Processing of high spectral resolution reflectance data for the retrieval of canopy water content information

This article examines how different processing, in particular different degrees of smoothing, affect the extraction of biophysical information from high spectral resolution data. Spectral reflectance data collected during the EISAC'89 Campaign from and experimental grassland site on Tadham Moor, Somerset, United Kingdom were used for the analysis, together with canopy biomass, chlorophyll, and water content variables. The analysis focuses mainly on the correlation between percentage water content per unit area of canopy and spectral parameters across the two water absorption features, around 940 nm and 1150 nm. Correlation analysis revealed that both the strength of the peak correlation between the first derivative of reflectance and water content and its wavelength position were sensitive to the degree of smoothing for the 1150 nm feature, although this was not the case for the 940 nm water absorption feature. This sensitivity is attributed to the fact that different levels of smoothing emphasize different spectral features across a wavelength region where the first derivative of reflectance and the biophysical variable are generally highly correlated. The effect emphasizes that the exact position of peak correlation between these two variables should be interpreted with caution, especially across spectral regions where the correlation is generally high. A simple reflectance ratio based on the relative depth of the water absorption feature was found to be more closely related to canopy water content and insensitive to the degree of spectral smoothing. (C) Elsevier Science Inc., 1998.