I had to resize the images by 400%, but you can clearly see Deimos go by the sun. I owe this find to the magnificent Midnight Mars Browser, I was watching the images for February download and saw this in Sol371 Pancam...

See attached animated GIF

Eric P / MizarKey

Great find, I wonder how many more we may have missed.

That is a very cool animation! And I'm not necessarily doubting that it is Deimos that is being imaged as the size makes good sense compared to the previous Deimos images...

But it does leave inspire me to ask : how can it show up so bright when it is within 1 degree of the sun? I would expect that, as on Earth, it would be very difficult to image the moon when it were so close to the sun. Is it being illuminated by 'mars shine'? We don't necessarily know the original brightness values in the actual image, but...unless some tone curve or very very extreme gamma curve is being used on the raw images, I wouldn't expect the reflections off the surface of the moon while it is that near to eclipse to be within the range of brightness as the sun itself. (the moon, appears as brightness values of 40,43 and 60, while the sun does clip to 255 but is 230 on average at its brightest). Am I crazy?

I could be completely off my rocker here, am I just being fooled by the brightness stretching, or am I wrong about the brightnesses of eclipsing bodies?

It might very well be due to errors in the system, but...

The JPL Horizon's system ( http://ssd.jpl.nasa.gov/horizons.html ) allows for the generation of observational characteristics from different locations. Luckily enough, if you use the telnet service, you can do it from other bodies as well (the Opportunity Landing Site is included as one of the observation sites on Mars).

For Feburary 8th around 14:05 UTC Deimos seems to be at its smallest angular distance from the Sun, at 8.865 degrees away. The previous and next sols both have the smallest angular distance as larger than on the 8th, so it would appear its as close as it gets to being a transit. I'm not sure if this means it wasn't Deimos, or if I'm just doing something wrong. That day in particular being the closest it comes in terms of angular distance certainly points towards it being the day for the bi-annual transit of any given location on Mars by Deimos, but I am curious why it comes back with 8.8 degrees seperation. It's quite possible I'm doing something wrong, and I'd encourage others to try it for themselves.

I can't believe that is Deimos either. The Sun images are taken through a solar filter and there must be something like 20 magnitudes of brightess difference between the Sun and Deimos. Deimos wouldn't appear in the images even if it was at its brightest, let alone when its almost black when close to an eclipse, no matter what stretching was applied.

This is some eclipse pictures from early on in the mission

http://marsrovers.jpl.nasa.gov/gallery/pre.../20040311a.html

Deimos is considerably smaller than Phobos so I think it could well be Deimos in the animation.

What time of day were the images taken at? If noon, perhaps it is some sort of Marsshine opposition effect. When callibrated images for that sol are available, it would also be worth seeing if the "bright Deimos" is some sort of wierd effect of stretching. If that isn't Deimos, what the heck is it? A Mars-Bird?

I've created an animated gif of the March 11th, 2004 Deimos eclipse...the one where we know it is Deimos... It's attached.

There does seem to be a difference in that Deimos is completely dark and well seen against the sun's disk but not outside of it. The first frame was taken 50 seconds before the frame that actually shows Deimos against the sun, then the frames are 10 seconds apart.

Maybe the object is not Deimos, is there anyone who can reexamine Slinted's work from the earlier post to verify Deimos position?



Eric P / MizarKey

Mars shne wouldn't cut it. With the kind of filter they're using to keep the Sun from saturating the CCD, even a full Earth moon wouldn't be visible in the same frame as the Sun. Most solar filters on telescopes cut out something like 99.99% of the Sun's light. That would make a full moon dimmer than the threshold of naked eye visibility.

Also, if calculations showed that Deimos was 8 degrees away from the Sun that day, it wouldn't be in the same frame as the Sun. The Sun's angular diameter from Mars right now is only about 1/3 of a degree.

With Earth and our Moon, we do get Earthshine making the Moon visible when it is close to new moon, but even then you can't see the earth-lit part in the daylight. Other than in an eclipse, I think the closest anyone has seen the Moon following or preceding new phase was about 13 hours, and 18 hours is pretty rare.

So I don't think there's any way that you would see Deimos in the same frame as the Sun unless there was an eclipse in progress. There must be something else at work here like an internal reflection or cosmic ray noise.

The sol 371 images are all taken with the L8 filter, which is 10^6 less transmittance than the geology filters, so whatever it was, it was very bright.

I wanted to qualify something I said in an earlier post: Feb 8th wasn't the closest that Deimos gets to the sun...I was reading the output incorrectly. This obivously all assumes that the Horizon's system is correct for usage of the Opportunity Landing Site as an observation location.

The closest approach of Deimos to the sun on Feb 8th (in 10 minute intervals) is 8.9 degrees from the sun at 14:00 utc, but the closest in the near future is out in March.
March 19th at 18:30 utc, Deimos is 0.21 degrees from the sun. Since it is only giving output every 10 minutes, it's hard to tell if this is the actual sol of the transit or if it is a couple sols in the past or future. The previous sols also show very close crossings (0.38, and 0.45 degrees from the sun).

I agree that the moon would be far to faint to show up in L8. I'd say internal lens reflection. I'm guessing the pancam is fixed during these sun sequences, and that some autocropping routine sends back the centred solar image? If so, the sun is actually drifting across the full field of view during the sequence, and the shots in question may be when the sun was nearly on-axis. The "moon" drifts at approximately 1/4 degree per minute of time relative to the sun, which is the sort of rate I'd expect in this case (though there may be factors of order 2 in there?).

This same feature was also spotted during a sequence identical to the one from Sol 371 (L8 solar image every 60 seconds for 11 minutes) on Sol 362. http://www.lyle.org/mars/imagery/1P1603293...06L8M1.JPG.html is the first of the 4 frames that show the bright spot. The motion is identical (3 pixels x, 15 pixels y relative to the sun every minute or 0.22 arc-degrees per minute) to the Sol 371 images.

I'm leaning towards agreeing with fredk, that it is a camera effect (lens reflection or possibly even a single bad pixel that only appears to move due to the movement of the subframe tracking the sun). If it is a single bad pixel, then the tracking that should occur because of the sun's motion (0.24 arc-degree per minute) is very close to the actual motion, probably within the error margins. Opportunity's left pancam has picked up quite a few hot pixels along the way, evident in most of the L7 (longer exposure) shots : http://qt.exploratorium.edu/mars/opportuni...DNP2376L7M1.JPG

The only question I have left, and this one could be chalked up to pure coincidence, is why these two sequences were done the way they were. If they were looking to catch this event, it would make sense that they would use a "one filter, every 60 seconds for 10 minutes" approach. But like I said, that could be purely coincidental.

I like the bad pixel idea - I was worried about the size of the reflected image being so small. The clincher would be the cropping info that presumably could be extracted from the img files (when available) which would tell us if the "moon" is at a fixed ccd pixel.

Or, if people are out there with oodles of spare time on their hands they can check for a similar occurence on sols for which we already have the img files!