Ice Found On Mars: Wet or Dry?

[Warren Platts]

Yes, but the scientific method is not composed of only hypotheses. First you form a hypothesis, then you test the hypothesis and re-test to eliminate variables, then you form a conclusion, and finally you try to falisfy your conclusions. If you can't, you don't have proof necessarily.. but you'd have a hard time convincing anyone you didn't.

Mathematically there is always an answer. If you examine a molecule and find it is composed of two hydrogen and one oxygen atom, it's water. If you try again with other instruments and get the same result, you've proven that it is water because the math adds up and it's a fact that the chemical composition of a water molecule = H2O.

That's exactly what I'm sayin'. We haven't had any positive lab results for water so far. They're shooting themselves in the foot by going out on a limb anyway: What's the point of all the mass spectrometers and such, if all we got to do is dig a trench and take some pictures in order to declare the mission a success?

Are they setting us up for failure? Are they afraid that everything else is going to flake out? Is their thinking that they need to point to at least one positive result so that they can justify the cost of this and future missions? What's the hurry? Why announce "proof" now?

And Mr. Vegan, are you really in Iraq right now, and if so, how do you stay drunk?

[slang]

What's the point of all the mass spectrometers and such, if all we got to do is dig a trench and take some pictures in order to declare the mission a success?

Can you provide a quote that the mission is declared a success, solely on the apparent finding of water ice?

It certainly means a success, in the sense that it landed in the conditions required for the planned research, for which "all the mass spectrometers and such" were brought. And a success since the earlier observations indicating hydrogen (thus probably water) from the orbiters are confirmed.

[Drunk Vegan]

And Mr. Vegan, are you really in Iraq right now, and if so, how do you stay drunk?

My location is essentially a synonym for "nowhere" :

http://politicalhumor.about.com/cs/q...feldquotes.htm

[spacestart.eu]

Why discuss about where the ice is made of, NASA has confirmed that they have found frozen water on Mars.

[geonuc]

Over at the APOD discussion forum, one of the posters has put together a very cool composite GIF that animates between the two Phoenix 'scoop' images. Scroll down to the 11th post:

http://bb.nightskylive.net/asterisk/...37e3f77464bfd9

[JonClarke]

Yes, but the scientific method is not composed of only hypotheses. First you form a hypothesis, then you test the hypothesis and re-test to eliminate variables, then you form a conclusion, and finally you try to falisfy your conclusions. If you can't, you don't have proof necessarily.. but you'd have a hard time convincing anyone you didn't..

Usually when deal with field sciences there are multiple hypotheses, which are tested more or less simulataneously. This case is a good example. We have an observation that several clods have suddenly disappeared while others have changed shape. Hypotheses that been proposed in this discussion are photoshop, dust burying the clods, and sublimation of ice.

Setting aside the photoshop hypothesis because of its unsubstantiated conspiratorial implications we have burial and sublimation as the two main ones. There are many problems with the burial idea, starting with the fact that there are no changes in the trench morphology apart from the disappearence of the clods, and no changes outside. there also has been no evidence of strong wind in the meteorological tell tale, or signs of dust clouds in the atmosphere, or build up of dust on the lander deck. All phenomena we have seen elsewhere on mars associated with the movement of dust and other fines. So that can be set aside.

Sublimation is expected because we already know there are various ices at the martian high latitudes and the polygon nets point to ice in the shallow subsurface. Several possible ices have been hypothesised in this discussion - CO2, H2O, CH4, NH3. The temperature ranges and the relative slow (days rather than minutes or hours) sublimation eliminates all but H2O ice. This is also consistent with what we know about the martian atmosphere, CH4 and NH3 are either absent or present only as minute traces, inconsistent with a reservoir of their ices in the regolith. It is also consistent with the evidence for abundant hydrogen in the soil from neutron data.

Obviously this has to be tested by further observations. But the H20 ice hypothesis stands where other hypotheses fail so it is the presumptive explanation at this moment in time.

Mathematically there is always an answer. If you examine a molecule and find it is composed of two hydrogen and one oxygen atom, it's water. If you try again with other instruments and get the same result, you've proven that it is water because the math adds up and it's a fact that the chemical composition of a water molecule = H2O.

But in field sciences the answers are rarely mathematical, rather they are probabilistic interpretations.

Jon

[Swift]

Over at the APOD discussion forum, one of the posters has put together a very cool composite GIF that animates between the two Phoenix 'scoop' images. Scroll down to the 11th post:

http://bb.nightskylive.net/asterisk/...37e3f77464bfd9

Brilliant! In my mind that clearly shows it to be something that is leaving (sublimating) and not something being covered up by dust.

[John Mendenhall]

Looks like ice, melts like ice - oh, to have a human there to say "Bust up some of that ice for the drinks while we wait for the mass spec results."

[MaDeR]

Well, I would not want to drink this water. Who knows what is in that.

[Paul Scott Anderson]

For those who doubt that the ice is water ice, there is a good new summary on the Phoenix blog about why they have come to this conclusion:

http://phoenix.lpl.arizona.edu/blogsPost.php?bID=211

The CCD camera showed a water-type spectrum of the white patches. TEGA didn't see any ice because the sample sat out in the open sun for a few days before finally being dumped into the oven; any ice would have sublimated away already. Plus, it's currently not cold enough for CO2 ice and salts don't sublimate.

[JonClarke]

The CCD camera showed a water-type spectrum of the white patches.

Water has some very ddfinitive and downright overwhelming absorbtion bands in the near IR, so this clinches it., if correct A TEGA result is just icing in this case, IMHO.

Jon

[slang]

Water has some very ddfinitive and downright overwhelming absorbtion bands in the near IR, so this clinches it., if correct A TEGA result is just icing in this case, IMHO.

I found this post in Q&A pretty convincing too, wrt to dry ice.

[Maksutov]

[edit]if correct A TEGA result is just icing in this case, IMHO.

Jon

Nice variant of "icing on the cake": good one!

[Warren Platts]

Originally Posted by Paul Scott Anderson
The CCD camera showed a water-type spectrum of the white patches.

Water has some very definitive and downright overwhelming absorbtion bands in the near IR, so this clinches it., if correct A TEGA result is just icing in this case, IMHO.

OK, I've looked at the animations, and I am forced to admit that it looks like something sublimated away. There is dust there, but that is easily explained by the fact that the brown nuggets were dust-covered. However, if the brown nuggets were in fact water ice, then why didn't the white patches nearby that are exposed to the Sun also sublimate away? Jon, you are on record as saying the white patches are probably not the same material as the brown nuggets:

And then there's the fact that the supposed wet [as opposed to dry] ice particles that are purported to have sublimated away are in the shade. Wouldn't we expect the white stuff in the sunny portion of the photo to sublimate first?

. . . The fact that the white material in the sun has not sublimed away indicates that it is probably not ice but white minerals. You can have more than one light coloured material in a soil you realise? Of course the white minerals being salt is a hypothesis for testing, but, based on what we know now, it is a good start.

Now let's review the facts:

• The ground temperature next to the brown nuggets has not been directly measured. If I had designed the probe, I would have embedded a thermocouple on the business end of the scoop, but that's why they get paid the big bucks. And I don't want to hear speculation about what the ground temperature should have been based on calculations on Earth before Phoenix landed. The fact is, we don't know the subsurface temperature based on direct measurements on a freshly dug hole.
• The brown nuggets were in the shade in both photos. I'm assuming the shady zone is perpetually shady this time of year. Until someone produces a photo or a direct report from the Phoenix team that says otherwise, that should be the working hypothesis for our purposes. Yes, I know the lander is above the arctic circle (by 4^o). I also know that at midnight, the Sun is a mere 3^o above the horizon. Thus, I suspect the trench is deep enough so that a portion on the south edge is perpetually in the shade.
• Also, note that there were no similar brown nuggets in the sunny portion of the trench that exhibited the disappearing act. Coincidence? Or could it have something to do with the distribution of sunshine?
• The big white patches in the sunny zone do in fact appear to be sublimating, but a much more slow rate than the brown nuggets. So it is quite likely that the brown nuggets and the white patches are not the same substance. And note in case you haven't already noticed that the spectrum as measured by the CCD cameras was of the white patches, and not the brown nuggets that are the subject of the poll question at the start of this thread.
• However, salts do not sublimate at all, and they do not give off water-like spectrums. Therefore, it's quite likely that the white patches are not salts or some other white mineral, but are in fact water ice.
• Since the brown nuggets cannot be salts, either, and since the brown nuggets and the white patches are different substances, then the only other explanation for the brown nuggets is that they were Martian dust cemented by dry ice.
• As to the argument that the brown nuggets could not have possibly been cemented by dry ice: I have been roundly criticized here for not having direct experience with 0.01 bar atmospheres. Oh, but I have! I have a nice, glass thermos bottle that's insulated by a near vacuum. It works great! So the idea that CO₂ ice will last about as long as an ice cube in an oven warmed up to 140^oF is exactly the wrong picture to have, and is guaranteed to give wrong, if intuitive, answers. The reason is that the vast majority of heat transfer at a 1 bar pressure is the result of conduction with the atmosphere, as opposed to radiation, which is the only option in a near vacuum. That's why thermos bottles and convection ovens work so well. So a proper picture would be to fill a vacuum insulated thermos with water, freeze that to well below freezing temperature, then stick one ice cube on top of the frozen block within the thermos, and then stick the thermos in the warm oven, and see how long the ice cube lasts. I'll bet anything it lasts more than four minutes. This is the situation in Trench 270. The lower left is the deepest part of the east-west trending trench is perpetually in the shade, so it cannot be heated by direct sunlight. Since it lies on top of permafrost, there is much thermal inertia there. But since the overlying atmosphere is 0.01 bar, radiation is the only method of heat loss. So, conduction works to keep the nuggets cold, whereas radiation works to warm the dry ice cemented nuggets to make them sublimate. Radiation eventually won out, but it took a lot longer than naive assumptions and calculations would suggest.

So, I stand by my initial idea that the brown nuggets are not water ice. I'm convinced now that they weren't covered over by dust, and now that it's been confirmed that the white patches are probably water ice rather than white minerals, then dry ice is the only explanation for the brown nuggets. My hypothesis also explains why there were no other similar brown nuggets in the sunny zone: there were, but since they were in the sun, they sublimated before the first picture was taken.

In fact, I'm so sure now, I'll start taking bets up to $100 USD! However, due to the poll results, I'm going to demand some odds: $0.94 USD will get me $9.06 USD when the Phoenix team reverses their position on the brown nuggets. I take PayPal!

[crosscountry]

maybe next time we can send some ice with the lander and watch how it reacts.

[Warren Platts]

Better yet, we'll send a probe with oven doors that work!

[Procyan]

When salty water freezes slowly, it tends to separate into water rich and salt rich crystals. The two types of crystals sublimate at different rates. I think the salty ice goes first, leaving salt behind... the "pure" water ice takes longer

[HypothesisTesting]

ok I am going to ask a dumb question here, but if water ice sublimates, then wouldn't it be considered "Dry Ice" as well? What I mean is. Isn't dry ice is considered to be dry ice because it sublimates under normal earth atmospheric conditions, or is "Dry Ice" only solid CO^2?

If water sublimed on earth, then it would also be considered a form of dry ice. On earth, only CO2 sublimes, so that's the only one they call dry ice.

But now on Mars, they have found water subliming (or maybe technically being liquid over an undetectable range of temps.) , because the Martian pressure is about at the triple point for water (6 millibars). So a case could be made to name water ice "dry ice" if you're referring to water on Mars.

[Squink]

On earth, only CO2 sublimes

Water ice sublimes quite nicely on earth.
One way to see the results of sublimation is to hang a wet shirt outside on a below-freezing day. Eventually the ice in the shirt will disappear.
A dusting of snow will disappear over the course of days, even with temps below 0°F; that's sublimation.

[John Jones]

Water ice sublimes quite nicely on earth.
One way to see the results of sublimation is to hang a wet shirt outside on a below-freezing day. Eventually the ice in the shirt will disappear.
A dusting of snow will disappear over the course of days, even with temps below 0°F; that's sublimation.

You beat me to it.

Yes, water has vapor pressure even below the freezing point, so it does sublimate on earth.

[Nerull]

I've got an experiment for you to try, Warren.

Take a large chunk of ice, and some small chunks of ice.

Do they all melt in the same time period?

(The answer is - no)

So far you've been *sure* about everything that's turned out to be wrong.

[Warren Platts]

I've got an experiment for you to try, Warren.

Take a large chunk of ice, and some small chunks of ice.

Do they all melt in the same time period?

Actually, the answer is that it depends on the local conditions. Your point?

So far you've been *sure* about everything that's turned out to be wrong.

Everything that's ever turned out to be wrong, I've been sure about. . . .

Makes sense to me. . . .

[aurora]

The CCD camera showed a water-type spectrum of the white patches. TEGA didn't see any ice because the sample sat out in the open sun for a few days before finally being dumped into the oven; any ice would have sublimated away already. Plus, it's currently not cold enough for CO2 ice and salts don't sublimate.

Water spectrum is pretty definitive.

[Warren Platts]

Water spectrum is pretty definitive.

True, but the water spectrum is off the white patches in the sunny zone, not the "whitish" gravels found in the sliver of shade.

[JonClarke]

OK, I've looked at the animations, and I am forced to admit that it looks like something sublimated away. There is dust there, but that is easily explained by the fact that the brown nuggets were dust-covered.

Good!

However, if the brown nuggets were in fact water ice, then why didn't the white patches nearby that are exposed to the Sun also sublimate away? Jon, you are on record as saying the white patches are probably not the same material as the brown nuggets

Correct, I was mistaken. How can the persistance of the ice patches in the walls be explained. Greater thermal intertia may be an explanation. They are much larger than the clumps (which from their colour are probably not pure ice anyway) and are surrounded on five sides by a cold insulating layer of regolith. There are changes visible in the white material too, but I had dismissed those as being just due to different contrast effects. Quite possibly in error.

Now let's review the facts:[list][*]The ground temperature next to the brown nuggets has not been directly measured. If I had designed the probe, I would have embedded a thermocouple on the business end of the scoop, but that's why they get paid the big bucks.

There is a temperature probe on the scoop.

And I don't want to hear speculation about what the ground temperature should have been based on calculations on Earth before Phoenix landed. The fact is, we don't know the subsurface temperature based on direct measurements on a freshly dug hole.

Not yet, but doubless we will, all going well.

The brown nuggets were in the shade in both photos. I'm assuming the shady zone is perpetually shady this time of year. Until someone produces a photo or a direct report from the Phoenix team that says otherwise, that should be the working hypothesis for our purposes. Yes, I know the lander is above the arctic circle (by 4^o). I also know that at midnight, the Sun is a mere 3^o above the horizon. Thus, I suspect the trench is deep enough so that a portion on the south edge is perpetually in the shade.

So what? If the air tremperature is above the sublimation temperature of CO2 it will sublime rapidly, regardless of whether it is in the shade or the sun.

Also, note that there were no similar brown nuggets in the sunny portion of the trench that exhibited the disappearing act. Coincidence? Or could it have something to do with the distribution of sunshine?

I would not draw too many conclusions from the random distribution of a few small clods.

The big white patches in the sunny zone do in fact appear to be sublimating, but a much more slow rate than the brown nuggets. So it is quite likely that the brown nuggets and the white patches are not the same substance. And note in case you haven't already noticed that the spectrum as measured by the CCD cameras was of the white patches, and not the brown nuggets that are the subject of the poll question at the start of this thread.

Ice can occur in soil as white or clear segration or almost invisible traces. So the different appearance of the white ice segrations in the trench wall and the clods is not germane. if there is ice in the segrations it is quite lifely that there was ice in the clods.

However, salts do not sublimate at all, and they do not give off water-like spectrums. Therefore, it's quite likely that the white patches are not salts or some other white mineral, but are in fact water ice.

More or less correct, although some salts do contain water bands, as I know from experience. And at low enough temperatures you can get ice-salt intergrowths.

Since the brown nuggets cannot be salts, either, and since the brown nuggets and the white patches are different substances, then the only other explanation for the brown nuggets is that they were Martian dust cemented by dry ice.

Not at aall. the air is too warm fro dry ice. You have been told this mant times. Please recognise this.

Also you can have both salts and ice in small but significant quantities without the soil appwaring lighter than normal.

As to the argument that the brown nuggets could not have possibly been cemented by dry ice: I have been roundly criticized here for not having direct experience with 0.01 bar atmospheres. Oh, but I have! I have a nice, glass thermos bottle that's insulated by a near vacuum. It works great! So the idea that CO₂ ice will last about as long as an ice cube in an oven warmed up to 140^oF is exactly the wrong picture to have, and is guaranteed to give wrong, if intuitive, answers. The reason is that the vast majority of heat transfer at a 1 bar pressure is the result of conduction with the atmosphere, as opposed to radiation, which is the only option in a near vacuum. That's why thermos bottles and convection ovens work so well. So a proper picture would be to fill a vacuum insulated thermos with water, freeze that to well below freezing temperature, then stick one ice cube on top of the frozen block within the thermos, and then stick the thermos in the warm oven, and see how long the ice cube lasts. I'll bet anything it lasts more than four minutes. This is the situation in Trench 270. The lower left is the deepest part of the east-west trending trench is perpetually in the shade, so it cannot be heated by direct sunlight. Since it lies on top of permafrost, there is much thermal inertia there. But since the overlying atmosphere is 0.01 bar, radiation is the only method of heat loss. So, conduction works to keep the nuggets cold, whereas radiation works to warm the dry ice cemented nuggets to make them sublimate. Radiation eventually won out, but it took a lot longer than naive assumptions and calculations would suggest

Have you done this experiment or are you guessing?

So, I stand by my initial idea that the brown nuggets are not water ice. I'm convinced now that they weren't covered over by dust, and now that it's been confirmed that the white patches are probably water ice rather than white minerals, then dry ice is the only explanation for the brown nuggets. My hypothesis also explains why there were no other similar brown nuggets in the sunny zone: there were, but since they were in the sun, they sublimated before the first picture was taken.

What would make you change your mind?

Jon

[JonClarke]

True, but the water spectrum is off the white patches in the sunny zone, not the "whitish" gravels found in the sliver of shade.

It's the next best thing. It shows the ice is there. They could not do the clods of course, because they have disppeared.

[geonuc]

On earth, only CO2 sublimes, so that's the only one they call dry ice.

I've noticed the effects of sublimation in my freezer's ice tray many times.

[HypothesisTesting]

I've noticed the effects of sublimation in my freezer's ice tray many times.

I have seen that too, good point. (But technically, thermodynamic sublimation can only occur for a atmospheric pressure below 6 millibars for water. That is, the temperature has to change across the phase point for it to be technically sublimation. What is happening in freezer is a non-equilibrium situation where the molecules slowly ablate off the solid ice. )

[Warren Platts]

Correct, I was mistaken. How can the persistance of the ice patches in the walls be explained. Greater thermal intertia may be an explanation. They are much larger than the clumps (which from their colour are probably not pure ice anyway) and are surrounded on five sides by a cold insulating layer of regolith. There are changes visible in the white material too, but I had dismissed those as being just due to different contrast effects. Quite possibly in error.

Wall??? It looks to me like the big white patches in the upper part of the photos lie on the floor of the trench. But your contention that the white patches are subject to the thermal inertia of the regolith would also apply to the vanished clumps. Yet they did not last as long as the white patches. That's consistent with the view that the clumps were cemented by CO₂.

There is a temperature probe on the scoop.

Too bad they didn't use it.

Not yet, but doubtless we will, all going well.

Still waiting. . . .

So what? If the air temperature is above the sublimation temperature of CO2 it will sublime rapidly, regardless of whether it is in the shade or the sun.

Now I get to ask for evidence; what is your evidence that sunlight has no effect on sublimation rates?

I would not draw too many conclusions from the random distribution of a few small clods.

Just look at the animation again. There are about eight objects that do the disappearing act and they are obviously confined to the shady zone, and that is a fraction of the total size of the hole. Not only that, but the eight clumps are confined to just the lower left of the shady zone itself, as if the shady zone does in fact shrink somewhat, but not all the way. But if you insist, I could run a Monte Carlo program for you to demonstrate that the apparent nonrandomness is in fact statistically significant.

Ice can occur in soil as white or clear segration or almost invisible traces. So the different appearance of the white ice segrations in the trench wall and the clods is not germane. if there is ice in the segrations it is quite likely that there was ice in the clods.

Perhaps, but the presence of CO₂ in the clods is also consistent with the pictures.

. . . since the brown nuggets and the white patches are different substances, then the only other explanation for the brown nuggets is that they were Martian dust cemented by dry ice.

Not at all. The air is too warm for dry ice. You have been told this many times. Please recognize this. . . . Have you done this experiment or are you guessing?

The fact is no one has done any definitive experiments on the sublimation behavior of dry ice under Martian conditions. The one reference that the 01101001 was able to dig up ("Sublimation Rate of CO₂ in Simulated Mars Conditions" by White and Sears; Abstract; HTML version) came to the following conclusion:

Conclusion: It cannot be determined how different the Mars sublimation rates are from the ones found in [the] lab due to the temperature difference between the chamber and Mars;

Mainly because the lab experiment was conducted at only -10^oC.

They did obtain a rough estimate of the sublimation rate of CO₂ on Mars based on the rate of the disappearance of dry ice with the onset of summer weather on Mars:

From the abstract:

The rate is then calculated to be 450 mm/h, regardless of particle size. Our own, admittedly approximate, calculation for the rate of sublimation of the northern and southern CO2 cap during spring and summer is 0.12 μm/h for the northern cap (Planum Boreum) and 0.95 μm/h for the southern cap (Planum Australe).

That's about a 5-6 orders of magnitude difference compared to their laboratory results.

What would make you change your mind?

Here's my bottom line take as an amateur geologist as to what's going on: There are two primary cycles going on with respect to ices on Mars, there is a water cycle, and there is a CO₂ cycle. Of the two, the CO₂ cycle is predominant. This is because Mars is so cold that water--near the surface anyway--is perpetually frozen, so it doesn't move around much. CO₂, on the other hand, undergoes major swings; apparently, the Phoenix site is buried by a 1 meter thick layer of dry ice during the winter (see above reference). Thus, CO₂ is the predominant species in the atmosphere, and water vapor exists only in trace amounts. Now, as it gets colder, water will precipitate out of an atmosphere before CO₂. Thus, where they occur together, they will form two layers, with water ice on the bottom, and CO₂ ice on the top.

At the Phoenix site, there are also two layers: there is a weakly cemented upper layer of dirt that the arm is able to easily dig through that overlies an impenetrable hard pan that shows white when it is scraped. We know from the spectrometer results that the white stuff that shows through when scraped is probably water. So we can safely conclude that the hard pan layer is either a strongly cemented mixture of dirt and water, or perhaps represents the top of a solid sheet of ice.

That leaves the comparatively soft, crumbly, upper layer. After the CO₂ surface "snow" sublimates in the springtime, we may assume that there is some residual CO₂ adsorbed into the upper layer; as the summer progresses, though, CO₂ will progressively sublimate out of the dirt from top to bottom until just about all the CO₂ is lost from the upper layer. But because of the thermal inertia provided by the frozen water hard pan layer and the interference of dirt particles, we would not expect all the CO₂ to sublimate out in a matter of hours; in fact, we shouldn't be surprised if this process takes several months to complete.

We may then ask what would happen to any water in the crumbly, upper layer. Although the temperature is below the freezing point of water, water would still be lost to the Martian atmosphere during the summer due to the ice cube effect noted by geonuc and others above. But then what would happen to the water vapor? Would it later return during the next winter? Not likely. Most of it would be redeposited and thus sequestered on the permanent polar water ice caps. (So there might be an interesting CO₂/water stratigraphy in the polar ice cap, where relatively thick layers of CO₂ are separated by thin bands of water representing individual Martian years. ) Thus, the depth of the hardpan layer represents the lower limit of the upper layer through which water can sublimate during the summer and early fall. So we should expect the upper layer to be quite "dry" indeed.

Thus, the hexagonal features in the dirt visible on the surface are not due to water at all, but are due to relaxation induced by the annual egress of CO₂ during the spring and summer.

(You heard it here, folks--once again--on bautforum.com)

A prediction of my theory, then, is that the crumbly, upper layer is cemented by CO₂, and not water. So, to answer your question, Jon, what would make me change my mind is if they could get their oven and arm and everything working properly together, and then if they could take a relatively deep sample of the upper layer and run it through the mass spectrometer in a timely fashion: if it shows more water than CO₂, then I'll admit that my theory is false.

[HypothesisTesting]

Water ice sublimes quite nicely on earth.
One way to see the results of sublimation is to hang a wet shirt outside on a below-freezing day. Eventually the ice in the shirt will disappear.
A dusting of snow will disappear over the course of days, even with temps below 0°F; that's sublimation.

I agree . But technically scientists refer to "equilibrium" phase changes like melting, sublimation, etc., occuring at a phase change temperature at a certain atmospheric pressure. Like for water, it has to be below 6 millibars for the system to be in equilibrium sublimation; if non-equilibrium, any phase change is possible at any temperature , pressure like evaporation at room temp. and pressure, etc., but very slowly.

The really Alice in Wonderland point is the triple point where solid-liquid-vapor water can exist in equilibrium 0 celcius and 6 millibars.
http://en.wikipedia.org/wiki/Triple_point