ESTIMATING LIGHT USE EFFICIENCY OF A PINE AND A BEECH FOREST FROM LEAF TO ECOSYSTEM SCALE USING THE PHOTOCHEMICAL REFLECTANCE INDEX Theofilos Vanikiotis (1) , Nikos Markos (1) , Stavros Stagakis (1) , Angelos Tzotsos (2) , Olga Sykioti (3) , Aris Kyparissis (1) (1) University of Ioannina, Department of Biological Applications and Technology, Laboratory of Botany, GR-451 10, Ioannina, Greece, Email: sstagaki@cc.uoi.gr (2) National Technical University of Athens, Department of Topography, Remote Sensing Laboratory, Iroon Polytechneiou 9, GR-157 80, Zografou, Athens, Greece, Email: tzotsos@gmail.com (3) Institute for Astronomy, Astrophysics, Space Applications and Remote Sensing, National Observatory of Athens, Metaxa and Vas. Pavlou str., GR-152 36, Penteli, Athens, Greece, Email: sykioti@noa.gr ABSTRACT The prospect of accurately tracking photosynthetic processes using satellite observations is very important for understanding and monitoring global carbon cycle and climate change. The present study investigates the efficiency of the Photochemical Reflectance Index (PRI) in detecting light use efficiency (ε) in different spatial scales. The study sites concern two dense and homogenous forests in the region of Epirus (Greece), one evergreen coniferous forest dominated by Pinus nigra species and one deciduous forest dominated by Fagus sylvatica. Field and laboratory measurements of canopy structure (Leaf Area Index - LAI, needle and shoot structure characteristics), leaf pigment concentrations, leaf photosynthesis and water potential were performed throughout the growth period. These measurements were used for an accurate description of the ecophysiological characteristics of the two species and thus the parameterization of a Canopy Photosynthesis Model in order to estimate canopy photosynthesis. During the same period, leaf and canopy reflectance measurements were performed in the field to test and evaluate PRI regarding it’s efficiency to track ε in leaf and canopy scale. In order to investigate the potential application of PRI for estimating ε in a broader spatial scale, satellite hyperspectral or superspectral sensors can be used. Compact High Resolution Imaging Spectrometer (CHRIS) and Moderate Resolution Imaging Spectroradiometer (MODIS) are currently available for this purpose and their performances were tested within the present study. An agreement between the fluctuations of CHRIS PRI and the field measured canopy PRI has been found, with both of them appearing to track the ε fluctuation efficiently. However, MODIS PRI shows no intense fluctuation and no relationship with ε and field measured PRI, probably due to lack of atmospheric correction and the effects of viewing and illumination geometry. 1. INTRODUCTION Low spatial resolution multispectral satellite sensors have been used extensively over the last decades for monitoring global vegetation dynamics. The ongoing research has already proved that vegetation green leaf area and therefore plant photosynthetic capacity can be efficiently determined and monitored using multispectral remote sensing imagery through empirical and modeling approaches [1][2]. However, directly detecting how much of this capacity is actually realized at daily and seasonal scales is a much more challenging goal, yet essential for an accurate quantification of the role of terrestrial ecosystems in global carbon balance and climate change. The light use efficiency (LUE) approach [3] expresses gross primary productivity (GPP) as the product of absorbed photosynthetically active radiation (APAR) by the canopy and the efficiency (ε) that the absorbed radiation is converted to biomass (GPP = ε * APAR). Only few methods can be used to estimate ε directly through remote sensing. Most of them employ techniques that detect plant photoprotective mechanisms that are functionally related to photosynthetic rates [4]. Xanthophyll cycle is an important plant photoprotective mechanism that is related with leaf reflectance around 531 nm [5][6][7]. The Photochemical Reflectance Index (PRI) uses the xanthophyll de-epoxidation signal at 531 nm (R 531 ) on the normalized difference formula: PRI = (R 531 – R ref ) / (R 531 + R ref ) (1) As reference wavelength (R ref ) PRI uses a wavelength not affected by the xanthophyll cycle. The most commonly used reference wavelength in literature is 570 nm [7][8]. PRI application requires narrow-band remote sensing instruments with FWHM of 10 nm or less [9] in order to detect xanthophyll de-epoxidation signal. A strong relationship between PRI and ε has been proven by several studies in leaf [7][8][10] and canopy scale [6][11][12][13] concerning different species and