Vegetation moisture dynamics plays a key role in wildland fire risk assessment. While dead fuel moisture can be
considered only dependent on the dynamics of the meteorological variables, live fuel dynamics is also related to the
phenological state of the considered species as well as the soil water content. Minimal variations in live moisture content
can cause a great effect in wildfire risk condition, since the fine fuel load is mainly composed by live vegetation.
This work presents a model able to predict the moisture content of live fine fuel starting from the phenological principles
of leaf growth cycle. This approach assumes that the phenological state of a given group of vegetation at a given time
instant (day of the year) can be modelled as a function of the local meteorological conditions, the soil parameters, and
some vegetational parameters. The Vegetation Index products of MODIS sensor have been used to parameterize and
calibrate the model. To this end, experimental parcels, fully representative of the Mediterranean vegetation cover of
Liguria Region (Italy), were used as test areas. Such areas are equipped with a suite of meteorological sensors, and are
periodically subject to sampling campaign aiming at characterizing the phenological state and the moisture contents of
their different vegetation species. The data collected during the field campaigns were completed by the observations of
MODIS-NDVI from 2001 to 2006.
The paper provides a calibration procedure of phenological module carried on using the whole data sets (meteorological
data, phenological and physiological data, NDVI imagery), and formalized through a mathematical programming
approach. The phenological model was implemented with reference to five areas placed on the Italian territory.