Hi,

I'm trying to propogate errors (uncertainties) through a equation which contains ratios of SeaWiFS bands (ie Lwn443/Lwn555).
I have estimates of the uncertainties for single bands (e.g. values from Zibordi et al 2006, GRL) but I can't seem to find any references for the uncertainties/errors for ratios of bands.  I'm guessing that the errors in each band are likely to be correlated (so it may not be suitable to assume that they completely cancel each other out when they appear in a ratio).

Does anyone know of any references that state the uncertainties for ratios of SeaWiFS bands ?
(I'm guessing that this would have been done for the OC3 and OC4 chlorophyll algorithm development but I can't find any references).

Thanks for your help.

Cheers
Jamie