Re: Flashers

[Arne Henden <aah@nofs.navy.mil>]

This post is what I wanted in the first place.  You just did not
give enough information in your original posting for anyone to
give reasoned guesses.  That is what I was referring to.  You
should know by now that I am not complaining, nor "knocking" anyone.
I just want the whole story before I am asked to give an opinion.
   Either the events are real (unlikely at this magnitude, since
Tycho would have seen them), or else there is an error someplace.
They seem real, though I note that the color is identical in
all cases to prior observations, which is not astrophysically common
for outbursts.
   The _first_ thing I do anytime I see something like this is to
go back to the original images.  One cannot place absolute confidence
in software.
   On the other hand, Tom is doing exactly what he should be doing;
checking for oddities.  Those are the cases where either something
interesting is going on, or else good checks for improving the software.
Arne

Tom Droege wrote:
> I remind you all that this was first put up as "I see a lot of this 
> annoying thing, can someone tell me what it is?"  I am not making any 
> claim.  I figure this list is like the lunch room at Fermilab.  I feel 
> free to sit down and say "Can anyone explain what I am seeing today?"  
> You come here at your own risk.  If I cause you to do some work with 
> some wild error, then no one is forcing you.  You are free to stay away.
> 
> Now that the first few simple explanations don't seem to work it is time 
> for some real data.
> 
> At 02:41 PM 5/18/03 -0700, you wrote:
> 
>> The lack of reasonable explanations is probably due to lack of
>> reasonable description.
> 
> 
> It seems the nature of this medium that everything that is posted is not 
> always read.  So one has to repeat it several times.  Also due to this 
> being a mail list, it is assumed that those that read it know what is 
> going on.  This is not a publication that stands alone.  I assumed that 
> everyone reading here knew that the tass cameras take simultaneous V, Ic 
> images.   Further, I assumed that everyone knew that we require 
> simultaneous V, Ic detections for a star measurement to make it to the 
> output file.  Not good assumptions, it turns out.  ;^(
> 
> OK, I realize that one is suspicious of the original images.  For 
> various reasons, it will take me some time to dig them out.  It is true, 
> we are looking at what comes out of the end of a pipeline, and there may 
> be something in the software that produces these bumps.  However, I have 
> looked at a lot of images and we are doing things to reject the crummy 
> images.  As you shall see, I have looked at the data around the images.  
> There may be some grand common cause, but I have not figured it out 
> yet.   As the six out of 18 statistic below indicates, there are enough 
> of these to be annoying.
> 
> First some background:
> 
> I am looking at a data set that contains 9,709,345 V, Ic measurement 
> pairs of 1,851,566 stars.  Most of the measurements were taken between 
> -6 and +18 degrees declination.  The Mark IV camera is turned on at dusk 
> and scans from -4 (center of field) to +16 in six fields.  It then 
> returns to -4 and repeats the process through the night.  This covers 
> all the sky that crosses the meridian from -6 to +18 with one 
> measurement pair a night.  There is some overlap of the fields so it is 
> possible that a star will be measured as many as 4 times a night.  This 
> rarely happens, though quite a few get measured twice an evening at 
> slightly different times.  The data spans about 100 days of observation.
> 
> The data was processed by the Michael Richmond pipeline version TARP 
> 1.0.  One can get more than you want to know about this at the tass web 
> site.  I think Michael will be the last to claim that all 62,065 lines 
> of code are perfect.  But that is what I am trying to do these days.  
> Look at the result and question anything funny.
> 
> As a first look, I have selected the 18 stars with WS statistic > 1000.  
> 10 of these were previously presented and all turned out to be known 
> variables.  Two others are also probably in this category.  There are 
> six that show "flashes".  That is they show one or more measurements 
> which are brighter than the others.
> 
> The first of the six has one bright point.  It happens to be the first 
> on the list below though it was one of the last in time for this star.  
> You will note a slightly different position.  This is about a pixel on 
> the Mark IV which has 7.5 arc second pixels.  This could be explained by 
> some shift in the position computation with brightness.  Pixels should 
> not be saturated at mag 9.  The bright point is of order 2 mag 
> brighter.  One sigma for all the data is more like 0.1 mag at this 
> magnitude.
> 
> For these data, the first number is a sequential star number for the 
> processing run.  Mostly the run is sorted by RA.  Next is the number of 
> measurements for this star.  Then RA and dec in decimal degrees, 
> date/fractional_day, V, vmag, v_err, v_flags, I, ...  See the pipeline 
> documentation on the web site for details.
> 
> 783941  19 129.3065 11.7692 2452679.73662 V  9.896 0.004   0 I  9.102 
> 0.002   0
> 783941  19 129.3081 11.7724 2452609.92696 V 11.644 0.014   0 I 10.946 
> 0.007   0
> 783941  19 129.3082 11.7718 2452618.90236 V 11.585 0.010   0 I 10.965 
> 0.005   0
> 783941  19 129.3082 11.7716 2452614.91324 V 11.645 0.010   0 I 10.969 
> 0.005   0
> 783941  19 129.3082 11.7716 2452615.91063 V 11.552 0.011   0 I 10.958 
> 0.005   0
> 783941  19 129.3083 11.7715 2452683.72567 V 11.674 0.011   0 I 10.989 
> 0.005   0
> 783941  19 129.3083 11.7718 2452619.89993 V 11.611 0.010   0 I 11.000 
> 0.005   0
> 783941  19 129.3083 11.7716 2452654.80499 V 11.702 0.012   0 I 11.010 
> 0.005   0
> 783941  19 129.3083 11.7714 2452651.81306 V 11.684 0.011   0 I 11.009 
> 0.005   0
> 783941  19 129.3084 11.7721 2452617.90528 V 11.625 0.010   0 I 10.984 
> 0.005   0
> 783941  19 129.3084 11.7716 2452624.88652 V 11.607 0.011   0 I 10.990 
> 0.005   0
> 783941  19 129.3084 11.7720 2452616.90789 V 11.639 0.010   0 I 10.983 
> 0.005   0
> 783941  19 129.3085 11.7721 2452678.73941 V 11.656 0.011   0 I 10.978 
> 0.005   0
> 783941  19 129.3085 11.7718 2452643.83490 V 11.570 0.011   0 I 10.957 
> 0.005   0
> 783941  19 129.3086 11.7720 2452662.78314 V 11.662 0.012   0 I 10.940 
> 0.005   0
> 783941  19 129.3086 11.7719 2452690.70657 V 11.611 0.011   0 I 10.971 
> 0.005   0
> 783941  19 129.3086 11.7717 2452648.82126 V 11.632 0.011   0 I 10.968 
> 0.005   0
> 783941  19 129.3087 11.7717 2452666.77211 V 11.766 0.013   0 I 10.981 
> 0.005   0
> 783941  19 129.3087 11.7719 2452647.82429 V 11.593 0.011   0 I 10.993 
> 0.006   0
> 
> For comparison I paged down the processed list until I found a star 
> close to the star above and with a similar number of measurements and 
> similar brightness.  The data is below.  You will note that there is a 
> measurement of this nearby star at the same time as the flasher above.  
> It is not significantly different from the other measurements for this 
> star.  It is in the same image and at near the same RA and a degree away 
> in declination.  There appears to be nothing funny common to all the 
> data on this day/time.
> 
> 783955  18 129.3085 12.3744 2452643.83490 V 12.211 0.017   0 I 11.338 
> 0.007   0
> 783955  18 129.3086 12.3738 2452624.88652 V 12.169 0.016   0 I 11.339 
> 0.006   0
> 783955  18 129.3088 12.3744 2452654.80499 V 12.112 0.015   0 I 11.356 
> 0.007   0
> 783955  18 129.3088 12.3723 2452666.77211 V 12.398 0.021   0 I 11.351 
> 0.007   0
> 783955  18 129.3088 12.3735 2452662.78314 V 12.235 0.017   0 I 11.347 
> 0.007   0
> 783955  18 129.3088 12.3734 2452678.73941 V 12.236 0.016   0 I 11.366 
> 0.007   0
> 783955  18 129.3088 12.3733 2452679.73662 V 12.338 0.018   0 I 11.379 
> 0.007   0
> 783955  18 129.3088 12.3732 2452619.89993 V 12.189 0.014   0 I 11.346 
> 0.006   0
> 783955  18 129.3089 12.3739 2452647.82429 V 12.278 0.018   0 I 11.374 
> 0.007   0
> 783955  18 129.3089 12.3739 2452618.90236 V 12.101 0.015   0 I 11.319 
> 0.006   0
> 783955  18 129.3090 12.3737 2452648.82126 V 12.241 0.017   0 I 11.359 
> 0.007   0
> 783955  18 129.3091 12.3736 2452651.81306 V 12.329 0.017   0 I 11.395 
> 0.006   0
> 783955  18 129.3091 12.3740 2452616.90789 V 12.126 0.014   0 I 11.353 
> 0.006   0
> 783955  18 129.3092 12.3741 2452609.92696 V 12.259 0.021   0 I 11.253 
> 0.007   0
> 783955  18 129.3092 12.3733 2452683.72567 V 12.340 0.018   0 I 11.419 
> 0.007   0
> 783955  18 129.3093 12.3730 2452690.70657 V 12.200 0.017   0 I 11.343 
> 0.007   0
> 783955  18 129.3098 12.3750 2452614.91324 V 12.291 0.016   0 I 11.364 
> 0.006   0
> 783955  18 129.3101 12.3729 2452615.91063 V 12.253 0.018   0 I 11.337 
> 0.006   0
> 
> The last of the six has two high points.  For this star position of the 
> bright points seem to be within the error in comparison to the other 
> positions.  Again the points happen to come first on the list but vary 
> through time as a plot will show.  Sorry, you will have to plot these to 
> see them.
> 
> 960657  12 168.0465 2.7291 2452678.84684 V 11.016 0.009   0 I 10.243 
> 0.004   0
> 960657  12 168.0469 2.7276 2452648.92867 V 10.649 0.008   0 I  9.840 
> 0.004   0
> 960657  12 168.0470 2.7303 2452740.67971 V 11.923 0.013   0 I 11.128 
> 0.006   0
> 960657  12 168.0471 2.7293 2452763.61215 V 11.791 0.013   0 I 11.335 
> 0.006   0
> 960657  12 168.0471 2.7293 2452690.81400 V 11.917 0.014   0 I 11.165 
> 0.006   0
> 960657  12 168.0472 2.7296 2452654.91238 V 11.805 0.012   0 I 11.310 
> 0.006   0
> 960657  12 168.0473 2.7297 2452651.92051 V 11.802 0.012   0 I 11.129 
> 0.005   0
> 960657  12 168.0475 2.7300 2452677.84950 V 11.984 0.014   0 I 11.179 
> 0.006   0
> 960657  12 168.0475 2.7295 2452762.61029 V 11.995 0.014   0 I 11.267 
> 0.007   0
> 960657  12 168.0477 2.7298 2452647.93183 V 12.009 0.014   0 I 11.196 
> 0.006   0
> 960657  12 168.0478 2.7295 2452620.00721 V 11.878 0.012   0 I 11.231 
> 0.006   0
> 960657  12 168.0478 2.7288 2452739.67689 V 11.795 0.015   0 I 11.043 
> 0.006   0
> 
> This time the next star on the list was comparable.  I notice that it 
> contains points taken on the same day as the pulser above, but at 
> different time, so they appear to be in different images.  This means 
> that the bright points are in the overlap region between two images.
> 
> 960658  14 168.0471 1.6509 2452648.92698 V 12.650 0.023   0 I 11.916 
> 0.009   0
> 960658  14 168.0474 1.6509 2452763.61046 V 12.497 0.021   0 I 11.934 
> 0.009   0
> 960658  14 168.0475 1.6501 2452677.84781 V 12.388 0.019   0 I 11.796 
> 0.009   0
> 960658  14 168.0475 1.6513 2452643.94062 V 12.459 0.021   0 I 11.859 
> 0.009   0
> 960658  14 168.0476 1.6511 2452654.91068 V 12.536 0.020   0 I 11.871 
> 0.008   0
> 960658  14 168.0477 1.6512 2452755.63427 V 12.463 0.023   0 I 11.826 
> 0.009   0
> 960658  14 168.0479 1.6508 2452762.60860 V 12.536 0.021   0 I 11.839 
> 0.010   0
> 960658  14 168.0479 1.6514 2452651.91881 V 12.554 0.020   0 I 11.861 
> 0.009   0
> 960658  14 168.0480 1.6504 2452690.81231 V 12.536 0.021   0 I 11.859 
> 0.009   0
> 960658  14 168.0480 1.6510 2452678.84514 V 12.549 0.021   0 I 11.873 
> 0.009   0
> 960658  14 168.0481 1.6514 2452679.84233 V 12.402 0.020   0 I 11.873 
> 0.009   0
> 960658  14 168.0484 1.6515 2452647.93013 V 12.560 0.021   0 I 11.860 
> 0.009   0
> 960658  14 168.0486 1.6502 2452620.00539 V 12.459 0.018   0 I 11.867 
> 0.008   0
> 960658  14 168.0491 1.6497 2452666.87784 V 12.719 0.033   0 I 11.930 
> 0.011   0
> 
> OK, any comments?  Yes, I know, more data is needed.  More data is 
> always needed.  ;^)  I am looking at a lot of this stuff.  There is no 
> good way to transfer it to you until I can get it into shape to put on 
> Mike Sallman's data base.  But I want to find problems first, that is 
> why I am making these posts.
> 
> Tom Droege
> 
> 
> 
> 
> 
>