Gridding of global cloud structure from asynoptic satellite measurements

Climate properties regulated by convection, like cloud cover and related distributions, are undersampled in asynoptic data from an orbiting platform. Following an overview of the information content of asynoptic sampling, we explore the consequences of undersampled cloud variance on (1) the construction of time-mean distributions and (2) the construction of synoptic maps, which describe the global organization of cloud and its evolution. The results are validated against true cloud cover in high-resolution Global Cloud Imagery that has been composited from 6 satellites simultaneously monitoring the earth. Undersampled diurnal variance leads to systematic errors in the time-mean distribution of cloud cover that exceeds 50% over tropical landmasses, where cloud cover undergoes a pronounced diurnal variation. The random component of undersampled variance can be treated successfully by rejecting small-scale incoherent variance. This reduces the error variance to 10% or less, enabling time series of synoptic maps to be constructed.