What's sea level on the moon?

More broadly. How do we express altitudes in celestial bodies without a belt of connecting oceans?

When you want to say how tall a mountain is, you say it is x meters above what?

WAG: I’d guess the landing of Apollo 11 would be used as the datum point for the moon.

Use the average diamater of the moon for that particular latitude as a reference point?

Go ahead and set benchmarks all over the planet. You could use a grid pattern, or a lat-long intersection pattern, or whatever you wanted to. I would be similar to how people in the middle of Kansas have benchmarks referenced to “sea level”. Just pick one to be “0”.

I suppose one could just take the lowest available point and call that zero.

Whatever system you come up with will be equally arbitrary. The issue, therefore, is reliability and ease of use.

For example, if you felt like it, you could measure every height in “distance from the center of the celestial body”.

That would certainly make sense, but it would result in a lot of values that are very, very similar and be irritatingly hard to use.

I’d guess that that too, and the Sea of tranquility does have parts that are pretty close to zero altitude.
However, it turns out that they measure altitude relative to the approximate gravitational shape of the moon, its geoid:

Topography of the Moon, referenced to the lunar geoid.

I’d go with mean surface level, with some fancy math for any non-sphericality.

The wikipedia article on Olympus Mons gives the height in meters above mean surface level.

Slight nit. I doubt it is referred to as a geoid, since “geo” refers to the earth. But, yes, altitudes should be given with respect to a standard, determinable reference shape.

For that matter, most measurements of Earth these days aren’t done in reference to sea level, but instead to a reference ellipsoid that is a mathematical approximation of the geoid. Most commonly used is WGS 84.

Wikipedia doesn’t include a citation for the detail, but claims:

How is this geoid measured? I was under the impression that it required some fancy gravimetry from orbiting satellites, which I don’t think we have done on the moon, or have we?.

The difficulty, or really the impossibility, of knowing the precise position of the geoid at every point is one of the reasons why reference models are generally used. Mathematical formulas can be used to convert a position measurement, taken by a radar, say, using a reference earth model such as WGS84. Whereas, actually using the geoid would require a grid of points. Refinement of the geoid determination, or even changes in the geoid then only matter in the determination of the best reference shape.

Obviously, we can do this much more accurately for the earth than any other body. IIRC, however several satellites have included mapping the Martian gravitational field, and undoubtedly we have done so for the Moon, also.

It actually is called the “lunar geoid”, however lexically absurd that might be. (See for example this NASA page on Apollo landing sites, and this paper (PDF) on lunar gravity models, where the term is used.)

Depends on who you ask, oddly enough. One common lunar geoid is defined from Clementine laser altimetry data, though I understand this is not universally accepted. See this presentation (PDF) from the USGS for a good summary of the topic.

We did a lot of sophisticated measuring of the Moon for the runup, and during, the Apollo missions. We had to, because the Moon has a “lumpy” gravitational field, and this caused problems with the missions. The final Apollo missions were absolutely crammed science projects, when compared to the first landing which was little more than a quick snatch and grab of a few samples. 16 and 17 deployed satellites which orbited the Moon and beamed back a wealth of data, the pilots left behind in the capsule while their partners ran around on the surface, spent a large portion of their time doing scientific experiments.

That doesn’t stop us from calling people who study the “geology” of the other planets “Planetary Geologists”.

Fascinating stuff.

From looking at this image:

Those red spots at 300 mGal are less than a thousandth of a g (if I am understanding the relevant wiki page right). How much can that affect an orbit? Say you put a satellite on an orbit calculated without taking that into consideration. How soon would it cause the satellite to fail? (or is it important for some other reasons?)

It’s nearly two thousandths of a moon g. I don’t know what the fluctuations will do to an orbiting object though.

Depending on many factors, the orbit could fail in as little as a few weeks. More on the topic here.

“Lunar mascons [ mass concentrations ] make most low lunar orbits unstable,” says Konopliv. As a satellite passes 50 or 60 miles overhead, the mascons pull it forward, back, left, right, or down, the exact direction and magnitude of the tugging depends on the satellite’s trajectory. Absent any periodic boosts from onboard rockets to correct the orbit, most satellites released into low lunar orbits (under about 60 miles or 100 km) will eventually crash into the Moon. PFS-2 released by Apollo 16 [ crashing 35 days after deployment ] was simply a dramatic worst-case example. But even its longer-lived predecessor PFS-1 (released by Apollo 15) literally bit the dust in January 1973 after less than a year and a half.

Wow. From your link:

Back to my OP, how do you measure altitude while you are there on the ground? Or is that just no longer necessary with accurate satellite maps and a link to a tracking satellite while you are there?

Going back to Bonzer’s quote from Wikipedia, it sounds like Mars has an actual “sea level” – that is, the zero reference is the point below which (due vto higher atmospheric pressure at lower altitudes) liquid water could exist if Mars had any. Am I reading that correctly.