Analysing long-term spatiotemporal land surface phenology patterns over the Iberian Peninsula using 250 m MODIS EVI2 data.

Iberian Peninsula ecosystems are especially vulnerable to global warming, and variations in climate patterns may alter the wide variety of services they provide. Despite this, seasonal variations in Iberian ecosystems have been understudied. Thus, this study aims to characterise land surface phenology (LSP) patterns in the Iberian Peninsula over 21 years (2001-2021), considering three phenometrics: the start (SOS), the end (EOS), and the length of the growing season (LOS). These were estimated from 8-day image composites of EVI2 (Two-band Enhanced Vegetation Index), derived from the surface reflectance product MOD09Q1 at a 250-metre spatial resolution. Phenometrics and in-situ human phenological observations of plant phenophases were also compared. Pearson's correlation coefficient, p-value, and absolute differences between paired phenometrics and phenophases were calculated to quantify uncertainty between both phenological approaches. Generally, SOS and EOS dates were later in the Alpine and Atlantic biogeographic regions. SOS occurred in March-April and EOS between October-December. Natural vegetation land cover types had similar phenological dynamics, with SOS occurring between late winter and spring and EOS between autumn and early winter. However, phenometric dates were earlier in Mediterranean savannas, grasslands, and scrublands. Atlantic evergreen broadleaf vegetation showed the strongest correlations between SOS and the first (r = 0.96) and second (r = 0.68) leaf unfolding. In contrast, Mediterranean evergreen broadleaf vegetation had unclear correlations. Atlantic deciduous broadleaf vegetation showed a moderate correlation between SOS and first (r = 0.52) and second (r = 0.57) leaf unfolding. The correlation between SOS and the first (r = 0.14) and second (r = 0.13) leaf unfolding was weaker for Mediterranean deciduous broadleaf vegetation. EOS and autumn phenophases generally showed unclear consistency. Higher spatial resolution satellite data may improve the consistency between EOS and autumn phenophases in Iberian ecosystems, as well as between SOS and spring phenophases in heterogeneous Mediterranean ecosystems.