Regarding the interpretation:
The proportions allow a quasi-absolute quantification of the observed greenness values in a long-term comparison. To do so, the absolute deviations from the long-term mean (more precisely, the median) are computed. Quantiles -which do not allow for quantifying the absolute deviations – should always be interpreted in connection with the proportions to get an idea of the absolute greenness deviations.
In order to enable a classification of the values, proportions were quantified for locations at which dying pines have been observed in spring 2016 as a result of the 2015 summer drought (references 2-3) as well as dying beech trees in spring 2019 (reference 7) as a consequence of the 2018 summer drought (reference 7). Proportions observed at these locations varied between -20 and -5 % at the time of tree death and averaged -10 %. In addition, theoretical considerations were made based on high-resolution, multispectral drone flights of dying pines (reference 2), according to which a greenness value of -10 % would indicate that at least a quarter of the population is under severe stress. A transfer of this interpretation, based on observations and theoretical considerations, to other sites seems plausible, but should be done with caution as due to the relatively coarse spatial resolution of the satellite images other subscale processes can influence the computed greenness values.
The proportions of former drought-years are to be found in the archive.