Dear Yoav,

At global scales, cloud cover does not change radically
over the 1 to 3 hours between SeaWiFS and Aqua-MODIS
views.  Land cover changes at an even slower (geological)
pace.  These two account for the majority of the masking.

Other masks which do change more between sensors are
caused by inter-orbit gaps and sun-glint.

I wouldn't call the masking similarities disturbing, but rather,
reassuring.

Regards,
Norman

Dear Norman,
Thank you for the quick response!

What strikes me is the similarity on a regional (e.g. over the north Atlantic) scale, where there is hardly any difference between the two datasets.

To my I understanding, over such spatio-temporal scales one should expect much more variability in cloud cover.

All the best
Yoav

Yoav,

Do you have any two particular level-3 images
in mind as an example of what you are referring
to?

Norman

Hi Norman,

See for example clouds over the north Atlantic for the 8-days-composite image from the 4-11/7/2007.

Yoav

Yoav,

The cloud similarities do not look unexpected to me.
Remember SeaWiFS and MODIS view the same bits
of ocean within a few hours of each other.  Clouds just
don't change that much over such a time scale.  Granted,
over an 8-day period, many of the clouds will get filled
in, but areas with persistent clouds should still retain
the same relationship between SeaWiFS and MODIS.

I'm attaching a SeaWIFS/Aqua-MODIS comparison
image for the 8-day period you named just to make
sure we are talking about the same thing.

Dear Norman,
Thank you for the thorough response!

I guess it means that cloud variability is smaller than I thought (given that indeed masked pixels mostly correspond to clouds)...

All the best
Yoav

I would also note that by mid-2007, the SeaWiFS orbit had drifted to about 1 PM equator crossing time, so the time difference with Aqua was small.

Fred