It's equal parts -- can't remember if that's by weight or by volume, but I think it's by volume -- of Lincoln 60 fire clay and food-grade diatomaceous earth. (FGDE is normally used for, among other things, de-worming horses and killing centipedes. I tasted it. Bleah.)

One entertaining afternoon a few months back, when we were testing out different mixes, fellow rover driver Paolo Bellutta and simulant designer Kim Lichtenberg (the mix is called "KimSim" :-) and I drove out to a local ranch, picked up a huge bag of food-grade diatomaceous earth, and drove back to Lab. Later, I was up to my elbows hand-mixing a batch of the stuff in a wheelbarrow. Ah, the things I get paid to do!

why aren't there six more of these things wandering around by now?

Because I can drive only two at a time. When these two are done (gods forbid), then it's time to send two more. :-)

Using the arm to help isn't completely off the table, but pretty close, largely for reasons you conjectured about in your post. First, we can't actually push while driving, because the motor controllers are shared between the arm and the wheels -- you can run one or the other, but not both at once.

We could, potentially, push down with the arm to lift Spirit slightly, then run the wheels. But Spirit's just not strong enough to make much difference. :-) In the best case, we can push down with maybe 70N of force, and that's if we had a hard surface to push on. (But if we had a hard surface to push on, we probably wouldn't be mired in this stuff in the first place.) For comparison, you'd need to apply ~ 650N to completely lift Spirit, so the arm can apply only about 1/10 of the needed force. As you can see, she wasn't designed to do one-handed push-ups. :-)

Further, doing so would pose a high risk of damage to the arm itself, and since four of Spirit's science instruments -- about 2/3 of the total science payload -- live on the end of that arm, the potential downsides are quite significant.

In addition, it's not completely clear that pushing down with the arm to partially lift Spirit would actually help: one effect of that would be to reduce the traction on the wheels, and not having enough traction is one of our big problems here.

Resculpting the terrain is a less unlikely scenario, but something we're keeping in our back pocket for now. There are few suitable rocks within reach, we've never tried it and (again) would risk damaging the arm by doing so, and on top of all that we don't even know if it would actually help, since the rocks might simply slip quickly under the wheels without moving us forward much. Even so, if things get desperate enough, we might possibly try that one.

The soil we're stuck in is very weird, and has some counterintuitive properties. It doesn't work like dirt or mud. We mixed up a batch of simulant to drive our test rover in, and while there are known differences between the simulant and the real soil, the experience of working with the simulant is quite illuminating. The stuff feels like flour and flows like water: run your hand through it, and it flows away from you like water does, it just stops moving sooner. Weird, weird stuff.

I'd like to point out that the engineers designing the rovers probably expected them to last longer than that (though certainly not 5 years). They probably budgeted for 90 days to keep the projected costs down so that NASA would chose the project. They knew that the budget would be extended once the rovers were there.

A lot of people seem to believe this, but it's really not true. I'm not saying we expected the rovers to drop dead at the stroke of midnight on sol 91, but even the wildest optimists on the project did not openly dare to hope that we'd even double that 90-sol lifetime. (We've just hit twenty times that number, as it happens. Incredible!)

Also note that underestimating surface survival time doesn't significantly reduce costs. Getting through the first 90 sols on Mars cost a little over $800 million. But most of that cost goes into design, development, testing, launch (about $100 million per rover goes to launch costs alone, IIRC), and so on. Operations, by comparison, is cheap: now that they're there, we run the rovers for ~ $20 million per year. If we'd known, for example, that we'd survive a year on the surface, we could have promised NASA four times the science for a ~ 10% cost increase; that would have made the project a better sell, and we'd have been fools not to do it.

Yes, the rovers have photographed both moons.

Excellent link to some of the astronomy Spirit and Opportunity have done. Considering they were designed to be mainly geologists, the rovers have done a decent amount of astronomy (some of it not covered by that page), including observing a Phobos transit and a Deimos transit.

We've even imaged the Earth! On sol 63, Spirit took the first picture ever taken of the Earth from the surface of another planet.

I'm one of MER's rover drivers; I've been on the project from the start. Which has been considerably longer than five years, as development started about 3.5 years before landing, so MER has been the focus of my life for nearly a decade now. I co-wrote the software (RSVP) we use to drive the rovers, and I've been using that software to drive Spirit and Opportunity ever since.

As a contribution to MER's five-year anniversary celebration, I'm blogging my personal mission notes from the early days of the mission. They'll be posted in "real time" -- roughly one update per day, five years after the fact -- at http://marsandme.blogspot.com/. First update will be tonight around 18:30 (Pacific time).

Be prepared to stick with it; it's a little slow for the first few days. And be aware that it's a personal activity, not a JPL-sponsored activity, so I occasionally swear and stuff. But if you're a fan of the rovers, it will, I hope, give you a new insight into what it's been like to be a small part of an historic adventure.

Ah, and for twitterati: you can follow the official MER feed at http://twitter.com/MarsRovers; you can follow me at http://twitter.com/marsroverdriver.

Isn't this just an effect of reduced sunlight during winter? Or is Spirit near the equator / other hemisphere? I know the Phoenix shutdown is at least partly due to seasonal changes

Time of year plays into it, certainly; this would be less worrisome at the height of summer. But it's not quite the dead of winter, either (solstice was something like 150 sols ago), so it could be worse.

Spirit is a little farther from the equator than her twin sister, Opportunity, so winters hit her somewhat harder than they hit Opportunity anyway. (Phoenix is at something like 60 degrees N latitude -- much farther than either rover.) On top of that, she was pretty well dust-covered already, thanks to a previous dust storm and the regular old ongoing dust-deposition process -- dust was blocking about 70% of the light hitting her solar arrays before this storm hit.

It's possible for this to turn out to be good news. Dust storms are caused by (and, in a classic feedback loop, cause) high winds, so it's possible that the winds will actually clean Spirit off and she'll end up better than she was before. Keep your fingers crossed.