Nevermind the reason, but my task is to predict unique climate
conditions by a combination of four attributes in a database, such
that every combination makes sense and all four attributes are
involved in every unique result.  I have four input channels available
and must use them all; I need to know what to monitor for each
channel.

Visualize a table in which there are five columns.  The first four
contain values for each of four weather factors.  Currently, I have
Wind, Humidity, Pressure, and Temperature.  The fifth column contains
the resulting climate condition (generalized, unique, two-word
descriptions like blizzard or thunder storm) for each combination of
weather factors.  By "unique" I mean climate conditions that are
sufficiently different from each other that it would make sense to
display a different symbol or icon (e.g. "rainy" and "drizzling" would
result in the same icon on a weather map, thus not unique).

My present system is not working, for at least three reasons: 1) I am
not familiar enough with meteorological nomenclature to briefly
describe enough climate conditions, 2) some weather factor
combinations do not make sense to me (e.g. is it possible to have high
wind + low humidity + low pressure + low temperature?), and 3) I may
have selected the wrong four factors.

In my current table, wind can be "calm", "1-5 mph", "6-25 mph", or
"greater than 25 mph".  The other three factors can be high or low.

What I need is for an expert to correct the system I described above, add or
modify the values available to the weather factors, and fill in the
results for each possible combination of the factors using only a few
words.

Details:

The results must be short because they will be used in another
database that limits the field length for the "weather description"
field to 20 characters.

There must be four weather factors involved in producing every climate
result, no more and no less.  If this is impossible then I will need
to know what combinations are invalid, so I can flag my input data
when those combinations occur, and remap them to an appropriate result.

Using my current system, there are 32 possible combinations.  Of
course, order has nothing to do with the number of unique combinations
(hot/dry is the same as dry/hot).

Currenty, wind is the only field with more than two possible values. 
It could be redefined with no less than three or no more than five
ranges because any more radical change would generate too few or too
many result possibilities.

Wind does not have to be the field with more than two values, it could
be another factor.  Another factor could have up to three values, but
the other two must only have two values.  There are two reasons: 1) of
my four input channels, one is variable, one has three possible
signals it registers (on/off/out-of-range), and the other two have
only two signals they register (on/off); and 2) these signal options
allow for 24 - 60 possible results; 24 might be too few possibilities
for my purpose and 60 would almost certainly be too busy to graph the
way I need to.

I set the base fee at $25 because the minimum answer would be to keep
my existing system and simply fill in all 32 UNIQUE
climate descriptions using no more than 20 characters each, which for
an average expert ought to take half an hour.  But, I can and will pay a
bonus for a complete solution depending on how closely it satisfies
the directions above.  If you are a pro and you expect $150/hr
then don't spend much time on this!  It's for a hobby project, not a
business or a job; I don't have a large budget to hire a pricey
consultant or I wouldn't be posting my problem here.

Hang a rock on a string. 
1. If it is wet,,,it is raining.
2. If it is dry and hot... the sun is shining, and it is summer
3. If it is white and cold...it is snowing, and it is winter
4. if it is moving on the end of the string...the wind is blowing and
it could be stormy.

Q.E.D.

JUST JOKING.

You say you've a fifth column? Be very careful with the fifth column.
During World War 2 the Fifth Column was hell-bent on sabotaging the
war effort! Buy bonds!

Why have you assumed that you can describe all possible weather
conditions with 4 factors?  Why have you assumed that there are only
32 possible climate conditions?

:-D
Well, at least my question is generating some good comments.
Weather rocks!
Bonds?  James's bonds?

frankcorrao-ga:

I haven't assumed either.  I realize it's not strictly feasible to
define weather so simply.  Super computers are still working on that. 
I'm merely limited by my existing data collection gear, expertise,
budget, and time.

I hope to trade a reasonable portion of my funds for a better
approximation of the 24-60 most recognized patterns based on the four
most common of many influences than I can devise myself.

Hi stuckfly-ga,

I can't really help much with the statistical aspects of this question
as my knowledge of databases and such is much too limited, but if
you'd like to increase your knowledge of meteorology, here's a good
place to start:
http://www.free-ed.net/free-ed/Science/EarthSci/MeteorAcad/default.asp 

As for short descriptions of weather, aviation weather reporting uses
what are known as METAR/TAF abbreviations, which might possibly be of
some help to you. Here's a list of them:
http://www.alaska.faa.gov/fai/afss/metar%20taf/metcont.htm 

Good luck, sounds like an interesting project.

Best,
Byrd-ga

Perhaps I got this wrong, if so, please bear with me..

In the 4 fields you described, how about rain, or snow?

I'd suggest to replace pressure (it's kinda vague anyway, at least for
me) with, well, I don't know the exact term for it, but here goes:

Wind: none - weak - mild - strong - extraordinary
Humidity: dry - normal - humid
Things-that-drop-from-the-sky: none - rain - snow - ice
Temperature: very low - low - mild - hot - very hot

Obviously, some combinations would be null, like snow on very hot
temperature. But still the above produces 300 combinations, and I
guess at least 100 would be logical.

Just my two cents.