Use of a MODIS-Derived Photochemical Reflectance Index to Detect Interannual Variations in the Photosynthetic Light-Use Efficiency of a Boreal Deciduous Forest
Guillaume
Drolet, Laval University, guillaume.drolet.1@ulaval.ca
K.
Fred
Huemmrich, University of Maryland, Baltimore County, karl.huemmrich@gsfc.nasa.gov
Forrest
Hall, University of Maryland, Baltimore County, fghall@ltpmail.gsfc.nasa.gov
Elizabeth
Middleton, NASA Goddard, Elizabeth.M.Middleton@nasa.gov
Andy
Black, University of British Columbia, andrew.black@ubc.ca
Alan
Barr, Environment Canada, alan.barr@ec.gc.ca
Hank
Margolis, Laval University, hank.margolis@sbf.ulaval.ca
(Presenting)
Biochemical changes in leaves during stress events have been shown to result in changes in spectral reflectance at 531 nm and the Photochemical Reflectance Index (PRI) was developed to measure this phenomenon. We calculated PRI from MODIS reflectance data for cloud-free days between 2001 and 2003 for the Fluxnet-Canada Old Aspen flux tower in Saskatchewan and the flux and meteorological data from the tower allowed us to calculate the photosynthetic light-use efficiency (LUE) at the time of MODIS overpasses. A linear relationship was found between the PRI and the LUE only when backscatter spectral data (minimal shadowing) was used. The relationship was stronger for top of the atmosphere reflectance data (R2=0,76) compared to data that had been atmospherically corrected with MODIS-derived aerosol optical depth values and the Six-S atmospheric correction model (R2=0.53). While our analysis of MODIS-derived PRI did not capture the seasonal variations in LUE, it did seem to detect interannual variations associated with drought versus non-drought years. An ability to reliably estimate LUE from satellites would significantly improve large-scale modelling of the carbon cycle.