I'm trying to organize a Science Fair project that will use fractal
geometry to model cloud formation and appearance.

The idea is to take or use photos of 'real' cloud types, and then do
computer modeling to create structures that form until they fairly resemble the
photos.

I'm limited to Windows PC computing, and although there are lots of
programs around that will generate pretty fractal pictures, I've not
found a program that's adaptable to computing in this manner.  I'm
especially _not_ interested in a 'canned' fractal program that throws
out pictures.  It has to be a process that can be controlled, so we
can fairly say that we created or at least implemented a computer
model of cloud formation.

Please help me find and implement the Windows software -- free or at
nominal cost -- that will get the cloud project 'off the ground.'

---

Request for Question Clarification by rcd-ga on 05 Dec 2002 01:58 PST

Hello richard-ga

Are you looking for programming source codes that would produce cloud
like effects?

As I have found a nice selection, but it would require programming
languages like C or pascal to change the variables.

---

Clarification of Question by richard-ga on 05 Dec 2002 07:33 PST

Unfortunately I have no experience with Pascal or C, nor is either
installed on my Windows XP PC.  No doubt I could obtain Pascal or C
software for the Windows machine, but I'm hoping to avoid learning to
program in those languages.

So for now I'd like to wait and see if a Windows solution is possible.
 If that doesn't appear in the next few days I'll post a later
clarification to invite Pascal or C ideas suitable for a novice like
me....

Thanks
-R

---

Clarification of Question by richard-ga on 05 Dec 2002 07:46 PST

Mathtalk:

Regarding the points raised by your comment,

1.  Unfortunately besides being ignorant of Pascal and C, I'm ignorant
of Java.  Again I'll wait a few days before inviting solutions that
would require me to attain minimal skill in one of these languages.

2.  My question does invoke the difficult problem of taking a picture
of a real cloud and approximating it by fractal techniques.  However
it only needs to be achieved at a basic Science Fair level of
sophistication--in other words we're not doing real meteorology here,
we're just looking to put together a demonstration.  If the patented
algorhythms that you refer to can serve to simplify the job they'd be
great--but if they're high level meteorological tools they might not
be practical for me.

Thanks
-R

---

Clarification of Question by richard-ga on 05 Dec 2002 11:07 PST

Mathtalk

If I'm understanding you correctly, the Science Fair project based on
your algorithms would end up resembling the Master's thesis cited in
diagonal-ga's comment.  And if I am capable of handling the
computational side (via your algorithms) that would be a very
satisfactory solution.

The Master's thesis is very much what I had in mind--an introductory
section that reviews the physics of cloud formation, and then a
computer program that accepts some sort of input that allows me to
draw clouds that resemble the photos [notwithstanding the fact that
the computer program has no meteorological basis and hence little or
nothing in common with the way Nature makes clouds].

I don't know the significance of the algorithms being 'patented.' 
Does that mean I have to license them in some fashion?  Perhaps the
non-commercial, educational use would be OK?  Are the algorithms
'fractal?'  I ask that because 'fractal' has good appeal in a science
fair context--in fact the poster display could end up in thirds-- 1/3
the physics of clouds, 1/3 the nature of fractals, and 1/3 the
payoff--using fractals to emulate [in some fashion] the formation of
clouds.

So if you think I can get past the two hurdles--me obtaining the legal
right to use these algorithms, and me being able to learn enough to
run these algorithms on a Windows computer and generate the desired
output--then I think you've got it.

Thanks,
richard-ga

---

Clarification of Question by richard-ga on 06 Dec 2002 05:45 PST

Researchers, my original question asked that you "help me find and
implement the Windows software ..."

If you're worried that answering this question will commit you to
teaching me how to use the cloud-modeling program, worry not.  You
only have to recommend which software you believe will work
appropriately for me in this situation, and point to a source from
which I can obtain it.

Also, I am now ready to accept answers requiring that I learn
something of Java, Pascal or C programming--the less the better!

-R

Hello Richard,

Here is my answer I had originally put together. The recommendations
get more specific and useful as you read further....

I think I have found a number of programs that may be suitable for
your project.

Something called Skyworks is available which includes the source code
http://www.vterrain.org/Atmosphere/Clouds/

There is a classic fractal program for MSWindows called Fracint. This
includes source code. You should be able to get the source code for it
here http://www.programmersheaven.com/zone10/cat101/14057.htm

A possible alternative search strategy instead of clouds would be to
try plasma, this has a very similar cloudy look.

This page has basic programs that relate to creating plasma type
patterns
http://www.basicguru.com/abc/graphics.htm

Here is the TurboC source code for a basic plasma, which you could
alter for your project at http://pascal.sources.ru/demo/plasma.htm
So you may need look at the code and how the colors are defined and
alter the parameters a bit.

There is a great list of source codes for plasma simulations  is at
http://www.programmersheaven.com/zone10/cat344/

Here is something a little more convincing as it has a picture with
it, that potentially looks more like a basic cloud.
http://www.shdon.com/weffects.html#plasma


I hope that this helps you alittle to get you 'of the ground'. You may
need to get a computer programming language like Pascal or C on your
machine, copy the source code into it and then play around with the
variables to make the clouds more or less grainy etc.

Oh I found one more thing. It's not quite plasmas/clouds but the
source code is available on the page and it looks pretty cool. They
are at http://www.reallyslick.com/

Clarification to help you get started right away with a little
programming

You can get a copy of turbopascal directly at
http://community.borland.com/article/images/20803/tp55.zip
This is on the page with other programming languages at
 http://www.devq.net/pascal/doc.php?cid=11

and then you could copy the simple the program for creating a plasma
that I mentioned at: http://pascal.sources.ru/demo/plasma.htm
just copy and paste it

or you could use  the programs at

http://www.shdon.com/weffects.html#plasma

Another source is at
http://www.tmt.com/ for a so called pascal compiler


Most of these programs are fairly small and have comments through them
to allow you to change the variables. As a science teacher this is the
sort of approach I would suggest to my students.

kind regards

rcd-ga


Google search strategy
"source code" fractal cloud formation
"source code" plasma OR cloud OR clouds
"source code" fractal plasma cloud OR clouds

---

Request for Answer Clarification by richard-ga on 09 Dec 2002 16:35 PST

Hello rcd-ga

I am working through your answer and mathtalk's comments to see what
programs might be practical for me.

Meanwhile, I also found a Windows program called Terragen

http://www.planetside.co.uk/terragen/
http://caroluk2.crosswinds.net/terratut/
http://home.cogeco.ca/%7Eedegiuli/TGtut/TerragenTutMain.htm

Do you think it might be a more workable cloud generator than the
sources cited by you and mathtalk?

Thanks,
richard-ga

---

Clarification of Answer by rcd-ga on 10 Dec 2002 00:50 PST

Hello Richard,

The program you mentioned looks very good. As you have probably noted
the clouds dialogue box at
http://www.planetside.co.uk/terragen/guide/dlg_clouds.html
May be what you are looking for. I had a look through the site but I
can't see the source code.

The only possible difficulty is determining how the parameters relate
to fractal dimensions.

I've continued my search on google with
programming fractal clouds OR plasma

Here is an example of a scientist's work on comparing a molecular
cloud in space with a fracal model can be found at
http://www.ph1.uni-koeln.de/~ossk/Myself/structure.html

Information about fractal clouds can be found at
http://climate.gsfc.nasa.gov/~cahalan/FractalClouds/FractalClouds.html

Here is a really good quick background information about
fractal/plasma cloud that would be useful background information for
your poject. It also includes a pascal program on the page
http://www2.vo.lu/homepages/phahn/fractals/plasma.htm

I found a fractal cloud generator but it's in German. It allows you to
control specific fractal parameters. It is located at
http://www.ph1.uni-koeln.de/~ossk/Myself/Fractalweb/fraktale-welt.html
This is what 	 would probably use, but  would need to either email the
professor who created the program or get an English German dictionary.
This is probably the most accurate and variable cloud generator.

Ahah!, just discovered Google has a translator so that page can be
translated
Here is the URL I used
http://translate.google.com/translate?u=http%3A%2F%2Fwww.ph1.uni-koeln.de%2F%7Eossk%2FMyself%2Ffractal-program.html&langpair=de%7Cen&hl=en&ie=ISO-8859-1&prev=%2Flanguage_tools


I have discovered that the technical name for clouds and coast lines
in fractal geometry is something called "fractional Brownian motion"

Well I'm not sure if this helps but it is starting to look a little
better, at least you can use all this information as background info.
But I do admit that the program you found is more realistic. I think
part of your challenge is to find out how the parameters listed relate
to fractional brownian motion.

I have certainly had fun looking for this information, it's one of my
pet topics.

regards

rcd-ga

Hi, richard:

There seems to be source code at various sites for simulating the
appearance of a cloud, e.g.

http://www.bekkoame.ne.jp/~ishmnn/java/cloud_s.html

Search Strategy: fractal cloud

See also the NASA stuff found this way, including:

http://climate.gsfc.nasa.gov/~cahalan/FractalClouds/Movies/Movies.html

which has both real and computer generated "movies" that display
structure at varying scales.

Perhaps your problem is a difficult one, in which you wish to take a
picture of a real cloud and approximate it by fractal techniques.  I
believe that there are some patented algorithms for this sort of
"image compression" that might be of interest to you.  Comment back if
you would care to know more about that.

regards, mathtalk-ga

Richard, the algorithms I mention have nothing to do with meteorology;
they only concern using fractals to approximate images, which might be
of any type.  Fractal clouds would just be one application for a
general method of image compression.

-- mathtalk-ga

A Swedish programmer has posted his cloud images here:

http://www.d.kth.se/~nv91-gta/exjobb/

and you can find out more about him here

http://www.nada.kth.se/~gustavt/

He sounds approachable but I don't know what would be required to run his software.

Another possible source:

David Ebert - you can see his images here.

http://dynamo.ecn.purdue.edu/~ebertd/cloud/

There's an excellent little introduction to the field at

http://freespace.virgin.net/hugo.elias/models/m_clouds.htm

which includes a program which will display some animated images (it
requires DOS4GW, which is available via the same page) but doesn't
appear to be configurable in any way.

A totally different (non-fractal) approach appears on

http://www.sinc.sunysb.edu/Stu/fdachill/clouds/clouds.html

where a cloud shape is produced by 'punching' an oval at different
scales - this may be the sort of thing you could demonstrate live if
you have access to this sort of modelling software.

There's also a little Java applet for cloud modelling available
through

http://www.dseffects.com/

but this appears to produce random forms. Yoiu can download if from 

http://download.com.com/3001-2411-5547171.html

Richard,

Setting aside the question of clouds for the moment, I suggest looking
at these links for software and ideas:

Keywords: fractal image compression
://www.google.com/search?hl=en&ie=UTF-8&oe=UTF-8&q=fractal+image+compression&btnG=Google+Search

regards, mathtalk-ga