Visualizing the Dance of a Melting Snowflake

From NASA and the “I’m sure there’s a snowflake joke in there somewhere” department.

NASA Visualizes the Dance of a Melting Snowflake

NASA has produced the first three-dimensional numerical model of melting snowflakes in the atmosphere. Developed by scientist Jussi Leinonen of NASA’s Jet Propulsion Laboratory in Pasadena, California, the model provides a better understanding of how snow melts can help scientists recognize the signature in radar signals of heavier, wetter snow — the kind that breaks power lines and tree limbs — and could be a step toward improving predictions of this hazard.

Snowflake research is one of many ways that NASA studies the frozen regions of Earth, collectively known as the cryosphere.

Leinonen’s model reproduces key features of melting snowflakes that have been observed in nature. First, meltwater gathers in any concave regions of the snowflake’s surface. These liquid-water regions then merge to form a shell of liquid around an ice core, and finally develop into a water drop. The modeled snowflake shown in the video is less than half an inch (one centimeter) long and composed of many individual ice crystals whose arms became entangled when they collided in midair.

This visualization is based on the first three-dimensional numerical model of melting snowflakes in the atmosphere, developed by scientist Jussi Leinonen of NASA’s Jet Propulsion Laboratory in Pasadena, California. A better understanding of how snow melts can help scientists recognize the signature in radar signals of heavier, wetter snow — the kind that breaks power lines and tree limbs — and could be a step toward improving predictions of this hazard.
Credits: NASA’s Goddard Space Flight Center/ Lauren Ward


Leinonen said he got interested in modeling melting snow because of the way it affects observations with remote sensing instruments. A radar “profile” of the atmosphere from top to bottom shows a very bright, prominent layer at the altitude where falling snow and hail melt — much brighter than atmospheric layers above and below it. “The reasons for this layer are still not particularly clear, and there has been a bit of debate in the community,” Leinonen said.Simpler models can reproduce the bright melt layer, but a more detailed model like this one can help scientists to understand it better, particularly how the layer is related to both the type of melting snow and the radar wavelengths used to observe it.

A paper on the numerical model, titled “Snowflake melting simulation using smoothed particle hydrodynamics,” recently appeared in the Journal of Geophysical Research – Atmospheres.Some of the most remote places on Earth are showing signs of change, with potentially global impacts.

In 2018, NASA is scheduled to launch two new satellite missions and conduct an array of field research that will enhance our view of Earth’s ice sheets, glaciers, sea ice, snow cover and permafrost. Collectively, these frozen regions are known as the cryosphere. Over the course of the year NASA will share an inside look at what the agency is doing to better understand this critical component of our home planet.


63 thoughts on “Visualizing the Dance of a Melting Snowflake

      • David

        The endless money pit called, Taxpayer.

        I have always recognised the value of taxation and happily contributed my fair share in the UK. Indeed as a policeman in my earlier years, I benefited from the taxpayer (not much though). When I moved into business I made good money, and understood my taxes were supporting valuable public services.

        I’m an Ambulance driver now (semi retirement was awful) and I do it because I want to, not because I need to. Something about wearing a uniform I guess!

        And whilst in house NHS (National Health Service, for those not familiar with the jewel in the crown of British public services) Ambulance provision was chaotic and expensive, privatisation is no less chaotic and expensive, the difference being is that predatory private companies are profiting from it at the expense of employees wages.

        However, no matter how bad these attempts at privatisation of traditional public sector services are, this example of taxpayer money spunking is utterly appalling.

        Sadly, socialisms appeal of everything for nothing has taken the world by storm. And whilst I value scientific endeavour as much as I value responsible taxation, this type of naval gazing research is jobsworth justification and science itself needs to get a grip.

        The concept of research, for research’s sake must be stopped. We are swamped by a deluge of failed research. As a non scientist, my perception of that august profession is that anything goes. Yet the best scientists emerging from a religious and superstitious world struggled to have even the most basic of science accepted.

        And whilst I’m not a religious man, the world has turned 180 degrees, science, not religion, is the new religion, and the money exchanges Jesus objected to (was it Jesus?… religious knowledge is poor) have moved from religion to science, which is now the centre of superstition, “we just need money to prove it’s superstition”.

        Sticking snow?

        Go ask a frigging Eskimo.They know more about snow than any scientific tosser, and have more terms for snow, I believe, than scientists can cope with.


    • Whenever some guy says “Hold my beer”, you know something extraordinarily stupid is going to follow.

      This is the sort of thing that happens when a science nerd says “Hold my beer”.

  1. “A better understanding of how snow melts can help scientists recognize the signature in radar signals of heavier, wetter snow — the kind that breaks power lines and tree limbs — and could be a step toward improving predictions of this hazard.”

    Pure excuse for spending money on this. Even if you can better predict the heavier, wet snow how is that going to help? What are they going to do, go out and reinforce everything immediately? If you don’t want to suffer from the damages of this kind of snow, don’t live where it falls.

    • You can pre-position men and equipment to more rapidly respond to downed power lines.
      I’ve read that they can shut down some power lines in order to drive more power in the remaining lines, which causes them to warm up, hopefully enough to melt the snow accumulating on them.

      • In Victoria, Australia it was, presumably still is, common practice to run the generators of the McKay Creek hydro power station in the Victorian high country for short bursts to heat the power conductors connecting the station to the main power grid in order to melt snow/ice from the high voltage lines.

      • 1) Snow does not break power lines, ice from freezing rain, which has nothing to do with this article, breaks power lines.

        2) Generators spin at a constant rotation rate to maintain the 50hz/60hz standard frequency. As user energy demand increases, that creates a greater force against the spin of the generator requiring more mechanical energy (steam) to keep the generator spinning at the standard frequency.

        3) You can’t ‘push’ more current through a power line – that’s not how electricity works. Current flow is determined by user demand at the other end. If everyone ‘unplugs’ at the user end, zero current will flow.

      • JKrob
        March 31, 2018 at 1:50 pm

        3) You can’t ‘push’ more current through a power line – that’s not how electricity works. Current flow is determined by user demand at the other end. If everyone ‘unplugs’ at the user end, zero current will flow.

        Er, yes you can ‘push’ more current through a power line. You simply increase the excitation of the alternator by a small amount. The emf (voltage) of the generator increases and that generator then feeds more current, via the power line, into the overall system.

      • Well JKrob firstly let me assure you icy snow most definitely can and does break power lines.
        Secondly it is possible to load (push more current through) a power line, firstly you can bring on line (synchronise) generators which have been off line and with the cooperation of load dispatch load them and the power lines that connect them to the grid to what ever percentage of their rated load (current) you choose. All that is necessary to achieve this is that the load set point of other operating generators is lowered and/or the generators doing frequency control have sufficient control range to adjust for the additional generation.
        Finally yes if every user were to “unplug” there would be no customer load to be met but that is not a credible situation in normal utility grids. In reality generation despatch can always select which generating units are used to meet load.

      • MarkW March 31, 2018 at 8:15 am

        “They try to, but it’s based on guesses.”

        I know you are not implying that any model they would create by this study would be any better than their current guesses.

      • JKrob March 31, 2018 at 1:50 pm

        1) Snow does not break power lines, ice from freezing rain, which has nothing to do with this article, breaks power lines.

        Wrong! You obviously have never seen snow on a power line. When you get wet heavy snow falling, after a cold spell when the aluminum conductor is well below freezing, it sticks and builds up massive weight to the point it either can break the conductor, or even break an old weak pole in half. Or you get line slap when the heavy snow falls off, and the primary hits the neutral, and either melt the wire in half, or always blows the cut-out fuse at the top of the pole. I can catch it in time by knocking on the poles with a 20 pound sledge hammer, which usually assists in dislodging some of the snow build up as it happens. I operate 3-4 miles of private power line, and has happened to me on several occasions.This is my single biggest problem, usually once every winter, depending on conditions, but I usually can catch it time. As for you other utterances, I see they were mostly already addressed, but you are also wrong about those as well.

  2. The model shows that light, pretty snowfalls are natural in origin, while heavy, damaging snowfalls are definitely anthropogenic in nature. /sarc

    • ScienceABC123


      can we please have ScienceABC123 comment deleted lest it provides the rent seeking scientists any ideas.

      I’m certain they cruise sceptic sites, to get ideas, simply to wind us up.

    • ScienceABC123 and HotScot,
      The science of water surface tension…and how to make it ‘wetter’ (that is, reducing the surface tension of water using surfactants) has been thoroughly studied by detergent companies and mining industries dependent on wetting, frothing and flotation separation/concentration techniques.

      • Firefighters also use the technology for fighting fires. The surfactants are added to the water before application to the flames!

  3. This is a MODEL, right? And, looking at the article, there has been no attempt to undertake observations to see how it matches reality.

    So it’s not science. It’s graphic art.

    • Normally I am right there with you on the anti-modeling sentiment. We have actually observed the melting using gossamer capture of snow flakes and bringing them to a just above freezing environment to observe and measure. The equations used to perform this model have been published for a while (decades even).

      I am not sure how they expect to simulate the radar signature part of this though without simulating billions/trillions of them in a column and then running the known radar algorithms against that column to characterize the simulated returns. All the flakes in the column are going to be in a different state of melting and experiencing browning-like motion. Though the radar return can possibly be applied to a statistically modeled sample to get fairly close result. Even that though would require a truck mounted experimental broad-band radar to follow storms to compare the returns against the model output and actual on-the-ground snow results to completely validate.

      So this could be a good first step to a useful engineering model for weather radar engineers.

      • I’d want to see the original idea validated before starting to build a lot of expensive kit based on it.

        I’m not against models per se. I’m against the idea that models should be treated as adequately representing reality without any validation against reality……

      • Could the radar density mapping be coupled with a thermographic mapping in an attempt to tease out combinations?

  4. I can see that this research might be of use.

    My question is why is it being done at NASA? Wouldn’t NOAA be the proper agency for this type work?

    • It has aerospace weather applications to radar design, so NASA is an appropriate place for it to happen. Aviation safety is one of their original mission areas from back when they were still NACA.

  5. Why do they keep calling the place “Jet Propulsion Laboratory”?

    As far as I can tell they have not done any work on jets or propulsion in decades.

    Is it just that they don’t want to call it “Useless Waste of Taxpayers’ Money.”

    • The “jet” in Jet Propulsion Laboratory refers to rockets, as well. Their last big hit was the Mars Science Laboratory, and the landing vehicle for the Curiosity rover. Lockheed-Martin was the prime, under JPL direction.

      In terms of propulsion, they came up with the concept of the Skycrane for the last part of the landing. After atmospheric entry, the descent propulsion system fired up to slow the vehicle down. Then about 7.5 meters above the surface, the velocity was reduced to zero, and as the assembly hovered under rocket power, the Curiosity rover was lowered to the surface on a nylon tether. When that was complete, the tether was cut, and the Skycrane flew away to a crash landing at least 650 m away.

      I’m a rocket propulsion guy, myself, and have been involved in several vertical takeoff and landing projects. I know how difficult they are, and usually involve several full-up flight tests ending in crashes before the system is made to work. From what I knew of the Skycrane and existing vacuum test facilities, I couldn’t see how they could have tested it. So I spoke with the MSL Chief Engineer, and asked about full-up testing. He said they had never conducted a full-up test of either the entire vehicle, or even the propulsion system! They just system-engineered the hell out of it, and tested all of the components and subsystems (including engines) to as high a level as possible. Then they hoped for the best.

      I salute that achievement. Of course, having $2.5 billion to work with helped, I’m sure.

      • Michael Kelly,

        Very interesting! Wow, there’s modeling for you, with a lot riding on it. Not a lot of uncertainty in those models, eh?

        Yeah, $2.5 billion. Our tax dollars go to all sorts of things.

        I often think about the enormous amounts of money we spend on space travel (manned or not). I have no problems with it, but it seems to me if a choice had to be made, I would choose to try to understand our planet.

      • Kristi,
        We have spent a Trillion dollar$ US on the climate modeling equivalent of catastrophic navel lint predicting.
        I would gladly have spent that huuuuuge sum of money on something useful, like nuclear rocket powered science packages for exploration of our solar system.

    • Not as bad as those pesky Russians. Their latest rocket thingy has just been given a NATO reporting name – SS-X-32 Snowflake.

  6. So NASA have spent god knows how much money on understanding something that a keen weather watcher already understands.

    When the air temp the flakes falling through is just above 0C snowflakes start melting from the outside and so they start to melt and become sticky. Because of this when they are falling through the air if they hit each other then they stick to each other and then become larger. So by the time they reach the ground these snowflakes can be as large as one and a half to two inches across.
    Its this size of flake along with been wet and sticky is what allows them to build up snow cover very quickly on trees and wires. These people really need to get into the world outside of their computer models more to see and understand what’s going on.

    • Other way round Rob. CO2 causes global warming which then causes global cooling, global wetness and dryness, and of course global megalomania among those hungry for political power.

  7. OMG

    Scientists told us smoking caused cancer, so a campaign was mounted against smoking in the UK. Anyone who smokes or vapes now, is consigned to doing so as a social pariah.

    Then it was booze, so traditional British pubs are shutting down daily because of government legislation. But kids get ‘pre loaded’ now instead of standing in a pub with a beer, surrounded by people keeping a weather eye on what they’re drinking.

    The latest announcement on the BBC today is that coffee is carcinogenic, just as the coffee shops were reviving the buildings shut down in the booze prohibition.

    I have no doubt whatsoever that water is carcinogenic…..or will soon be declared so.

    When will this utter scientific nonsense end? It’s net effect is to polarise individuals, communities, and countries.

    Science is a wonderful institution, provided it’s occupied by, and presided over by responsible individuals who won’t allow their noble profession to be contaminated by irresponsible individuals colluding with the media to sensationalise whacko, crackpot research.

    If there was ever a time for scientific institutions to step in and halt this out of control monster, it’s right now.

    I’m sick of scientific experts, unfortunately the reasonable scientists who contribute to this blog, and all other reasonable scientists, get tarred with the same brush by the general public.

    Once again, the minority holding sway over the majority. Socialist science!

    • ” But kids get ‘pre loaded’ now instead of standing in a pub ”

      We used to get “pre-loaded” too, 40 years back. Mainly because pubs were expensive. Those prices now looks free beer in relation to what you have to pay for a decent pint of ale in the UK these days.

      We could got out and get pissed, and have fish and chips on the way home for a price that will not even get you half a pint of shandy these days.

      The price of beer in a pub is insane these days. That is why pubs are shutting.

      • Greg

        I must say, 40 years ago I never got pre loaded.

        And whilst I agree with you about pubs closing because of expensive beer, in my opinion, this has been a concerted government effort over the years to alleviate stress on the NHS from drink related events. They have done the same thing with smoking and priced it out of reach of most whilst demonizing smokers and making them stand in the rain and snow for a ciggie.

        What other section of society would tolerate, or be subject to such discriminatory behaviour?

      • Forrest

        Meanwhile, the simple pleasures of living are being sucked out of us all by holier than thou government and civil servant officials.

  8. Interesting. College kids that hear anything other than the approved cultural marxism have been modeled…..

  9. Being uncurious about the natural world around us gets us climate skeptics labeled anti-science. Don’t be anti-science.

    I can see how this has direct application to my research on radar propagation and information fusion for advanced air-traffic flow. Frankly, I can’t wait to talk to these potential new colleagues.

    • I no longer believe in science. First it was the physicists that invented the pink elephants of dark energy and dark matter to try to explain why the cosmos was expanding in an accelerating manner. Why didn’t they just say we don’t know why the universe is accelerating? However new studies put into question even that supposed truth. So where does that leave dark energy and dark matter? Hanging out on a very slim branch. then it was the medical and biological and drug studies that depended on each other and that had confidence intervals of only 2 sigma. So if a study depends on another study which depends on another study…….etc each with only a 95 % confidence interval for a false positive then eventually the studies at the bottom have no way in hell of being correct. That is exactly what has happened. A mathematician has proved that most biological and medical studies are now flawed and cant be replicated. They should have stuck to the 5 sigma standard like the chemistry and physics world. The problem was of course cost. Now we are in one hell of a mess. Climate science is even worse. Not only do they not try to replicate any climate study for its validity, climate scientists wont even let you have their data. Then they top it all off by worshipping their models as Gods. You heard it in the talks how much they worshipped their models. Even worse Pat Frank has proved that every climate model has a cloud error factor so bad that it makes them worthless for projections. Even worse than that is that the cloud scientists admitted in 2003 that there was no aircraft instruments providing measurements of the supersaturation of clouds. So what did they do? They turned to their own models. So by 2012, 4 of the best cloud scientists in the world ( 2 of them on that earlier study in 2003) never even bothered to discuss what efforts were being made on in situ (on site) measurements. They just went ahead and parameterized dozens of equations on cloud saturation BASED ON COMPUTER CLIMATE GENERATED DATA. Yes you read that right. So now they are describing reality and forging climate science based on computer calculations . THE CLIMATE WORLD HAS REALLY VENTURED INTO THE LAND OF ALICE IN WONDERLAND AND THEY ARE DRAGGING ALL SCIENCE WITH IT.

      • Overly pessimistic Alan. Science doesn’t need people to believe in it. The trick is distinguishing between what is labelled as science and what deserves the label.

    • DrSandman,
      That was what I was trying to get across earlier. It needs a lot of work to make something practical, but if they can use this to properly model a snow column under any arbitrary conditions, those of us that care about radar and flight safety can use it to make better tools. Having a radar processor that can tell you exactly what kind of snow you are flying into can be a life saver, particularly for rotary wing operations in search and rescue. Knowing when it is safe to go in and extract the hiker as soon as possible will save many people.

      Also, warning that snow conditions are likely to bring down trees and power lines two hours before it happens can help utilities mobilize repair crews and materials more quickly and maybe even know exactly where to send them to inspect the main damage from the radar analysis.

      It could also lead to nothing but some neat art projects, but we will never know until we explore it.

      • OweninGA, Thanks for pointing this out. I like learning about this sort of thing – the practical as well as theoretical.

        The model was pretty cool, I thought. One can think of them melting conceptually, but seeing it like that makes one appreciate the array of optical qualities a single flake might have it its lifetime.

  10. Now we should be able to model the compaction of snow in ice cores, simulate the reverse process and prove conclusively that global warming has changed the distribution of snowflake formation, and then demonstrate conclusively (by modeling) that faster melting snowflakes in a warmer world will lead to the extinction of the marmoset cave beetle of Lower Linguina.

  11. We were all taught in grade school that every snowflake that has ever existed in the history of the planet is different from each other with a different structure. So how would they ever be able to model any snowflake accurately when it melts because you never have the exact same temperature on one side of a snowflake as on the other side and the structure is different for each one. So what they are really modelling is the magic snowflake just like they model the magic molecule of CO2. Are they going to ban snowflakes from now on? I think that mother nature will have the last and only word on that possibility.

  12. NASA = National Aeronautical and Space Administration = snowflake research? Just fire them all.

  13. A short documentary about Snowflake Bentley – Wilson A. Bentley (1865-1931), the first man to ever photograph a snowflake.

  14. Now get them to do the reverse.
    Then do it with anything other than 100.00% pure H2O.
    Then admit how far the modellers have got to go in modelling a world dominated by water.

  15. “When he applies a high voltage to an ice crystal, says Libbrecht, a strong electric field emanates from each advancing tip. The field is strong enough to “polarize,” or align, the randomly floating water vapor molecules, and the small tip size creates large field gradients that attract them. The effect pulls water molecules onto the crystal at a higher rate than normal, and Libbrecht and Tanusheva describe a mathematical extension of the conventional theory that accounts for this effect. But above a threshold voltage the theory breaks down–exactly at the point where they observe runaway crystal growth. At that point, say the authors, the electric field effects overwhelm the tip’s surface tension, which normally keeps the crystal growth in check. Old “cloud chamber” particle detectors used a related effect: a muon speeding through a supersaturated vapor blazes a trail of ions which cause liquid droplets to form along the particle’s path. The droplets can form only because each ion provides a sufficient electric field.”

    “Electric fields can affect the growth of ice crystals in several interesting ways,”

Comments are closed.