PDA

View Full Version : Discussion: Dark Spots on the Moon Show a ...



Fraser
2005-May-26, 06:08 PM
SUMMARY: Only when the Apollo astronauts finally stepped out onto the surface of the Moon was the mystery of its dark patches finally revealed - they're ancient impact basins that filled with lava 700 million years after the formation of the Earth and Moon. Scientists now think this late heavy bombardment might have come from a time when the giant planets in the Solar System were changing orbits; Jupiter moved inward, while Saturn, Uranus and Neptune headed away from the Sun. The gravitational side-effect of this shuffling sent a rain of smaller objects that struck the Moon and other planets in the Solar System.

View full article (http://www.universetoday.com/am/publish/dark_spots_moon.html)

What do you think about this story? Post your comments below.

om@umr.edu
2005-May-26, 06:15 PM
There is little doubt that lunar basins formed late, but it is difficult to decide from this news report if there is any merit to the conclusion that lunar basins are evidence of changes in the orbits of the gaseous planets.

With kind regards,

Oliver
http://www.umr.edu/~om

antoniseb
2005-May-26, 08:50 PM
Originally posted by om@umr.edu@May 26 2005, 06:15 PM
it is difficult to decide from this news report if there is any merit to the conclusion that lunar basins are evidence of changes in the orbits of the gaseous planets.
Agreed.

There was a paper on arXiv a few days ago about this, discussing the computer simulations that let them come to this conclusion. I know you aren't too impressed with the physical validity of computer models, and I too think it's always good to take them with a grain of salt.

Nereid
2005-May-26, 09:18 PM
Nothing wrong with healthy skepticism, of course.

The use of simulations, however, is (IMHO) an increasingly important part of modern science. In some cases, 'simulations', broadly defined, are the only way to proceed (an example - QCD (http://www.pparc.ac.uk/Nw/qcdoc.asp)) - there are simply no closed analytic forms by which an otherwise powerful and successful theory can be further tested.

The good news about simulations is that (as far as I can see) many of the early pitfalls have been addressed (e.g. degeneracy, robust metrics for evaluating fits), so why not add simulation results to the array of testing scenarios?

Going to another level, as the instruments which we use to gather the data are themselves to be judged as much by simulation as by the underlying physics (after all, how many modern instruments have been through rigourous post-observation validation, like that used by the Hubble Instrument Physical Modelling Group (http://www.spacetelescope.org/news/html/heic0112.html)?), if we reject simulations, must we also reject much of the 'observational data' too?

dave_f
2005-May-26, 09:36 PM
Originally posted by antoniseb@May 26 2005, 03:50 PM
There was a paper on arXiv a few days ago about this, discussing the computer simulations that let them come to this conclusion. I know you aren't too impressed with the physical validity of computer models, and I too think it's always good to take them with a grain of salt.


Computer models make good supporting evidence. Of course, I always prefer real-world examples and observational data myself. ;)

antoniseb
2005-May-26, 09:46 PM
Originally posted by dave_f@May 26 2005, 09:36 PM
Computer models make good supporting evidence. Of course, I always prefer real-world examples and observational data myself. ;)
Thanks Dave,

This is pretty much my point. I will prefer a model that plays out well in a good simulation over one that doesn't. I certainly like to know what simplifications were used in the simulation. I'm also interested in knowing the starting conditions etc.

In the case of planet formation, we DO have some coarse grained observational evidence for what developing systems look like, so the computer model will need to do a good job of including these snapshots in its output without forcing it.

I am intrigued by the idea that Jupiter used to be the outer-most of the four gas giants, and that somehow the celestial dynamics allowed the other three to move out while it moved in, without a major collision (okay Uranus may have had a major collision).

Still it's a simulation, and subject to later improvements which may show some radical differences from what this one shows.

Greg
2005-May-27, 02:34 AM
This is very interesting. The problem I have with simulations is that it is more than likely possible to create numerous starting conditions by varying one or several variables and still achieve the same result. I agree that a simulation or model that explains all the end results plausibly is much more believable than on that doesn't. I am very curious and would like to see what their initial starting conditions were, as this bears directly on how likely it is to be realistic. If in fact the starting conditions they chose are unrealistic based on current observations of young protoplanetary discs, then the model instantly loses a degree of credibility. If their starting conditions are typical for other observed discs, then it is much more believable. I am also curious to learn the mechanism by which Jupiter was moved in and the other planets moved out. I would guess that since Saturn is smaller than Jupiter and revolving faster it would tend to be pulled towards Jupiter and that as a result the outer gas giants would tend to sort out by size with the smaller planets being boosted to more distant orbits. Perhaps this is a more general phenomenon with the largest gas giants always sorting out and ending up closest to the star?

Greg
2005-May-27, 02:38 AM
Speaking of what happened to Uranus that caused it to be knocked on its side, has anyone simulated what would hapen to Neptune once Triton fianlly spirals in an impacts into it?

Nereid
2005-May-27, 11:55 AM
The problem I have with simulations is that it is more than likely possible to create numerous starting conditions by varying one or several variables and still achieve the same result.
While I can't comment on this particular simulation, in general those who do this kind of work take great pains to factor this in to their work, and there are now (I think) a variety of tools and techniques to assist (Monte Carlo is one example).

Similarly, reviewers should (if they're doing their work properly) insist that the input parameters are appropriate, and choices and tests (of input parameter ranges, for example) sufficiently robust (of course, the researchers do the work, but the reviewers' job is, partly, to make sure that that work is written up properly ... and if the work is too sloppy, they should simply recommend that the paper not be published at all).

PIF
2005-May-30, 02:02 PM
So if the larger planets moved/changed orbits, what was
the earth doing? Was it also moving or remaing stationary?

And if either, why is one an exception and not the other?

I think once the concept of planets changing orbits is introduced the entire
classical model and the underlying physics begins to crumble.

This also raises the question, given some of the recent observations of unusual or unexpected activity solar system wide, if the orbits
changed once, can they change again?

Are they truly stable or are they still changing?

Just thought I'd ask.

PIF

antoniseb
2005-May-30, 03:07 PM
Originally posted by PIF@May 30 2005, 02:02 PM
Are they truly stable or are they still changing?
Hi PIF,

They idea is that they changed orbits because they were interacting with material in the proto-planetary disk, which is now mostly gone. They might be changing orbits a very little bit now, but only microspcopically compared to what they did when the Kuiper Belt and Oort Cloud were being populated.