A comparison of multiscale variations of decade-long cloud fractions from six different platforms over the Southern Great Plains in the United States

This study compares 1997-2011 observationally based cloud fraction estimates from six different platforms (three ground-based estimates and three satellite-based estimates) over the Southern Great Plains, United States. The comparisons are performed at multiple temporal and spatial scales. The results show that 1997-2011 mean cloud fractions from the Active Remote Sensing of CLouds (ARSCL) and from the International Satellite Cloud Climatology Project (ISCCP) are significantly (at a 2% significance level, two-sided t test) larger than the others, having 0.08 and 0.15 larger mean diurnal variations, 0.08 and 0.13 larger mean annual variations, and 0.08 and 0.14 larger interannual variations, respectively. Although more high (low) clouds are likely a reason for larger ARSCL (ISCCP) cloud fractions, other mechanisms cannot be ruled out and require further investigations. Furthermore, half of the estimates exhibit a significant (at a 1% significance level, one-sided t test) overall increase of 0.08-0.10 in the annually averaged cloud fractions from 1998 to 2009; another half of the estimates exhibit little tendency of increase in this decade. Monthly cloud fractions from all the estimates exhibit quasi-Gaussian frequency distributions while the distributions of daily cloud fractions are found dependent on spatial scales. Cloud fractions from all the estimates show much larger seasonal variations than diurnal variations. Findings from this study suggest caution when using observationally based cloud fraction estimates for climate studies such as evaluating model performance and reinforce the need of consistency in defining and retrieving cloud fractions.