Inversion of net ecosystem CO2 flux measurements for estimation of canopy PAR absorption

The fractional absorption of photosynthetically active radiation (f (PAR) ) is frequently a key variable in models describing terrestrial ecosystem-atmosphere interactions, carbon uptake, growth and biogeochemistry. We present a novel approach to the estimation of the fraction of incident photosynthetically active radiation absorbed by the photosynthetic components of a plant canopy (f (Chl) ). The method uses micrometeorological measurements of CO2 flux and incident radiation to estimate light response parameters from which canopy structure is deduced. Data from two Ameriflux sites in Oklahoma, a tallgrass prairie site and a wheat site, are used to derive 7-day moving average estimates of f (Chl) during three years (1997-1999). The inverse estimates are compared to long-term field measurements of PAR absorption. Good correlations are obtained when the field-measured f (PAR) is scaled by an estimate of the green fraction of total leaf area, although the inverse technique tends to be lower in value than the field measurements. The inverse estimates of f (Chl) using CO2 flux measurements are different from measurements of f (PAR) that might be made by other, more direct, techniques. However, because the inverse estimates are based on observed canopy CO2 uptake, they might be considered more biologically relevant than direct measurements that are affected by non-physiologically active components of the canopy. With the increasing number of eddy covariance sites around the world the technique provides the opportunity to examine seasonal and inter-annual variation in canopy structure and light harvesting capacity at individual sites. Furthermore, the inverse f (Chl) provide a new source of data for development and testing of f (PAR) retrieval using remote sensing. New remote sensing algorithms, or adjustments to existing algorithms, might thus become better conditioned to 'biologically significant' light absorption than currently possible.