MODELLING THE CARBON SINK IN ITALIAN FOREST ECOSYSTEMS USING ANCILLARY DATA, REMOTE SENSING DATA AND PRODUCTIVITY MODELS
Input data
The preparation of the input strata requested a large effort and takes almost one year of intense activity of the research group.
All the information were standardized to match the same geographic characteristics: a spatial resolution of 1 km with the geographic projection UTM on fuse 32N with datum WGS84. We adopted the IDRISI file and metafile format.
SPOT vegetation: remotely sensed images of Normalised Difference Vegetation Index were acquired as decadal images from VITO.
Vegetation: the CLC dataset is available in Italy with a special in-depth thematic elaboration for natural and semi-natural environments. The main dominant forest species are mapped with the same geometric details of the CLC project (minimum mapping unit of 25 ha). The map is available for the year 2000 and 2006. In the standardization procedure we calculated for each 1 km x 1 km pixel the percentage of forest type cover.
Soils: for soils we used the data created within the framework of the Carboitaly project (Papale, 2006). Soil dataset includes the following information: percent of sand, percent of lime, percent of clay, carbon in the dominant profile (kg/m2), soil depth, soil pH, apparent density (g/cm3). Maps are created on the basis of the data from soil profiles acquired within the European project SPADE2. Data from the dominant soil profiles are related to the soil mapping units of the European Soil Atlas of Europe. More info at EUSOILS.
DEM: the digital elevation model is a generalization of the 90 m resolution data from the SRTM (Shuttle Radar Topography Mission at http://srtm.csi.cgiar.org).
Meteo: the BIOME-BGC model requires input daily meteo data for minimum and maximum temperature and precipitation. To create this (very) large dataset (365 daily maps x 3 variables x 10 years = 10950 maps) we used the ENSEMBLES daily spatial dataset E-OBS (Haylock et al., 2008) available across Europe at 0.1° downscaling at 1 km resolution by local regression with climate maps we created at national level (Blasi et al., 2007). Local regression analysis was performed with an alghorythm developed in-house from previous work of Maselli (2001 and 2002). The daily spatial estimates we obtained were validated in 2011 with meteo station from the national net of SIAN-UCEA. Since the results were positive (r between predicted and measured data of 0.916 for minimum temperature, of 0.946 for maximum temperature and of 0.637 for precipitations) we adopted this approach for all the entire study period. These first results were presented in Chirici et al. (2011) at the 15th national Conference of Federation of Scientific Associations for Territorial and Environmental Information (ASITA 2011).
Go to:
C_FORSAT home page
C_FORSAT MODEL
C_FORSAT VALIDATION
C_FORSAT RESULTS