Evaluation of airborne and terrestrial laser scanning in Mediterranean forest fuel type characterization

In the last twenty years the airborne Light Detection and Ranging (LIDAR) has became a widespread remote sensing technology in several environmental applications, such as the vegetation characterization and the fuel inventory. In this field, airborne LIDAR technology can be an effective alternative to overcome the difficulties of the conventional ground based small area forest inventory approaches: high costs, time consuming methodologies, limited accuracy when the ground data were extended to landscape level. Several works reported the use of the airborne LIDAR approach in order to estimate and map several parameters of vegetation used in fire spread and behaviour modelling applications: fuel type, plant height, canopy cover, crown bulk density, etc. However, the accuracy and the capabilities of the former approach depend on the factors affecting the laser pulse reflection from both the vegetation and the soil, i.e. the plant cover and species. A method to overcome this problem may be the application of the terrestrial laser scanner in order to merge the data coming from the two different approaches. The aim of this work is to assess the capabilities of the airborne and terrestrial laser scanners in predicting vegetation and fuel characteristics of coniferous and broadleaf forests of a mountain area of central-eastern Sardinia. The analysis of data demonstrate similar accuracies between the two approaches with conifers forests; on the contrary, the terrestrial laser scanner provided better results in broadleaf forests, where accurate estimations of the live and dead fuel load were obtained; the integration of the two sets of data provided improved estimates of the crown bulk densities.