SAR and optical data comparison for detecting Trees Outside Forest in agroforestry landscapes

Small woods, tree hedgerows, scattered and isolated trees, are also known as Trees Outside Forest (TOF). TOF have an important role in agroforestry landscapes, enhancing their ecological connectivity, hosting biodiversity and having significant impact on biomass and carbon stocks. The identification and classification of TOF on a small area are easy to be accomplished. However, identifying, classifying and mapping TOF at regional or national level are complex, expensive and time-consuming tasks. Precise and fast techniques to estimate the agroforestry surfaces at both regional and national levels are needed. Despite the increasing number of studies combining remote sensing and field surveys for the identification and classification of TOF, guidelines for TOF inventory and mapping in agroforestry systems are still lacking. Furthermore an accurate and objective estimate of the extent and geographical distribution of agroforestry systems in Europe is crucial for the development of supporting policies. In this study we compare the use of Synthetic Aperture Radar (SAR) and optical data, derived from the Sentinel mission dataset, for detecting, classifying and mapping TOF in Italian traditional agroforestry landscapes. We tested the methodology in two areas of interest located in Umbria region (central Italy) where oak trees and hedgerows coexist with crops.