Ozone flux measurement and modelling on leaf/shoot and canopy scale.

The quantitative study of the ozone effects on agricultural and forest vegetation requires the knowledge of the pollutant dose absorbed by plants via leaf stomata, i.e. the stomatal flux. Nevertheless, the toxicologically effective dose can differ from the stomatal flux because a pool of scavenging and detoxification processes reduce the amount of pollutant responsible of the expression of the harmful effects. The measurement of the stomatal flux is not immediate and the quantification of the effective dose is still troublesome. The paper examines the conceptual aspects of ozone flux measurement and modelling in agricultural and ecological research. The ozone flux paradigm is conceptualized into a toxicological frame and faced at two different scales: leaf/shoot and canopy scales. Leaf and shoot scale flux measurements require gas-exchange enclosure techniques, while canopy scale flux measurements need a micrometeorological approach including techniques such as eddy covariance and the aerodynamical gradient. At both scales, not all the measured ozone flux is stomatal flux. In fact, a not negligible amount of ozone is destroyed on external plant surfaces, like leaf cuticles, or by gas phase reaction with biogenic volatile compounds. The stomatal portion of flux can be calculated from concurrent measurements of water vapour fluxes at both scales. Canopy level flux measurements require very fast sensors and the fulfilment of many conditions to ensure that the measurements made above the canopy really reflect the canopy fluxes (constant flux hypothesis). Again, adjustments are necessary in order to correct for air density fluctuations and sensor-surface alignment. As far as regards flux modelling, at leaf level the stomatal flux is simply obtained by multiplying the ozone concentration on the leaf with the stomatal conductance predicted by means of physiological models fed by meteorological parameter. At canopy level the stomatal flux is calculated by SVAT models often based on the energy balance of the soil-vegetation-atmosphere system and on the big-leaf concept. This latter assumes the canopy as equivalent to a single leaf having a leaf area equal to the total area of all the plant s leaves and lying at a certain height above the ground. The complexity of SVAT models ranges from one-dimensional to three-dimensional models. The most used are one-dimensional models in single-layer, dual-source or multilayer version. The main uncertainties in flux modelling are currently associated to the estimation of the non-stomatal flux component and to the up-scaling process from leaf to canopy and stand level. For the latter a separate representation of sun and shaded leaves is recommended.