Correcting MODIS 16-day composite NDVI time-series with actual acquisition dates

Remote sensing phenological works often use vegetation index (VI) time-series (TS). Since ground-observed phenological metrics occurrences vary by a few days from year to year, TS temporal accuracy became mandatory, but it is less strict in composite data. A technique to recover the temporal accuracy of 250 m 16-day composite VI from the MODIS MOD13Q1 product is proposed, relying on acquisition dates contained in the Composite day of the year layer. We demonstrated that the correction process significantly affected the VI TS during most of the year, especially in spring and autumn when the starting of season (SOS) and the end of the season (EOS) are expected. As a consequence of the TS correction process, SOS estimation showed to be affected too.