Simulation of 100 albedo measurements per year of a linearly rising
albedo of 1 percent per decade. The simulation is based on the mean monthly global albedo estimates over several decades from a suite of models as in Bender et al . 2006 (Tellus, 58A, 320) figure 3.

grey crosses : individual measurements

green line — monthly averages

red circles — annual averages

the black lines are the fits to the monthly averages — they both come out with the correct slope – so it does work!

the two panels show the effect of 1 percent (lower) and 0.1 percent scatter (upper) on
the individual albedo measurements due to natural variability.

the 0.1% case allows the underlying model to be seen clearly.

1% scatter in the daily albedo makes things look a deal noisier — but the right slope comes out and one can get a pretty good seasonal pattern
from the monthly means and better stull by folding the data into a year.

we are looking further into what daily variations naturally arise — the Bender paper is not clear on this.