#: 509632 S1/Weather 30-Dec-93 11:16:53 Sb: #Ice, wtr & supercooling Fm: Ed Williams [LVK CA] 72347,1516 To: John Deakin 76702,310 (X) John: I've moved the thread from "When to use Carb Heat" since this a Physics of Weather lecture coming up... First some basics to make sure we're on the same wavelength. Water vapor, liquid water and ice are all made out of the same identical water molecules. At any temperature, these molecules are in constant motion, at high temperatures rapidly, at low temperatures slowly. The physicist's definition of "absolute temperature" is just the mean kinetic energy of the water molecules measured in convenient units. At short ranges there exists a fairly weak attractive force between the water molecules, so there is a for them tendency to clump together, with this clumping tendency being opposed by their thermal motion. At sufficiently high temperatures then, clumps never build beyond a few molecules and the water forms a gas that fills whatever container it is put in. As the temperature is lowered fewer molecules have the required "escape velocity" to leave the clumps, allowing macroscopic regions of condensed water to form. For subtle reasons, there are several forms that condensed water can take. First, liquid water, where the molecules are still free to move past each other and are somewhat randomly arranged, and second, solid water, which itself can in several crystalline forms, where the water molecules are locked into a crystal lattice like neatly stacked billiard balls. It is possible for different "phases" of water to co-exist. Consider a sealed jar partly filled with liquid water and held at a fixed temperature. At any given temperature the water molecules have a distribution of energies (their mean energy *is* the temperature, remember). The more energetic molecules in the liquid surface can exceed escape velocity against the attraction of their neighbours and evaporate. Molecules in the vapor above collide with the water surface, are captured and condense. After a while the processes of evaporation and condensation come to an equilibrium and the water vapor above the liquid surface is now *saturated*. Equal numbers of molecules are evaporating and recondensing. The rate of evaporation is controlled by the temperature, as this determines the fraction of the molecules in the liquid that can escape. OTOH, the condensation rate is controlled by the pressure of the water vapor, as this is a direct measure of the rate at which the water vapor molecules whack into the liquid water. The net result is that the temperature determines the so- called "saturation vapor pressure" of the water. At 0C this is about 7mb, at 35C it's 50mb, at 100C 1013.2mb, roughly doubling for every 15C rise in temperature. Be patient, we'll get to supercooling eventually... Before we go on let's relate this to some well known weather facts. 1) "Warm air can contain much more moisture than cold air." (If you've been paying attention to my previous railings on this subject, you'll know that this is not quite correct. The *air* has nothing to do with it. The temperature determines the saturation vapor pressure of the water, independent of any additional air pressure.) 2) Water in an open container boils when the saturation vapor pressure equals the atmospheric pressure. The boiling point decreases with pressure, so be very careful opening your Thermos of coffee at FL250 in an unpressurized plane... 3) If you cool unsaturated air (ie where the actual vapor pressure is less than saturated) you can saturate it. The temperature at which this occurs is called the dewpoint (wrt liquid water/water vapor equilibrium) and the frostpoint (wrt ice/water vapor equilibrium). We'll get back to this. Back to the lecture... We can imagine our sealed container at constant temperature, this time containing ice and water vapor. If you wait, condensation and sublimation will come to equilibrium and the water vapor will come to saturated pressure which again is determined by the temperature. It will NOT however be the same saturation pressure as for the water/water vapor equilibrium because the binding energy of water molecules to ice is not precisely the same as to liquid water. It is in fact somewhat less, that is, per unit area, ice will sublimate slower than a water drop at the same temperature, even though the condensation rates are the same. This will be a key fact when we move onto the process of cloud "glaciation", coming up soon. We can look at the third possible equilibrium between water and ice. Because both are relatively incompressible, it takes a lot of pressure to affect the equilibrium temperature, so for atmospheric physics purposes the melting temp is always 0C. There is an measurable effect though. In fact increased pressure lowers the melting point- which is how ice-skates work. Below 0C, water droplets and ice crystals can *separately* exist. However if they are placed in contact, the water would immediately freeze. The water is said to be supercooled. It's an unstable equilibrium, like balancing on a point. Similarly, if you rapidly cool some water vapor in a superclean container it can become supersaturated. Introduce water drops or other convenient condensation nuclei and it immediately condenses. In the mean time though, it can remain supersaturated for the following reason: a very small droplet evaporates faster than a larger one because of its large surface to volume ratio. So starting from scratch, with no droplets and no condensation nuclei, it is hard to get enough molecules to stick together for long enough to get the condensation process going even though with a critical mass available the process would go like gangbusters. So what's the story in below freezing clouds? (I may not have this all quite straight, as it's even more complicated than the version I'm providing here.) Suppose we lift some moist air. It cools and saturates and it's below freezing. Even though ice crystals are ultimately more stable, it's easier for a small water droplet to get organized and grow, starting maybe on dust particles or suspended salt from the ocean. So we first get a cloud consisting mostly of supercooled water droplets. With rapid uplifting, as in a CB, the drops will grow considerably and may lag the ambient temperature. Fairly soon, some ice crystals will form on especially suitable nuclei (silver iodide is very effective and is used for cloud seeding). With all three phases present (water vapor, water drops and ice crystals), the situation is unstable. Remember that water drops evaporate more rapidly than ice crystals sublimate at any given temperature. Because of this ice crystals grow rapidly at the expense of the water drops and either drop out as snow or remains suspended. The cloud has now "glaciated" and represents a greatly reduced structural icing hazard. So return to your original picture. In the case of freezing rain and big CBs, it is probably correct that the water drops are warmer than the surrounding air and may in fact be above freezing. They freeze on contact with your airframe which *is* at the ambient, below freezing, temperature. This is not the case though in your typical stratus cloud. early in its life cycle, prior to glaciation, the water drops are supercooled, below freezing. I'm not a long message writer (like you ) so I'll quit here. Ed There are 2 Replies. #: 509972 S1/Weather 31-Dec-93 04:26:33 Sb: #509632-#Ice, wtr & supercooling Fm: John Deakin 76702,310 To: Ed Williams [LVK CA] 72347,1516 (X) I'm going to have to read that one several times more. I had no idea the process was so complex, and my poor mind boggles at the thought you have simplified it! Thanks very much for that tutorial, which you ought to upload to the libs. I'd do it, but it should have your UID on it. You are one wonderful resource, Ed. Happy New Year, while I'm at it! There is 1 Reply. #: 510156 S1/Weather 31-Dec-93 10:27:49 Sb: #509972-Ice, wtr & supercooling Fm: Ed Williams [LVK CA] 72347,1516 To: John Deakin 76702,310 John: I don't think the simplifications miss anything essential, but if they do someone here will undoubtedly straighten me out. The complication I had in mind that affects the dynamics of real clouds: the distribution of drop sizes etc, is the role of collisions between the drops. This is however pretty much independent of the basic thermodynamics of ice/water/water vapor I was trying to describe. Happy new year to you too! Ed #: 510088 S1/Weather 31-Dec-93 08:43:56 Sb: #509632-#Ice, wtr & supercooling Fm: Paul Nafziger (MHV) 73007,3141 To: Ed Williams [LVK CA] 72347,1516 (X) Excellent explanation, Ed. So how about explaining the following? I spent my youth on a farm in central Illinois. We kept a bowl (metal) of water outside for the pets year round. I walked past it every very early morning as I went to do chores for about 10 years. Twice, I noticed that the water was unfrozen even though the air temperatures were extremely cold (teens). The time was early morning and the air was extremely still. The water was obviously supercooled. When I tapped the container, the water froze instantly, almost explosively. Really interesting, but I have no idea of the mechanism. Ever hear of anything like that? Any idea why jarring the water would cause the change of state? Naf There is 1 Reply. #: 510155 S1/Weather 31-Dec-93 10:27:42 Sb: #510088-Ice, wtr & supercooling Fm: Ed Williams [LVK CA] 72347,1516 To: Paul Nafziger (MHV) 73007,3141 Paul: Your pets must believe in licking the bowl totally clean. This supercooling stuff is a threshold phenomenon. An ice crystal needs to be a minimum size before it will grow, and it can take a very long time for the minimum size crystal to appear spontaneously. So in practice the ice requires some sort of nucleus to grow on: a dust speck, a scratch on the surface of the bowl or similar. Tapping the bowl presumably dislodged such a speck catalysing the freezing of the whole bowl. Another thing that would have worked is to have dropped a tiny snowflake into the bowl. Back in chemistry class, I remember similar stuff with crystal growing. You have some solution which is supersaturated with solute, but the crystals won't appear. The tricks were to scratch the side of the vessel with a glass rod, or to find a crystal of the material (maybe from someone more successful in the lab than you ) and to drop it in. The interesting wrinkle with cloud physics is that there are typically many more nuclei available from promoting water droplets than ice crystals, so except at very low temperatures, clouds form as water drops first. However, as I explained earlier, as soon as there are some ice crystals around in addition, the combination of water vapor, water drops and ice crystals is unstable, with the ice crysals growing at the expense of the water drops which evaporate. So when it snows out of a wet stratus cloud, at no point did the water drops freeze. What happened is they evaporated while the snow crystals grew at their expense with the transfer through the medium of the water vapor. Ed #: 510156 S1/Weather 31-Dec-93 10:27:31 Sb: #509972-Ice, wtr & supercooling Fm: Ed Williams [LVK CA] 72347,1516 To: John Deakin [76702,310] John: I don't think the simplifications miss anything essential, but if they do someone here will undoubtedly straighten me out. The complication I had in mind that affects the dynamics of real clouds: the distribution of drop sizes etc, is the role of collisions between the drops. This is however pretty much independent of the basic thermodynamics of ice/water/water vapor I was trying to describe. Happy new year to you too! Ed #: 510328 S1/Weather 31-Dec-93 17:39:12 Sb: #510155-Ice, wtr & supercooling Fm: Paul Nafziger (MHV) 73007,3141 To: Ed Williams [LVK CA] 72347,1516 Hmmm. Makes sense, except that bowl wasn't all that clean. Maybe that's why I only encountered it twice in ten years. I haven't studied cloud physics for almost 30 years and had forgotten much of what you've discussed. I don't recall ever hearing about snow crystals growing at the expense of water droplets through evaporation. Makes a lot of sense, though. Is this fairly recent theory or did I sleep through one of my classes? Naf #: 510371 S1/Weather 31-Dec-93 18:50:16 Sb: #510328-Ice, wtr & supercooling Fm: Ed Williams [LVK CA] 72347,1516 To: Paul Nafziger (MHV) [73007,3141] Paul: I think the glaciation stuff is pretty old, but it might not be thirty years old. I didn't yet have my head in the clouds at that point. Ed