This is a very bright event, but very short. Essential it be recorded at 1x to get decent timings. The rank is high, given the uncertainty limits. Cabrillo Observatory is inside the path. Santa Cruz is north of the narrow path, and Karl is far north and out of the path. I'm planning on trying it from Cabrillo Observatory, if clear. If fog rolls in, it'll be hard to get above the fog, as the path slopes NE and doesn't get very high in the hills of rural Aptos.
Alt=26, Az=211 in Capricorn, 10 degrees below the right corner of the crescent.
|
|
3 Successful positives from me, Kirk, and Bernard. Bernard and I were at Cabrillo Observatory.
I observed from Cabrillo Observatory with the 8SE, from the corner of the large storage building. 11 feet away was Bernard Huynh, and play back showed he had a short positive as well. My positive came at the predicted moment, differeing by less than 0.2s and certainly within the prediction times which are printed in OW rounded to the nearest second, unfortunately.
magDrop report: percentDrop: 63.4 magDrop: 1.090 +/- 0.518 (0.95 ci)
DNR: 2.48
D time: [03:32:30.1896]
D: 0.6800 containment intervals: {+/- 0.0085} seconds
D: 0.9500 containment intervals: {+/- 0.0422} seconds
D: 0.9973 containment intervals: {+/- 0.1108} seconds
R time: [03:32:30.2696]
R: 0.6800 containment intervals: {+/- 0.0085} seconds
R: 0.9500 containment intervals: {+/- 0.0422} seconds
R: 0.9973 containment intervals: {+/- 0.1108} seconds
Duration (R - D): 0.0800 seconds
Duration: 0.6800 containment intervals: {+/- 0.0141} seconds
Duration: 0.9500 containment intervals: {+/- 0.0543} seconds
Duration: 0.9973 containment intervals: {+/- 0.1287} seconds
NIE about 2.2 sigma, shown below
Note that Bernard Huynh recording at 3x finer time resolution, had an identical short drop to zero at this same time, and again at the predicted event time. So despite the less than 3-sigma NIE histogram distribution, the event looks very highly likely to be real and should be combined with Kirk Benders longer and stronger detection near the centerline.
An event of duration 0.040 seconds with magDrop: 7.0 is likely detectable.
 The sky counts here are all zero, a possible problem with the inability of the software to find this event |
 |
 |
 |
 |
PyOTE's detectability graph |
This event could only be ID'd by severely trimming the rest of the light curve. See my comments and reasoning below...
|
Bayesian probability even more obviously needed in this event, and which is not applied in PyOTE. The priors are:(1) the predicted event duration was only 0.1s, so that raises the a priori probability of a single point drop. (2) This single point drop happened exactly at the predicted event time. The two neighboring points are low but not below the bottom of the scatter of un-occulted points, but may indicate the occultation began during those two integrations. (3) Bernard H saw 3 consecutive ~zeros at the same moment, from 15 ft away. (4) Kirk Bender had a solid 27 sigma positive of the max duration from the centerline, indicating the prediction was on-target, and further suggesting my/Bernard's chord should have had a positive at the predicted time. We did; it was less than the max duration, appropriate since we were halfway to the path edge.
No probability in PyOTE was associated with the depth of the occultation, which was 9 magnitudes for this very small asteroid, and yet PyOTE kept putting higher detection odds on other drops during the recording which were quite clearly only partial drops but happened with 3 consecutive similar values likely noise and within the envelope of noise distribution.
My Conclusion: I'll again make the point that I strongly maintain that IOTA reviewer Tony George's belief that "every event must stand on its own", is bad reasoning, and instead Bayesian statistics must be used in assessing the probabilities of an event being valid. The existing predictions of the event time and depth must be priors, and nearby chords must be factored in as well.
I got a 0.1 sec event for 1991 PR16, at 1x from the Moran lake parking lot, very close to the predicted center line. To avoid saturation I turned gain down to 34. There was a lot of flicker at 1x, I suppose from being so close to the beach, but I got a distinct 6 integration event even with the noise. NIE sigma distance 27.0.
magDrop report: percentDrop: 99.7 magDrop: 6.198 too much noise; cannot calculate error bars
DNR: 3.19
D time: [03:32:30.0565]
D: 0.6800 containment intervals: {+/- 0.0027} seconds
D: 0.9500 containment intervals: {+/- 0.0071} seconds
D: 0.9973 containment intervals: {+/- 0.0141} seconds
R time: [03:32:30.1658]
R: 0.6800 containment intervals: {+/- 0.0027} seconds
R: 0.9500 containment intervals: {+/- 0.0071} seconds
R: 0.9973 containment intervals: {+/- 0.0141} seconds
Duration (R - D): 0.1092 seconds
Duration: 0.6800 containment intervals: {+/- 0.0040} seconds
Duration: 0.9500 containment intervals: {+/- 0.0091} seconds
Duration: 0.9973 containment intervals: {+/- 0.0168} seconds
 |
 |
 |
 |
 |
 |
 |
 |
Bernard, running OCCULT4 to find all Quaoar occultations in our hemisphere for the next 2 years |
 |
Successful event of a tiny asteroid. Celebrating. |
 |
OCCULT4 predictions for Quaoar. No good events suitable for detection of the new moonlet we discovered June 25 this year at MIRA, alas. |
Jordan and I watching the playback, and both of us gave a little "oh!" at the moment of the extremely brief drop of the occultation. |
Bernard doing some imaging on the 12" scope, with Jordan. Climbed out on the roof to check the dome rotation |
 |
... and celebrate! |
