The Occultation of a 13.3 Star by TNO 2012 YF12 (but C2A says r=14.9)

Apr 9, 2025 Wed eve at 10:58:30pm +- 18s 1-sigma error on time, for CA coast.

OWc page

kmz file , click on link to open up GoogleEarth with path and limits. This path is the one which Jose has more confidence in, vs. some of the other paths w/o the new astrometry.

This is a high value TNO occultation, subject of a PhD thesis... José María Gómez-Limón Gallardo and I am a PhD student at IAA-CSIC in Granada, Spain. This email is to inform you about the TNO 2012 YF12 occultation opportunity that will be visible from Hawaii and USA West Coast on 2025-04-10T05:57 UTC, see attached files. This is a target of my thesis project. Since I am targeting smaller and hence fainter targets, the uncertainty region was originally higher than usual. However, the use of VLT high precision astrometric data allowed us to considerably narrow down the shadow path.

The aim of the campaign is to get at least one positive detection so that the ephemeris of the object can be further refined. Since the object is embedded in a dense stellar field, new occultation opportunities will arise (and this time with smaller uncertainty and greater chances for more positive chords).

If it is possible for you to observe this event, the data would be of GREAT value for my research. The attachment pdf document has more details to prepare your observation. Please report your intention to observe preferentially through the Tubitak occ portal https://occultation.trgozlemevleri.gov.tr/ or occultwatcher website, as I will be monitoring those. For any further questions do not hesitate to contact me. In order to streamline the analysis process, I kindly request that you provide your data in FITs images when you can (CMOS, CCDs), video is ok for video cameras. To make this process as smooth as possible, you can upload your data to the Occultation Portal used by our Lucky Star collaboration ( https://occultation.trgozlemevleri.gov.tr/ ). Include dark and flat calibrations in your data when possible. If you are new to this portal or have not used it for a long time, here (https://occultation.trgozlemevleri.gov.tr/media/docs/Quickguide_OccultationPortal.pdf ) you have the quickguide on how to use it

     

 

Alt=17.5 Az=256 in Monoceros about 13 degrees due left of Betelgeuse. Max duration = 12.3 seconds

The moon is 75 degrees away and 93% big, so the sky will be bright and with low altitude, even brighter. This is a tough event and we may not be able to preserve sky contrast when we get to a high enough integration. What may help is to take out the focal reducer. I don't think lowering the gamma will help, in a bright sky it'll just make it worse.

The actual magnitude does not look like 14.9, nor 13.3 as I saw in the LS predictions. It looks betwee, at 13.8. That should be quite do-able if we can have clean skies and integrate at 32x. Looking out over the ocean will help. So I will be setting up at the bike crossing "Twin Gates". Kirk plans to be at "The Berm". I hope he's careful about the horizon. The star is only 17.5 degrees up at event time.

This photograph indicates the visibility of the target star is much like the r=13.7 mag star in the small curve of 3 stars nearby.

    

 

Results:

Richard Nolthenius

I set up at the cul-de-sac of Pine Ridge Rd in Bonny Doon. During the event time, the sky over the target was pretty clear, otherwise significant cirrus around the sky. Low alt=17 deg and 93% moon made for bright sky. I integrated at 64x as a compromise. I could see the star on nearly all the 1-sec integrations. I removed the f/3.3 reducer to lower sky brightness, and the drift of the scope was a problem. PyMovie lot tracking on each of the moves of the scope, no matter how small. However, I did get about a solid minute including the predicted time, when tracking on star was OK. A suspected occultation at 10:58:44 lasting only perhaps 2-3 seconds, but low confidence, and Kirk Bender only a transverse distance of a mile or so away, saw nothing at 128x at that time. Probable noisy miss.

My PyMovie OCR's were contaminated by the very bright star seen in the finder. However, there were no dropped frames, and by using 'manual time insertion' in PyOTE I will be able to recover the time stamps for all frames. I will use the 'finder' feature to look separately at each integration's 32 frames and damp out sky noise and see if either of my dips might be more convincing. Not done yet as of 4/12/25.

A 'fourier finder' stack of 77 frames beginning at 5:57:02 UT, showing the target star, and how the OCR'ing of the times was contaminated at times by the very bright star at bottom.

Play back on Windows Media Player and a stop, and screen capture, 10s after the drop point seen in my PyMovie data.

Playback on Windows Media Player and a stop and screen capture at the dip in my light curve at 5:58:44 UT, un-enhanced just as image at left. No target star seen.

Same screen capture, but using Photoshop 'curves' and AstronomyTools "Local contrast enhance" to amplify any faint trace of the target star, but none is seen.

  

Let's do a whole sequence of 64 frame 'finder stacks'. 64 frames is 2x the integration rate,  The finders are definitely quieter in sky background noise. 64 frames x 2/50s per frame = 2.56 s per finder, The integration rate was 1.28s per integration in the Watec 910hx. So by starting my finders at 1s intervals there will be overlap from one image finder to the next, below. Hence the satellite crossing shows on both of the first two images for example. Here is the image sequence. The time stamp put onto the image at the start of the finder stack is what is on the image itself. Since the output time is actually applied at the END of the integration, it's a little confusing as to what is the representative time at the middle of these finders. From 5:58:39 to 5:58:49 UT. The target is definitely there at the beginning and the end of this sequence, but not so clearly there in the middle... I didn't do any post-processing of the finder images to change contrast or levels, as that would obscure the effects of any clouds that may have come through. The target appears to be missing from the 5:58:44.225 image and the 5:58:45.225 image, and faintly back in the 5:58:46.225 image. Suggesting a 2-3 s event. We were nominally north of the predicted line, but if there was an event for me, that would indicate the path just nicked me on the south side of the shadow, not the north side, since Kirk's data did not have a hint of this. 3km separation is pretty small and would indicate the shadow path was at Kirk's track. But that's a lot of supposition at this point. Roger Venable has volunteered to run it through LiMovie updated, and see what he can see.
       

Here's Roger Venable's judgement after processing this in the modern updated LiMovie....

Venable's notes about Nothenius's video of the 2025-04-10 event by 2012 YF12


This video appears to have been recorded with a Watec 910HX camera at 64x exposure setting, so that blocks of 32 frames are seen. On the middle 3500 frames video -- that is, with the early and late frames that could not show an occultation based on their frame times -- the point at which the occultation star crosses the darkest line of the horizontal background artifact in the field of view is ~frame 1440. This frame is stamped at 05:58:18, which is within the 1-sigma time uncertainty (of 17.8 seconds) of the predicted time at Richard's location, which was 05:58:30 UT.

The 140 seconds-duration central part of the original video contains 70 seconds before the predicted central event time and 70 seconds after the predicted central event time. The 140 seconds video contains 4 sigma of time error before and after the central event time, and therefore the parts of the original video that are before and after this central 140 seconds do not need to be analyzed. The 140 seconds video was split into two parts to allow Registax 5.1 to stack them, as that stacking program is unreliable for very long videos.

The occultation star is right at the limit of detectability in the video, so that it is visible on some frames and not on other frames. I identified it for certain by making a "screen shot" of the first video frame of the 140-seconds central part of the original video; then finding the occultation star by using its coordinates in the Guide9 sky software; then labelling nearby stars on both the video frame and the Guide9 view, and using Photoshop to copy, rotate, and scale the two views so as to superimpose them fairly accurately. (This method did show some slight distortion of the field of view, but it's not worthwhile to figure out where it came from.) I am attaching an arrowed copy of the first frame of the 140-seconds sequence that I analyzed, designating which star is the occultation star. I am certain that this is correct.


During the recording, the observer (Richard Nolthenius) made comments that the star seemed to have briefly disappeared, and stated he was not certain about it. These few comments were made as a star crossed a dark, artifactual, horizontal line in the video field. The star crossing that artifactual line at that time was not the occultation star. I point this out so that the comments will not cause any confusion.

The video exhibited considerable star drift and also some wind shake, and to prevent this from affecting the analysis I aligned the video using Registax 5.1. Because Registax sometimes has difficulty with large files, I cut the video into two halves before aligning each one separately. Further processing is described in the following paragraph.

The stars appear to be out of focus. This may be due in part to the low altitude of the event in the sky. The seeing was poor, so that many faint stars suffered from intermittent visibility on the video. The large sizes of the star images make it possible to blur the video slightly to make star images stand out from the background better, especially with contrast enhancement. I increased contrast by a factor of 2, and added a Gaussian blur with a 1-sigma radius of 2 pixels, and I think that the star images are better after these changes. The useless black pixels that remained in the background were then lessened in their effects by reduction of brightness to the brightness level of the original video
.

Large star images sometimes become more stable upon reduction of the image scale by combining pixels. Therefore I produced 2x2 reductions in the frame size using the "high quality" reduction algorithm of VirtualDub. I plan to analyze the event recording both with and without such frame size reduction to ascertain whether either method yields clearer results.

RESULTS: Neither the original video modified only by registration, nor the videos with contrast enhancement and brightness changes, nor the videos with reduced sizes of the frames show anything that can be considered to be an occultation. The severe noise in the light curves is exaggerated by the faintness of the star, which faintness causes the range of brightness to be small, thereby exaggerating random noise in the light curve as it is diagrammed as though stretched over the entire range of the ordinate.


This should properly be called a "no observation", as the star was too faint to show up reliably. Any aperture (such as Limovie uses) or aperture analog (such as AOTA and Tangra use) will have misleading results because the brightness values to be measured are not reliably brighter than the background noise. Unfortunately, the occultation event was beyond the capability of the instruments that were used.

I am attaching some light curves measured from the video by Limovie. The consistency of the brightnesses within each 32-frame group attests to the accuracy of Limovie. The exaggerated wide swings in brightness from one 32-frame group to another attest to the star's intermittent visibility due to mixing with the background noise.

-- Roger Venable 2025-04-14

 

 

Kirk Bender

Observed from "the Berm" on Empire Grade in Bonny Doon, 3km south of my station at Pine Ridge Rd. He observed at 128x on his Watec 910hx EIA with IOTA VTI, 8SE telescope at f/3.3.

I don't see an event in my data, but at 128x it is very chunky. I don't see a drop at 5:58:45 like yours. I did use the f 3.3 reducer. I was adjusting gain, etc. and started perhaps later than I should have, at 5:57:26,  but I recorded for 3 min 40 sec, so at 128x I could get more points for a noise baseline. In pymovie I did a TME and a nest of several static apertures. attached are graphs, Pyote1 is not block integrated, the apertures are from TME at the bottom, then static 2.0 to 5.1 then a reference star at the top. Red dot on the left is at predicted 5:58:30, dot on the right is 5:58:45. Pyote2 is block integrated, unsmoothed. Pyote3 is no-star at the bottom, then aperture size 2.4 smoothed, then reference star. The detectability test said: "An event of duration 4.300 seconds with magDrop: 8.0 is likely detectable." (Max predicted was 12.32) Doesn't look like there's any event, but I had pyote try to find one anyway, and it didn't find any, just gave the message:"No event fitting search criteria could be found." This is specifying min/max of 2/10  or 2/5. So doesn't look like an event, but prediction wasn't very precise, and we very well could have got a miss. I'll upload my recording to the Lucky Star portal anyway, and let them examine it. I did take a dark frame at the time. Too cloudy this evening to take a uniform twilight flat.