Abstract
The impact of cloud and cloud shadows on clear-sky Moderate Resolution Imaging Spectroradiometer (MODIS) vegetation index (VI) composite images of the Korean Peninsula was evaluated using four different composite schemes, including weekly, 10 day, biweekly, and monthly maximum value composites. Daily MODIS surface reflectance data were acquired and processed to compute normalized difference vegetation index (NDVI) and enhanced vegetation index (EVI) data sets with the four compositing periods. On average, 65% of the year was cloudy, and biweekly and monthly compositing periods were required to produce spatially continuous images. The means of composite VI maximums from cloud-shadowed pixels were slightly higher than those of clear-day composite data. It seems that the unusual VI highs from cloud-shadowed areas are often influenced by the data acquisition time of the year and viewing geometry of MODIS data products. Although biweekly composites have been frequently used for various applications, their regularity of pixel sampling intervals was lowest among the four data compositing schemes. This data sampling issue should be carefully examined for analyses of VI time-series composite data sets because leaf phenology studies often assume regular sampling intervals between composite periods to estimate the timing of green-up, maturity, senescence, and dormancy.
Acknowledgements
This research was financially supported by the Basic Science Research Programme through the National Research Foundation (NRF) of Korea funded by the Ministry of Education, Science and Technology (grant number 20100613442) and Pusan National University Research Grant 2010. The author thanks two anonymous reviewers for their critical comments and constructive suggestions on this work.