Monitoring live fuel moisture content and phenological behaviour of Mediterranean shrubland species using an automated digital system

Wildland fires are an important disturbance factor for natural ecosystems in the Mediterranean Basin. Wildland fire occurrence and behavior are determined by many different factors such as topography, weather, and fuel. Among them, vegetation type and water status are very important as drier fuel contributes to facilitate fire ignition and propagation. Shrublands are an important component of Mediterranean vegetation, which constitute the surface fuels primarily responsible for the ignition and the spread of wildland fires. As fires will easier propagate when live fuel moisture (LFM) is low, monitoring the plant water content of live vegetation can be a useful tool to identify critical periods of high ignition danger for Mediterranean ecosystems. However, because field surveys are expensive and time consuming, LFM is generally measured in the field at only a few discrete locations and time to determine the level of fire danger. Vegetation monitoring based on "near-surface" remote sensing techniques have been proposed in recent researches, in particular for monitoring the physiological and phenological status of vegetation communities. These techniques use radiometric instruments or imaging sensors to quantify, at high temporal resolution, the seasonal changes in the optical properties of the vegetation canopy. In this paper, an Automated Phenological Observation System (APOS), based on digital image sensors, used for monitoring the seasonal changes of vegetation water status and phenological behaviors of Mediterranean shrubland species is presented. The APOS system was developed under the EU-funded research infrastructure project INCREASE (an Integrated Network on Climate Change Research) and installed in a Mediterranean shrubland area. The APOS system is set to acquire a large panorama of vegetation every day at noon (36 shoots per panorama). The analysis of digital images and RGB (Red, Green and Blue) signal reveals both color variations related to the seasonal rhythms of plant phenology and those related to the physiological changes in the water status of vegetation. Primarily, this system would allow to (i) analyze digital images in the visible spectrum to evaluate physiological traits associated with the water status of plants and (ii) monitor the phenology and water status of vegetation based on technical of proximal sensing. The use of systems based on "near-surface" remote sensing techniques can provide new nonintrusive predictive tools to identify variation in the vegetation water status, and consequently in the LFM content seasonal pattern, which can allowed field surveys less expensive and time consuming than traditional monitoring. Moreover, they can be a valid support to identify critical periods of high ignition danger for Mediterranean ecosystems.