Validation and Application of European Beech Phenological Metrics Derived from MODIS Data along an Altitudinal Gradient

• Main Authors: Veronika Lukasová, Tomáš Bucha, Jana Škvareninová, Jaroslav Škvarenina
• Format: Article
• Language: English
• Published: MDPI AG 2019-01-01
• Series: Forests
• Subjects: validation; phenology; NDVI; LAI; spectral analyses; European beech; altitude
• Online Access: http://www.mdpi.com/1999-4907/10/1/60

Monitoring plant phenology is one of the means of detecting the response of vegetation to changing environmental conditions. One approach for the study of vegetation phenology from local to global scales is to apply satellite-based indices. We investigated the potential of phenological metrics from moderate resolution remotely sensed data to monitor the altitudinal variations in phenological phases of European beech (Fagus sylvatica L.). Phenological metrics were derived from the NDVI annual trajectories fitted with double sigmoid logistic function. Validation of the satellite-derived phenological metrics was necessary, thus the multiple-year ground observations of phenological phases from twelve beech stands along the altitudinal gradient were employed. In five stands, the validation process was supported with annual (in 2011) phenological observations of the undergrowth and understory vegetation, measurements of the leaf area index (LAI), and with laboratory spectral analyses of forest components reflecting the red and near-infrared radiation. Non-significant differences between the satellite-derived phenological metrics and the in situ observed phenological phases of the beginning of leaf onset (LO_10); end of leaf onset (LO_100); and 80% leaf coloring (LC_80) were detected. Next, the altitude dependent variations of the phenological metrics were investigated in all beech-dominated pixels over the area between latitudes 47°44′ N and 49°37′ N, and longitudes 16°50′ E and 22°34′ E (Slovakia, Central Europe). In all cases, this large-scale regression revealed non-linear relationships. Since spring phenological metrics showed strong dependence on altitude, only a weak relationship was detected between autumn phenological metric and altitude. The effect of altitude was evaluated through differences in local climatic conditions, especially temperature and precipitation. We used normal values from the last 30 years to evaluate the altitude-conditioned differences in the growing season length in 12 study stands. The approach presented in this paper contributes to a more explicit understanding of satellite data-based beech phenology along the altitudinal gradient, and will be useful for determining the optimal distribution range of European beech under changing climate conditions.