1) It should be geometrically (x,y,z) and radiometrically
(e.g, to percent reflectance) calibrated so that remotely
sensed data obtained on different dates can be compared
with one another.
2) The remotely sensed data must usually be calibrated
(compared) with what is on the ground in terms of
biophysical (e.g., leaf-area-index, biomass) or cultural
characteristics (e.g., land use/cover, population density).