Robust Satellite-Based Identification and Monitoring of Forests Having Undergone Climate-Change-Related Stress

Abstract

Climate-induced drought events are responsible for forest decline and mortality in different areas of the world. Forest response to drought stress periods may be different, in time and space, depending on vegetation type and local factors. Stress analysis may be carried out by using field methods, but the use of remote sensing may be needed to highlight the effects of climate-change-induced phenomena at a larger spatial and temporal scale. In this context, satellite-based analyses are presented in this work to evaluate the drought effects during the 2000s and the possible climatological forcing over oak forests in Southern Italy. To this aim, two approaches based on the well-known Normalized Difference Vegetation Index (NDVI) were used: one based on NDVI values, averaged over selected decaying and non-decaying forests; another based on the Robust Satellite Techniques (RST). The analysis of the first approach mainly gave us overall information about 1984–2011 rising NDVI trends, despite a general decrease around the 2000s. The second, more refined approach was able to highlight a different drought stress impact over decaying and non-decaying forests. The combined use of the RST-based approach, Landsat satellite data, and Google Earth Engine (GEE) platform allowed us to identify in space domain and monitor over time significant oak forest changes and climate-driven effects (e.g., in 2001) from the local to the Basilicata region scale. By this way, the decaying status of the Gorgoglione forest was highlighted two years before the first visual field evidence (e.g., dryness of apical branches, bark detachment, root rot disease). The RST exportability to different satellite sensors and vegetation types, the availability of suitable satellite data, and the potential of GEE suggest the possibility of long-term monitoring of forest health, from the local to the global scale, to provide useful information to different end-user classes.

Keywords

• climate change
• drought stress
• remote sensing
• RST approach
• Google Earth Engine