Remote sensing of canopy light use efficiency using the photochemical reflectance index: Model and sensitivity analysis

Abstract

A growing number of studies have shown that reflectance changes at 531 nm, associated with the xanthophyll cycle and the related thylakoid energisation are widespread among plant species. The photochemical reflectance index (PRI), derived from narrow band reflectance at 531 and 570 nm has been related with some success to photosynthetic light use efficiency (LUE). Such a relationship would enable the estimation of stand photosynthesis from remotely sensed data. However, canopy PRI is an integral of the component leaf response weighted by the strength of the signal from each leaf to the sensor. This analysis investigates the extent to which canopy structure, view, and illumination angles are likely to influence the measured canopy PRI. A one-dimensional ray tracing radiative transfer model was used to estimate light distribution within a canopy and the dynamic response of individual foliar elements, based on a published relationship between PRI and LUE and a simple photosynthetic light response function. The model estimated the LUE of the canopy, based on both incident and absorbed light, and reflectance of the canopy at the desired wavelengths and hence the canopy PRI. A range of solar zenith, leaf area index (LAI), leaf angle distributions (LAD), and soil types were used to determine the likely influence on measured canopy PRI. The results show a positive correlation between PRI and LUE variation at canopy scale. However, the index shows a greater variation of view angle than most vegetation indices. The index is strongly influenced by varying soil background for LAI<3. At large viewing angles (>30°) the index is also sensitive to LAD. Correction for Rayleigh scattering is necessary to relate the index to ground measured PRI. Results show that the PRI value is most sensitive to changes in LAI. Utilisation of the relationship to predict or improve estimates of canopy LUE based on either absorbed or incident light will require an independent estimate of LAI change between dates/locations of in situ measurements and of remote sensing observations.