Correcting orbital drift signal in the time series of AVHRR derived convective cloud fraction using rotated empirical orthogonal function

摘要

The Advanced Very High Resolution Radiometer (AVHRR) instruments onboard theseries of National Oceanic and Atmospheric Administration (NOAA) satellitesoffer the longest available meteorological data records from space. Thesesatellites have drifted in orbit resulting in shifts in the local timesampling during the life span of the sensors onboard. Depending upon theamplitude of the diurnal cycle of the geophysical parameters derived,orbital drift may cause spurious trends in their time series. We investigatetropical deep convective clouds, which show pronounced diurnal cycleamplitude, to estimate an upper bound of the impact of orbital drift ontheir time series. We carry out a rotated empirical orthogonal functionanalysis (REOF) and show that the REOFs are useful in delineating orbitaldrift signal and, more importantly, in subtracting this signal in the timeseries of convective cloud amount. These results will help facilitate thederivation of homogenized data series of cloud amount from NOAA satellitesensors and ultimately analyzing trends from them. However, we suggestdetailed comparison of various methods and rigorous testing thereof applyingfinal orbital drift corrections.