Connectix ball cameras as weather sensor.

[Chris Albertson <chris@topdog.pas1.logicon.com>]

I have been looking at the sky the last few days.  Lately
there are always clouds to the south but the north is clear.
I think I understand why there is less MkIII data than I would
expect.

It seems this is a big limitation with a fixed mount drift scan
instrument like the Mk III.    The percentage of nights where it
is clear at the specific small patch of sky where the system aimed
may be low, while the percentage of nights where _some_ patch of 
sky is clear is fairly high.  The Mk III system cannot be pointed
to the clear areas.
It is likely just a law of statistics that the probability of
it being clear _some place_ is always greater then the probability
of it being clear at a _given place_.

It seems that in principle the Mk IV could be aimed to where the
clouds not not present.  If so it could be much more productive
then if it simply scanned out a fixed pattern.  I think you would
call this "opportunistic scheduling".  The question is how to 
detect where the clear areas are.

The problem is easy if an operator were willing to stay up all
night and supervise the system.  The problem is the the MkIV is
an instrument you would like to run every night so you'd like to
do "automated opportunistic scheduling". 

It occurred to me that one of those cheap "ball" cameras made by 
Conectix could be used.  These things cost only $80.00 complete 
with cable and have a wide angle lens.  You can control the exposure
time in software.  Would one of these things be usable as a "clear
area detector"?  If we are lucky the camera's limiting magnitude is
mag 3 or so but let's assume it is more like mag -4 instead.  
Clouds should appear as bright areas and holes would be darker.  The
pixel sixe would be on the order of a few arc minutes (I guess)

I read an article recently in one of the trade mags about a new line
of very simple image sensors.  These devices typically have a three
wire serial interface to a computer and have on chip the sensor array,
timing electronics and analog to digital converter.  An entire digital
camera on a chip.  Some of these camera chips sell at <$10.00 each.
The article said that many engineers are reading the specs and seeing 
numbers like "64 gray levels, 32 X 32 image array size" and writing 
these things off as useless.  The author made a point that you should
not think of this as a poor CCD camera but as a _very_ smart photo diode.
This is what got me thinking that a connectix could make a decent
detector.

Des anyone have one of these cameras? If so what happens if you point it
out the window at night?


-- 
--Chris Albertson

  chris@topdog.logicon.com                Voice:  626-351-0089  X127
  Logicon RDA, Pasadena California          Fax:  626-351-0699