Spatiotemporal land cover change trajectories across protected areas in Greece during the last decades

Ongoing environmental changes challenge climate-sensitive regions worldwide with Mediterranean ecosystems being also particularly affected. Satellite remote sensing techniques allow for quantitative insights into these patterns by capturing land cover characteristics across a wide range of spatiotemporal scales. Here, focusing on natural protected areas of the European network Natura2000, we quantify the trajectories of their land cover dynamics during the last decades. To do so, we explored satellite imagery from the publicly available Landsat archive, together with advanced cloud computing services, and land cover change detection algorithms. We analysed the spatiotemporal variability in land cover, as quantified with optical remote sensing (Normalized Difference Vegetation Index, NDVI), for more than 50 Natura2000 sites distributed across Greece. The selected sites cover a wide range of environmental conditions. land cover compositions and Mediterranean vegetation patterns (i.e., tall forests, evergreen shrublands, phryganic areas and/or grasslands). Three main land cover change trajectories are examined: (1) abrupt shifts, e.g., due to natural disturbances, such as wildfires, and gradual alterations, namely (2) increase, i.e., ‘greening’, triggered, for example, by more favorable environmental conditions, and (3) decrease, i.e., ‘browning’, following, for example, drought and heat stress. Across the examined sites, the NDVI showed substantial variability, reflecting different land cover characteristics and site-specific demographics and environmental conditions. Site-level long-term mean NDVI ranged from 0.3 to 0.8 with the overall temporal trends being weakly positive. When these lumped spatiotemporal dynamics were disentangled, sites heavily affected by wildfires were identified (showing >50 % losses of their total vegetation cover) as well as sites with chronic decrease or increase in vegetation cover. Given the high significance and numerous services provided by such protected areas, a comprehensive quantification of their land cover dynamics not only enhances our process understanding, but also offers valuable insights to policy makers for the development of mitigation strategies.