series reconstruction methodology to monitor urban trees with Venμs satellite

This article studies the influence of number satellite remote sensing acquisition dates and their sampling on performance a time series reconstruction method developed in Granero-Belinchon et al. 2020. initially aimed at monitoring urban London plane (Platanus x acerifolia) trees, was tested with Sentinel-2 imagery spatial resolutions 10 20 m temporal revisit 5 days. Due to its higher frequency 2 days while having similar resolution m, Venμs consequently used present fulfill purpose this study. The strategy relies building different date configurations based by considering uniform non-uniform samplings total acquisitions ranging from 45 14. Thus, aim is examine annual needed describe properly vegetation phenological cycle impact these final reconstructed series. To end, study carried out using widely Normalized Difference Vegetation Index (NDVI). Results showed that one hand, applied an configuration composed least 18 uniformly sampled throughout year, methodology able correctly NDVI dynamics but leads inaccuracies description intra-annual ones. Nevertheless, descriptions are improved increase acquisitions. On other strongly lead inaccurate undersampled periods correct rest curve. case planes located Toulouse (France) cloud-free images during year 2019. Finally, work emphasizes main limitations studied when few or very available thus identification borderline cases future applications cases.