Cloud detection in sea surface temperature images by combining data from NOAA polar-orbiting and geostationary satellites

Cloud contamination in images from satellite-borne, advanced very high resolution radiometers (AVHRR) largely limits their utility by yielding absent and erroneous observations of sea surface temperature (SST). Accurate cloud detection and masking techniques are essential to improve timely areal SST coverage. In the present study, a preliminary algorithm for cloud detection based on wavelet transform (WT) is developed. Results from WT algorithm and two previously published algorithms, Clouds from AVHRR Phase-I (CLAVR-1) and Cayula and Cornillon are compared. It is concluded that the Cayula-Cornillon algorithm gives the best cloud detection, while CLAVR-1 algorithm provides too aggressive masks and WT algorithm shows similar cloud masks as those from Cayula-Cornillon algorithm. Potentials of using Geostationary Operational Environmental Satellites (GOES) SST data to compensate cloud-contaminated values on SST images from the polar-orbiting satellites of National Oceanic and Atmospheric Administration (NOAA) are discussed.