Celestia Forums

It?€™s often been debated and discussed here in the forums as to the extent of what Earth's cloud cover is and what it looks like from orbit. I have finally found enough data to put this to rest. Many people will think that the cloud cover is still to extensive for their tastes. But in the end we are trying to replicate as accurate model of the Earth as we can in Celestia at this time.

The data imagery I have discovered is going to allow me to make several new cloudmaps and cloudmap normalmaps that will all be at the 8k level. These will be totally independent of the old BlueMarble cloudmap of the past. I will be able to keep any image cloning down to a minimum or to nonexistent. These clouds cover the Northern Summer Southern Winter at this time. I should be able to continue to get data for the other seasons as time goes by. That way I can build up a selection of seasonal cloudmaps to go along with the seasonal BlueMarble Next Gen textures.

I can't give a timetable at this point but i can show a very rough example of what the average cloud cover looks like and it will also give everyone an idea of what to expect and what the condition of the imagery is in at this time.



I superimposed two close time line images to fill the blank data areas in this image and also used an Earth texture to fill in the missing Southern Polar data. The pale banding is the suns reflection of the oceans as the satellite orbits the Earth. As the images are taken the reflection of course changes as to the satellites position so it makes the reflection look stretched out. We have the same effect in the old BlueMarble cloudmap. But image processing and cloning cleaned a great deal of this noise up. I will be attempting to do the same thing.

So any comments are welcome at this point. I am hoping to have the first example of a clean cloudmap loaded in Celestia by the end of the month for everyone to see. At that time I will make a new topic in the Textures section as it will then be an official texture subject.
Wish me luck on this.

Don. Edwards

It's definitely busier, but it looks more realistic! I'll be interested to see how this turns out.

you might try using CImg-1-14's program called greycstoration
i used it on small 2k tiles of 64kvenus ( about 30 min.) it dose not suport transparency
so a b & w map would do
it builds in mingw and
|-- osx_xcode_2_2
|-- unix_gnu_gcc_3
|-- win_bloodshed_devcpp_4
|-- win_borland_bcc32_5
|-- win_digital_mars_compiler_8
|-- win_intel_icl_8
|-- win_microsoft_visual_NET_2003
`-- win_microsoft_visual_cpp_6

Just to verify what I think I see, The clouds in the image are all new data except for the poles? I think this would be about right for true cloud cover. Even the 2k "real" time cloudmaps generated by xplanet about every 3 hours are showing about the same amount of clouds.

A couple of links that I think would support this taken by astronauts from ISS.
http://eol.jsc.nasa.gov/scripts/sseop/p ... 431562.tsv
http://eol.jsc.nasa.gov/scripts/sseop/Q ... 13-E-10149
http://eol.jsc.nasa.gov/scripts/sseop/p ... 434402.tsv
http://eol.jsc.nasa.gov/scripts/sseop/p ... 434402.tsv

Good luck with this Don.

cartrite