© 2015 Bajocco et al.Traditionally fuel maps are built in terms of 'fuel types', thus considering the structural characteristics of vegetation only. The aim of this work is to derive a phenological fuel map based on the functional attributes of coarse-scale vegetation phenology, such as seasonality and productivity. MODIS NDVI 250m images of Sardinia (Italy), a large Mediterranean island with high frequency of fire incidence, were acquired for the period 2000-2012 to construct a mean annual NDVI profile of the vegetation at the pixel-level. Next, the following procedure was used to develop the phenological fuel map: (i) image segmentation on the Fourier components of the NDVI profiles to identify phenologically homogeneous landscape units, (ii) cluster analysis of the phenological units and post-hoc analysis of the fire-proneness of the phenological fuel classes (PFCs) obtained, (iii) environmental characterization (in terms of land cover and climate) of the PFCs. Our results showed the ability of coarse-resolution satellite time-series to characterize the fire-proneness of Sardinia with an adequate level of accuracy. The remotely sensed phenological framework presented may represent a suitable basis for the development of fire distribution prediction models, coarsescale fuel maps and for various biogeographic studies.
Bajocco, S., Dragoz, E., Gitas, I., Smiraglia, D., Salvati, L., & Ricotta, C. (2015). Mapping forest fuels through vegetation phenology: The role of coarse-resolution satellite time-series. PLoS ONE, 10(3). https://doi.org/10.1371/journal.pone.0119811