When reviewing the 7 cloud removal steps above, it became obvious to us that using a standardized cloud removal algorithm on our coastal sea surface temperature data was a bad idea. In test 2 for example, you may want to change the sensitivity (delta t) based on the image. Less sensitivity would be good for an upwelling image while more may be appropriate for a non upwelling image. In test 6, a standardized reflectance threshold for cloud brighness cannot be used because sun angle, image pass time, time of month, and time of year all play a part in the reflectance values of clouds. A morning image maximum reflectance (value 8) may be the minimum in an afternoon pass (ie it's brighter in the afternoon than early morning).

So, we simply analyze each image and determine the cloud removal specifications on a case-by-case basis. This process is more time consuming than a standard algorithm, but sometimes you must sacrafice time to get perfection.

Now let's look at some SeaWiFS navigation information. .