This thesis presents a comprehensive approach to satellite Fraction of Absorbed
Photosynthetically Active Radiation (fAPAR) product validation. This draws on 3D
radiative transfer modelling and metrology to characterise the biases associated with
a satellite fAPAR algorithm and the uncertainty associated with fAPAR estimates.
This extends existing approaches which tend to assume that the in situ measurement
technique produces the same fAPAR quantity as the satellite product.
The validation procedure involves creating a closure experiment where every
aspect of the satellite product definition and its associated assumptions can be tested
from the perspective of the in situ and satellite sensors. The intrinsic differences created by the satellite product assumptions are also assessed, where a new reference
is created. This is known as the “true” fAPAR since it is perfectly knowable within
the context of the radiative transfer model used.
Correction factors between the in situ and satellite-derived fAPAR are created
to correct data collected over Wytham Woods. The results indicate that the corrections reduce differences of >10% to near zero. However, the uncertainty estimates
for the satellite-derived fAPAR show that it does not meet the requirements given by
Global Climate Observing System (GCOS) (≤(10% or 0.05)). The wider implications of the retrieved uncertainties are also presented showing that it is unlikely that
the GCOS requirements associated with downstream applications that use satellite
fAPAR can be met currently.
This work represents an important step forward in the validation of satellitederived fAPAR because it is the first time that the absence of satellite and in situ data
uncertainty and traceability, and satellite product definition differences have been addressed. This paves the way for the improvement of satellite fAPAR products
because their uncertainties can now be quantified effectively and their validation
conducted fairly, meaning there is now a benchmark to base improvements on.