Speaking of long-term monitoring, something caught my eye about the latest rhazcam frame from sol 2072. At first I thought the rover had shifted a bit counter-clockwise between 2072 and the previous frame, 1998. But then I noticed that the entire frame, including the underside of the solar deck, shifted to the right. So it doesn't look like the rover shifted, but instead the rhaz camera itself rotated slightly towards the left (counter-clockwise from above) by sol 2072. I would've thought the camera was mounted pretty snugly, so this is surprizing.

Here's an animation showing the movement. You can tell it's not a lighting effect by concentrating on the silhouettes of features against the bright sky:

I confirm.

In a few Sols, the camera may be laying on the ground, conveniently aimed underneath Spirit

Wait... that's not a good thing

I have tweeted the appropriate people

One could explain the solar array moving with wind - but not the whole scene. Perhaps strong winds have been moving the rover around ever so slightly and the array as well. 120deg FOV on Hazcams - a 3 pixel shift would be the rover moving by 0.3 degrees.

So, if nothing else works, by sol 20,000 we should be out of here?

I think you're on to something. If we wait a couple of billion years, the ground will erode out from under Spirit like it did for the Meridiani meteorites. Then we can just drive away with impunity.

fredk is right, It does not look like the rover is moving, but rather that the camera is turning. Since the shift corresponds to how bright the surroundings are is it possible that thermal expansion is causing the camera mount to twist to one side? (Under recent darker, colder, conditions is the shift not as pronounced?) If the shift has been monotonically increasing over time then that might indicate seasonal temperature changes or something unrelated to temperature... (what else could cause the camera mount to shift? I doubt wind could, and wind would not cause a monotonic increase)

A great idea! But if that rock under Spirit is sturdy, then by the time the ground significantly erodes, Spirit may be trapped further. Maybe she'll be up at enough tilt to get some traction though. How would the rock scraping up against the underside of Spirit fare on the rover's health in such a situation?

I've noticed these 3-4 pixel wabbles before. I'd have to get back to my other computer, but I did an animation once using Navcam images looking at the layers in HP and the presumably fixed camera position moved by 3-4 pixels during the course of about 30 images.
My thought at the time was as temperatures changed throughout the day that either the soil or parts of the rover itself were expanding and contracting slightly changing the cameras position. That was my theory at least, but perhaps the explanation lies elsewhere.
I'll find that animation and post it later.

Scott's going to have a look at it.

It's a lot easier to imagine a variety of things causing the navcams way up there on the mast to drift around a bit. But hazcam really surprizes me. Bgarlick's idea of thermal expansion could be an explanation though.

Is it possible that the hazcams are attached with some kind of gasket for vibration dampening that could be expanding and contracting after so many thermal cycles?

I'm just having another thought about this apparent movement in the Hazcams. I think that there is also a tiny movement in the left Haz images as well as the right. Maybe not as pronounced.

Just looking at the timing between these frames, perhaps someone with more time could check, but did this movement occur in the interval when they were testing whether one of the middle wheels was jammed. Didn't they do a quarter/full wheel rotation at some stage?
Perhaps this motion moved the rover a little.

Maybe, but I think we are here dealing with a movement of the camera relative to the rover itself, isn't it?

I think Scott's looking at it a bit more closely. If we've discovered that RHAZ DOES move - then there's perhaps implications for generated terrain meshes.

My brain keeps recalling something I saw in the Navcams which suggested to me some odd movement.
I've tried to find it on my computer but to no avail. Possibly deleted it.

However, going back through the Navcam images I found something similar.


This sequence taken on sol1972 (left navcam) between 13:27.17 and 18:26.41 local time shows the movement which I was putting down to the effect of thermal changes on either the soil or vehicle. The alternative suggested above maybe movement of the IDD or wheel tests - not sure about the activity going on through this sol on that afternoon.

Of course, it could all be a trick of the light, but a giveaway to me is the position of the 'hot pixel' lower right (~1/4 up, ~1/4 from right) which stays in place while the background shifts.