[X] My Assistant

[imipak]

New work (Pelletier, McEwen, Kolb and Kirk; 'Geology', March 2008) apparently suggests dry debris is a more likely candidate for producing gullies of the sort observed by HiRISE.

This seems to me a more naturally "Occam-friendly" explanation than convoluted scenarios involving frozen aquifers with ice plugs that burst from time to time. What do the experts make of it?


(Edit: sorry for the mangled topic title, BTW; I'd blame the software tools, but I know what that'd make me )

[nprev]

Sounds plausible, but the trigger mechanism for such events also seems more problematic. I guess diurnal thermal variations in soil temperature might ultimately trigger 'sandslides', though.

We really need to catch one of these things in the act, just like other active processes (dust devils, possible hot springs, etc.) in order to figure out what's really going on. A cool Discovery or New Frontiers proposal might be an orbiter that takes little multispectral movies of interesting sites...we might get lucky.

[JonClarke]

New work (Pelletier, McEwen, Kolb and Kirk; 'Geology', March 2008) apparently suggests dry debris is a more likely candidate for producing gullies of the sort observed by HiRISE.

This seems to me a more naturally "Occam-friendly" explanation than convoluted scenarios involving frozen aquifers with ice plugs that burst from time to time. What do the experts make of it?

Beware of simple applications of Occam's razor when dealing with complex systems, and from extrapolating from one example to an entire planet.

Gullies can form by many processes, just because some of those on Mars may have formed by dry flow does not means that aall, or even most did.

The following examples I regard as being near impossible to explain without liquid water:

http://www.msss.com/mars_images/moc/2005/11/19/

http://www.msss.com/mars_images/moc/june2000/eg_crater/

http://www.msss.com/mars_images/moc/news2002/gullies/

http://www.msss.com/mars_images/moc/2006/03/29/

To pick just a few.

Jon

[djellison]

The following examples I regard as being near impossible to explain without liquid water:
...

On what basis?

[JonClarke]

On what basis?

Sinuosity, cut and fill terraces are very good evidence for fluid flow. Dendritic distribuaries are also good evidence. The source region for multiple gullies from the same stratum is strongly suggestive - dry flows are less stratigraphically controlled. Also strongly suggestive are high order (fourth order) dendritic networks of sinuous gullies.

[djellison]

So 'it looks like' then - and to be fair, that's the conclusion Malin et.al. made from MOC imagery as well. No offense, but results of CFD study of HiRISE generated DEM's holds more water (pun intended) at this stage, imho.

The problem I have with the water theory is this. Where's it coming from? To be a squiring gun for millions, tens of million, hundreds of millions of years or more - to be doing it for as long as Mars has been as it is today - surely, any such reservoirs would have been exhaust in the significantly geological past. It's a one shot activity. Squirt - boil - evaporate - gone. I don't think we can realistically expect a water cycle to put that volume back into the soil.

What's going to be most interesting (and I'm surprised we've not seen it yet) is CRISM results of the 'outflow'.

In one forum I've been called a water-on-mars nut, and on another, a dry mars conspiracist. All I'm really after is facts and data to help us find out what Mars is really like.

[mcaplinger]

So 'it looks like' then - and to be fair, that's the conclusion Malin et.al. made from MOC imagery as well. No offense, but results of CFD study of HiRISE generated DEM's holds more water (pun intended) at this stage, imho.

The pretty pictures in the UofA press release aside,
in general I'd be skeptical of computer modeling results, as they can easily be based on flawed assumptions, initial conditions, be too oversimplified, etc. For one thing, I'm not sure how accurate the DEM was with all the albedo variations caused by the flow. Malin and Edgett did not arrive at their water conclusion lightly. Branching statistics as used in geomorphology is not just "it looks like"; go back and read their papers, especially the 2000 paper.

[JonClarke]

So 'it looks like' then - and to be fair, that's the conclusion Malin et.al. made from MOC imagery as well. No offense, but results of CFD study of HiRISE generated DEM's holds more water (pun intended) at this stage, imho.

"looks like" is powerful evidence it refers to specific complex landforms. How many dry flows do you know of that produce meandering channels, but and fill structures, terraces, lateral channel migration, distributary channels, etc.?

I have not yet read the paper in []geology[/i]. I will get to it today. But as far as I can tell it is about a specific fature, not te ones I have indicated. Given the diversity of features called gullies it is unjustified to extend conclusions from a few examples to all. landscapes don't work like that. On Earth we see glullies formed by wet plastic flow, piping collapse, sapping, runoff snownelt, placial melt, spring discharge, dry flow, high density and flow density fluid flows. Many of these are likely on Mars, some are not.

The problem I have with the water theory is this. Where's it coming from? To be a squiring gun for millions, tens of million, hundreds of millions of years or more - to be doing it for as long as Mars has been as it is today - surely, any such reservoirs would have been exhaust in the significantly geological past. It's a one shot activity. Squirt - boil - evaporate - gone. I don't think we can realistically expect a water cycle to put that volume back into the soil.

Then clearly there has to be some sort of water cycle. At what rates and mechanisms remains to be seen. But we cannot ignore the morphological evidence.

What's going to be most interesting (and I'm surprised we've not seen it yet) is CRISM results of the 'outflow'.

And what will CRISM say? Spectroscoptically a wet flow will look the same as a dry flow once the water as gone. It is a very limited tool for understanding landforms as very few landforms are conclusvely linked to specific mineraology, and almost all landforms can be found in materials of a wide range of composition.

Jon

[djellison]

"Spectroscoptically a wet flow will look the same as a dry flow once the water as gone.

Assuming it's nothing but pure water. Some of the flows ARE Spectrally different - they're bright to MOC, when the surrounding terrain is darker.

I'm not saying that Malin et.al. proposal is wrong. I'm not saying the new HiRISE theory is right - I just like to see discussion, and I don't like to see people pinning their 'hopes' or 'desires' onto one 'side' of an argument. The moment one has a pet theory ( and that's what I think this has become to many ) then scientific integrity and the honesty of interpretation is slightly compromised.

Doug

[nprev]

If there is in fact water involved here; I would think that the effluvent would be very thin low-viscosity mud based on the apparently ubiqutious presence of hyperfine dust all over Mars.

Question then becomes whether hyperfines might act the same way dry in Martian conditions, and flow down a slope duplicating the effects of fluid erosion. I'm not saying it's likely, but it should be considered; and, always remember, this ain't Earth.

[mcaplinger]

Some of the flows ARE Spectrally different - they're bright to MOC, when the surrounding terrain is darker.
...The moment one has a pet theory ( and that's what I think this has become to many ) then scientific integrity and the honesty of interpretation is slightly compromised.

Brighter doesn't mean spectrally different; this could be a simple albedo change, not influencing the spectrum.

As for your "pet theory" remark: it's equally likely that refutations of a successful theory are made for compromised reasons, because it's fun to upset the apple cart.

That said, I completely agree with you that discussion is good, it'd just be better if it wasn't mostly stuff already discussed and rejected in the original paper. At least the UofA work is using new data and new simulations and should serve well to further the debate.

[ElkGroveDan]

it's equally likely that refutations of a successful theory are made for compromised reasons, because it's fun to upset the apple cart.

Boy that sure hits the nail on the head in so many different areas of technical debate that rage on today - this discussion notwithstanding. I might also add that there is a certain segment that believes that their apple-cart tipping makes them look thoughtful.

[Stephen]

Realistically, barring some serendipitious snapshot(s) by an orbiter or a passing rover the only way the mystery of those gullies is going to be solved is by carrying out detailed onsite investigations below the surface, whether by a roving robot with a (long) drill, a human drilling crew, or some other mechanism. Relying on inferences of what is going on there derived from hypotheses based on purely surface observations and/or theoretical models (coupled in either case with assorted presumptions about the conditions underneath) basically amount to making an educated guess. Such guesses may be right, but then again they may also be wrong. Or only partly right.

Not until such "guesses" can be tested by those subsurface investigations can the rest of us be certain which of those guesses (if any) is right, whether for all cases or some subset.

======
Stephen

[dburt]

New work (Pelletier, McEwen, Kolb and Kirk; 'Geology', March 2008) apparently suggests dry debris is a more likely candidate for producing gullies of the sort observed by HiRISE.

This seems to me a more naturally "Occam-friendly" explanation than convoluted scenarios involving frozen aquifers with ice plugs that burst from time to time. What do the experts make of it?...

Good discussion above. FWIW, I've published papers to the effect that if liquid water was involved, it probably was a brine and I also spoke to this effect (including frost leaching of chloride salts) at the Martian Gullies Workshop at LPI in Houston last month.

No clear consensus was reached by the end of that meeting, but wet debris flows were probably favored by a majority, with dry debris flows coming in second (with recognition that these possibilities are not mutually exculsive for different gullies). Relatively few favored a pure water scenario with subsurface aquifers and ice plugs and so on, in part because this seemingly would require localized uphill flow of water (i.e., an artesian aquifer, such as has been suggested on a huge scale for Meridiani) to explain gullies surrounding central peaks in craters and around other isolated hills. The very steep slopes on which gullies occur also favors debris flows rather than simple streams. The consensus was that if liquid water is involved, it probably represents local melting of snow or ice mixed with dust (and, I might add, ubiquitous salts).

I'm a huge fan of Occam's razor, as most here know, but recognize that it is only useful in evaluating competing interpretations of the same observations and sets of data that deal with a specific problem. The statement that "all gullies form the same way" would be an obvious oversimplification and not a useful application of the principle. You'd have to apply it to detailed observations on a specific gully or set of gullies. That seems to be what the U. of Arizona team has correctly done, with their dry debris flow model. And despite the headline, the news release does state: "They added that their research does not rule out the possibility that the images show flows of very thick mud containing about 50 percent to 60 percent sediment. Such mud would have a consistency similar to molasses or hot lava. From orbit, the resulting deposit would look similar to that from a dry avalanche." In other words, a conclusion not greatly different from the meeting's, or even from Malin and Edgett's (2000) original suggestions regarding wet debris flows with natural levees. The meeting, BTW, was convened by LPI's Allan Treiman, who first published the dry debris flow argument.

-- HDP Don

[centsworth_II]

Good discussion above. FWIW, I've published papers to the effect that if liquid water
was involved, it probably was a brine and I also spoke to this effect (including frost
leaching of chloride salts) at the Martian Gullies Workshop at LPI in Houston last month.

For those interested, here's the link to HDP Don Burt's paper from the meeting:
http://www.lpi.usra.edu/meetings/gullies2008/pdf/8035.pdf