[meteorite-list] Catalina Sky Survey

From: Linton Rohr <lintonius_at_meteoritecentral.com>
Date: Sun, 28 Mar 2010 22:53:00 -0700
Message-ID: <960D898FCD8C4C32B8CCC5B6374793BD_at_D190TH71>

Thank you, Richard.
It's always bothered me when I've seen article's refering to the "automated"
Catalina Sky Survey discovering 2008 TC3, since I knew one of our own was
there. Not knowing the details though, I appreciate your "more detailed
look" and applaud your "lengthy" response.
I too, can speculate on why someone would minimize your contribution to this
historic event, but I keep being interupted by the persistent image of a dog
retreiving a thrown stick. It landed rrriiight...... there!
Linton

----- Original Message -----
From: "Richard Kowalski" <kowalski at lpl.arizona.edu>
To: "Meteorite List" <meteorite-list at meteoritecentral.com>; "Shawn Alan"
<photophlow at yahoo.com>
Sent: Sunday, March 28, 2010 9:29 PM
Subject: Re: [meteorite-list] Catalina Sky Survey


> All,
>
> I've changed the subject line to bring it in line with the discussion.
> As I started writing this, I thought I could get by with a simple
> response, but
> to give a clear idea of why Catalina is not an automated system, I thought
> I'd
> give you all a more detailed look at what we do each night. Sorry for the
> length
> of this response...
>
>
>
> Shawn I'll answer your query about the article first.
>
> That article is one of a line of articles all based on Jenniskens' initial
> and
> continuing misrepresentation of how we operate.
> I can't blame the reporters because they can only go on what they are
> told.
> "Automated" has been repeated often enough that it is becoming accepted
> fact.
>
> At Catalina the observer chooses the fields that will be observed and in
> what
> order, making this decision depending on the conditions, previous
> coverage,
> needs of follow up observations, etc. The telescope then observes those
> fields
> in order. Once this first set of fields has been completed, the telescope
> repeats the cycle three more times.
>
> Each time an image comes down from the camera our computers process the
> images,
> and then record in x-y coordinates every "object" it sees. Once all four
> images
> have been made and processed the four sets of x-y coordinates are compared
> with
> each other. The objects that are in the same locations are stars, since
> they
> haven't moved. They are ignored.
>
> Then a number of filters are run to remove groups of false detections
> around
> bright stars and "objects" that appear to be moving too fast (artificial
> satellites) or too slow.
>
> At the beginning of the night we download the file of the known 300,000+
> minor
> planets from the Minor Planet Center. Our software uses this file to
> calculate
> what object should have the same position and apparent motion as the
> remaining
> suspects that are left after these filtering steps.
>
> These, along with the remaining candidates that can't be identified are
> are
> given one final calculation before the are presented to the observer for
> validation as an actual or false detection. This calculation assigns a
> number
> called the "Digest". The Digest is a determination that the apparent
> motion of
> the object is that of a Main Belt asteroid or not.
>
> Objects that are identified are presented at the bottom of our validation
> candidates. Those that have low digest numbers, representing objects that
> have
> apparent motion indicative of Main Belters, but which have not been
> identified,
> are presented in a group second from the bottom. Finally, objects which
> have
> higher digests, meaning they are not Main Belters, are presented at the
> top.
>
> The observer has to go through and actually look at each and every one of
> the
> candidates in this first group of detections. This could number from just
> a
> couple to 100 or more. Most of these will be marked "N" for "No" they are
> not
> real, but every so often an object is presented to the observer for
> validation
> that is real and clearly has a motion that is not typical of a main belt
> asteroid. Of these four observations, sometimes one or more detections may
> be
> "bad". In that case the observer can accept all four, or throw out one,
> two or
> all four observations. (You never submit a single observation).
>
> The observer also can manually measure each position to increase the
> accuracy of
> the positions.
>
> After a real object is discovered, the observer then compares the position
> of
> all known Near Earth Asteroids with the position and motion of this object
> to
> determine if it is known or not. Sometimes it is but anywhere from 1 to 10
> or
> more times a night it is a newly discovered NEO. The observer then reports
> the
> observations to the Minor Planet Center, which posts these observations in
> near
> real time. The observer also schedules additional same night follow up
> observations before continuing the search for more new NEOs.
>
> We repeat this process for anywhere from 9 to 20 fields per set and up to
> 20 or
> so sets in a single night. As you can see this means we look at thousands
> of NEO
> candidates each and every night.
>
> Part of our success is we operate at a detection sigma of 1.2. Most
> consider a
> sigma of 3 a minimum to provide a reasonable Signal to Noise Ratio, but we
> have
> determined that the human eye - brain system can pull out real objects
> that are
> barely detectable above the noise level. True automated surveys don't go
> below 5
> sigma because you can't teach a computer and software to reject false
> positives
> at such a low SNR.
>
> So, to summarize.
>
> A human determines where to observe each night, all night
> A human looks at each and every candidate to determine if it is real or
> not
> A human determines if all the positions are accurate or need to be
> remeasured.
> A human determines in real time if a real candidate is known or not.
> A human submits the observations for public dissemination in real time.
> A human determines if same night follow up observations need to be
> scheduled and
> accomplished.
> A human determines if the expected plan of observations can be followed or
> needs
> to be modified because of weather or discoveries.
>
> As one additional perk, a human determines if a candidate shows activity,
> going
> through the same process of identification, reporting, scheduling follow
> up,
> etc. By being an active candidate, I mean seeing a coma and a tail. IOW,
> determining the candidate is a comet.
>
>
>
> As for Petr, I could only speculate why after being informed of his error,
> that
> he would continue to minimize the efforts that it took to discover 2008
> TC3 and
> predict its impact point down to +/- 1 kilometer so he could head out to
> that
> exact point months later to search for black rocks on white desert sand...
>
>
> I hope this helps explain why Catalina is not an automated survey.
>
>
> Richard
>
>
>
>
> --- On Sun, 3/28/10, Shawn Alan <photophlow at yahoo.com> wrote:
>
>> Hi Richard and List,
>>
>> That's is a great point that you brought up about the
>> automated Catalina Sky Survey 1.5 m telescope at Mount
>> Lemmon, Tucson, Arizona, and how Catalina is not an
>> automated survey. I wonder why they say that about her? Is
>> it the software that is used for Catalina or the way she
>> scans the sky making it automated? Or is it all done
>> manually these days and one by one scientist scan the sky
>> asteroids?
>>
>> Here is an article from 2009 that uses automated when
>> referring to Catalina Sky Survey....
>>
>>
>> The four-meter-diameter asteroid, called 2008 TC3, was
>> initially sighted by the automated Catalina Sky Survey
>> telescope at Mount Lemmon, Ariz., on Oct. 6. Numerous
>> observatories, alerted to the invader, then imaged the
>> object. Computations correctly predicted impact would occur
>> 19 hours after discovery in the Nubian Desert of northern
>> Sudan.
>>
>>
> http://esciencenews.com/articles/2009/03/25/asteroid.monitored.outer.space.ground.impact
>>
>>
>> Shawn Alan
>
> ______________________________________________
> Visit the Archives at
> http://www.meteoritecentral.com/mailing-list-archives.html
> Meteorite-list mailing list
> Meteorite-list at meteoritecentral.com
> http://six.pairlist.net/mailman/listinfo/meteorite-list
>
Received on Mon 29 Mar 2010 01:53:00 AM PDT