Multitemporal Cloud Detection and Masking

The aims of this study are to investigate the spectral properties of cloud and to carry out cloud detection and masking using MODIS (Moderate-resolution Imaging Spectroradiometer) data. To do this we make use of the spectrally rich satellite data provided by MODIS sensor, which is equipped with 36 bands ranging from visible to thermal wavelengths. Cloud detection and masking were first carried out individually using single date of MODIS data. Multitemporal cloud analysis was later carried out using MODIS data from 24 different dates from 2004 to 2005. The eastern parts of Malaysia were found to have more cloudy days than the western parts, in which consistence with the meteorological observations made by the Malaysian Meteorological Services.