do you mean the temporal standard deviation of the LAI?
You could calculate the LAI for various dates (e.g. one per month), then merge them all in a stack (collocate) and then apply Radar > Coregistration > Stack Tool > Stack Average
and select Standard deviation. This will give you a measure on how constant the LAI is over the entire range of images per pixel.

Sorry should be more clear - “The standard deviation of the estimation errors”

Example: " MODIS LAI product - Next to the observations itself, we have to provide a standard deviation of the estimated value as a measure of accuracy of the observation.

For this example, we will assume that the LAI estimates have a standard deviation of 5% of the estimated value. The latter is very accurate compared to the accuracy reached by typical satellite LAI products which is more in the order of 20%."