Reading the story of Pluto’s surface

Active Convection in Sputnik Planum

Sputnik Planum cells blur crop dim  The pattern of irregular polygons is typical of Rayleigh–Bénard convection (see below). New Horizons’ images contain evidence that the convection is continuing and imply that Pluto has recently gone through some cataclysmic events.


irregular convection cells

Bénard convection cells tend to have quite irregular shapes when the situation is changing rapidly or the vessel is irregular as in this frame    from the Wikipedia video which can be viewed below. Even though it is on a vastly smaller scale this pattern of Bénard convection cells is surprisingly similar to the irregular polygons of Sputnik Planum.



(click picture to enlarge text)Pluto Sputnik speculations

Timescales: Evaporation vs Convection

Sputnik Plunum evaporation vs convection cropPerhaps the clearest sign that convection is ongoing is the contrast between areas of Sputnik Planum that are pitted due to evaporation and neighboring convection cells that are smooth. Evaporation must also be happening from the surface of the convection cells but the developing pits are transported to the cell edge and disappear down into the interior before the pits are bi enough to be visible. So the age of the surface within the convection cells must be a small fraction of that of the pitted zones. Compression artifacts in this preliminary New Horizons image hide any signs of initial evaporative pitting within convection cells but when uncompressed high resolution images are eventually down-linked, if smearing effects of sub-surface sheer motion are not to strong, it may be possible to measure the rate of convective transport across the surfaces of the convection cells by comparing the degree of evaporative pitting within them to that of the non-convecting pitted areas.


‘Black Blobs’

One consequence of the convection hypothesis is that the ice of Sputnik Planum must be soft.  So rather than sitting on a hard surface the ‘black blobs’ must be floating. They must be less dense than the ice so as not to sink. The ice is likely a mixture of methane, nitrogen and carbon monoxide, with methane and nitrogen as the candidates to be the primary constituent. Both of these are less dense than water ice. So the black blobs cannot be icebergs composed of dirty water ice.

The fact that only a few isolated black blobs lie within the polygons is consistent with them being carried to the surface within the convective flows then fairly rapidly conveyed to the peripheral troughs formed by sinking ice.

A few black blobs have been identified as possible sites of geysers due to apparent downwind dust streaks. These also all lie within the polygonal boundaries of convection cells implying that they have only recently reached the surface. Perhaps these ‘smoking blobs’ are the result of a dusty outbursts that occurs when a vertically extended deposit of dust and some gas even more volatile than methane, nitrogen or carbon monoxide reaches the surface.

The black blobs are an important clue to understanding Pluto overall, in that they demonstrate the existence of a solid substance on Pluto that has a low enough density to float on the ice of Sputnik Planum.

In the pitted areas of ice the lines of black blobs mark the boundaries of previously active but now dead convection cells. Their more dispersed distribution compared to lines of black blobs in the troughs surrounding active convection cells is further evidence that they are floating and free to move with the ice when its surface relaxes. Likewise, the concentrated distribution of black blobs in troughs, such as the U-shaped hills named Astrid Colles, is further confirmation that convective motion is still happening.

More Rayleigh–Bénard convection videos (page link)

Rayleigh–Bénard convection on Wikipedia (outside link)

The big questions are:

Where does all the heat come from to drive convection on such a small old frigid world?

What can this tell us about Pluto’s internal structure?

– CO may be the key.

– A new source of heat may not be necessary if heat that has long been bottled up inside Pluto was suddenly released. This could have happened due to a recent cataclysmic asteroid impact or more correctly a Trans-Neptunian Object impact.

Competing models of Pluto’s crust – a thin hyrocarbon crust or a thick water ice crust

(1) If Pluto has a thin crust an impacting asteroid could punch a hole through it and shock waves from the impact could crack the crust over a Tombaugh Regio sized area. Chunks of shattered crust might sink exposing a mantle of warm soft ice underneath. In this case Pluto’s crust would be something like an Earthly ice shelf extending planet wide, except that the ocean beneath is more like whipped cream, though not as dense. The crust itself may be more like pumice (the volcanic rock that floats), although given that hydrocarbons are raining out of  Pluto’s sky it is probably some sort of low density tar or pitch (behaving like hard toffee).

1 pluto pre-impact annotatedPrior to the impact a thin (~1 km) crust overlies CH4/N2/CO ice. Pluto’s meager supply of internal heat is trapped beneath the crust leading to higher temperatures there.

Cracks in the crust might release some of the pressure on the underlying ice, allowing vaporization and a venting of gases in the form of geysers, like those on Triton, although Pluto’s atmosphere would have been far more tenuous than now.

The sheer vastness of space beyond Neptune means that impacts on trans-Neptunian objects (TNOs) like Pluto must be quite rare compared to the rate in the inner solar system. Pluto’s 2:3 orbital resonance with Neptune keeps it well clear of the giant planet but Pluto does pass fairly close to Neptune’s orbit when Neptune is not there. “Scattered disk objects” make up a sizable fraction of the trans-Neptunian population. They were slung into their highly elliptical orbits due to close encounters with Neptune so their elliptical orbits all regularly carry them back to the vicinity of Neptune’s orbit. The Sputnik Planum impactor is likely to have been one of these scattered disk objects at perihelion.

2 pluto during impact annotatedThe impact could have punched through the thin hydrocarbon crust, vaporizing a section of it. Pluto’s weak gravity would have allowed some ejecta to be thrown directly into space, more would land back on the surface, but any chunks of ice would rapidly evaporate giving Pluto a new atmosphere.

Shock waves from the blast crack the crust in a wide area around the crater. Even though the crater could have been small compared to the size of the Sputnik Planum, the last could have triggered a chain of events taking in Tombaugh Regio and effecting the entire planet.

 3 pluto initially exposed ice cools and sinks annotatedThe softness of Pluto’s CH4+N2+CO ice means that the crater flattens out rapidly. Ice in the crater floor, previously insulated and kept warm by the hydrocarbon crust, is now exposed to the chill of space.

Evaporation cools the surface while supplying gases to Pluto’s atmosphere. The sheet of ice at the surface becomes denser as it cools. Eventually, some part of the dense surface ice sheet starts to sink and draws more of the dense sheet after it. Somewhere the surface sheet tears exposing new warm ice underneath, thus triggering the pattern of solid state convection we see on Sputnik Planum. The warmest ice is exposed in the centers of convection cells. Convection cell centers appear darker in the New Horizons high resolution image. This might be newly exposed ice or we might just be seeing through a thinner layer of cool ice. In either case it is clear that the properties of Pluto’s sub-crustal ice change significantly when it reaches the surface.

New Horizons’ observations may be able to resolve a temperature difference between the centers and fringes of convection cells. This would be incontrovertible evidence of active convection. Being warmest the centers of the convection cells should also be the richest source of new gases for Pluto’s atmosphere. 

The relatively small size of the crater means that initially there is likely to have been only one or at most a few convection cells. The size scale of the convection cells depends on the rheology of the ice, properties such as stiffness, the dependence of density on temperature, thermal conductivity and compressibility (or more importantly incompressibility and lack of thermal conductivity).

4 pluto convection zone widens as cracked crust sucked in annotatedThe size of the convective region grows as chunks of fractured crust around the initial crater are jostled by the convective motions, like icebergs caught in currents.

The hydrocarbon crust is likely to be only marginally buoyant so pieces may tip over and pile up or even be sucked down.



(2) The alternate,  more conventional, view of Pluto’s surface it that it has a thick crust of water ice. Water ice Lunar Seas impact crater in a thick crust flooded with lava from much deeper via cracks would be as hard as rock at Plutonian temperatures. In this case an asteroid impact could excavate a deep crater that was then flooded with warm ices flowing up through cracks from an even deeper warm mantle. This would look similar to the creation of Lunar Maria (Seas of the Moon).

Glaciers, Ice Floods and Tsunamis

There is a lot more going on in Tombaugh Regio than a sudden release of internal heat through convection. At the northern shore of Sputnik Planum ice appears to be flowing onto the landscape in a vast glacial flow, while on the eastern shore of Sputnik Planum, towards the right ventricle of the ‘heart’, and the southern shore, towards the Norgay Montes and Pandemonium Dorsa, ice appears to be receeding from the landscape.

N2 ice slopping around crater in water ice crustIn a conventional thick water ice crust (2) view of Pluto, schematically these ice flows would be something like this . . . .   →  .   But in this thick crust model there seems to be no reason why the ice would flow to and fro like water sloping around in a bathtub. There is no force to cause such slopping about. In the thin crust model (1) the crust itself can rise and sink, triggering the ice flows.


Norgay Montes – jagged mountains on the edge of Sputnik Planum are likely chunks of broken crust washed up on the shore of the ice sea.

The sharp peaks of the 3500 m high Norgay Jagged peaks of Norgay Montes look like beached icebergsMontes cast longer shadows as they approach the terminator. The angles of Norgay Montes are in sharp contrast to the rounded forms of the mountains beyond. Note the line of a crack running through the rounded mountains. It appears to have been filled in or annealed, perhaps due to the accumulation of sticky hydrocarbons (tholins) raining down over eons. The rounded mountains may too have once been jagged, but have been buried under a blanket of accumulated hydrocarbons.

The Norgay Montes are clearly new compared to the undulating mountains beyond. With their random placement and jagged edges the Norgay Montes look a lot like icebergs thrown up onto a shore by a tsunami and this may be exactly what they are (although not made of water ice).  If Pluto has a thin crust that is floating on a mantle ices, as in scenario (1) above, then the Norgay Montes are likely chunks of the crust that was shattered in the impact. They were carried up onto still intact crust when the soft ices of Sputnik Planum welled up and inundated the still intact crust of the rounded mountains (east Cthulhu Regio?) and left there when the soft ice receded.

A sign of this icy inundation is the ice still dammed up between peaks on the left side of the image.  It is hard to see how this pool of icy could have been filled by snow storms since it is so localized. There is no sign of the pools of ice around the other Norgay Montes peaks or in the folds of the rounded mountains beyond that would be expected if this ice was deposited as snow. It appears that the dammed up ice might be flowing, glacially, out of the dam and down onto the plain, although the hint is at the limits of the resolution of this image.

The ice tsunami

In the north of Sputnik Planum ice appears to be flowing onto ancient terrain, along with its irregular polygon shapes.

ice tsunami over northern Sputnik Plnum

n.b. It’s not yet clear whether the ice of Sputnik Planum is predominantly frozen methane or nitrogen.


Processed closeup image of ice flows in Sputnik Planum(Ron Baalke)

At the northern shore the ice is flowing around obstacles. Here the shapes of convection cells have become because the obstacles are smaller than the 30 km size scale of the convection cells.

It would be interesting to have observations of this flow over time to know just how fast this ice is engulfing the land here.

The thin crust model of Pluto’s surface (1) offers an explanation of these flows without the need for giant ice waves as in the thick crust model (2). If the crust is somewhat flexible it might be pulled down allowing Sputnik Planum’s soft ice to flow over it, while the ice itself maintains an almost perfectly flat surface (or at least one that follows an equipotential surface).

5 pluto downflow draws crust down annotatedIce cooled at the surface will always sink where it butts up against intact crust, so there is a natural tendency for the icy solid state convection to pull the edge of the crust down. If some ice does get on top of the crust the extra cooling area will increase flow of colder descending ice over the edge of the crust, tending to drag it down even more strongly. Of course the 3-dimensional pattern of warmer ice flowing onto the crust to replace the cold ice falling off it, must be much more complicated than can be represented in this schematic cross-section.

Any chunks of fractured crust floating on the ice will be carried with it when it flows over a sunken section of crust. The chunks will go aground and when the sagging crust eventually springs back to the surface they will be mountains. The ranges of mountains composed of big angular chunks along the western side of Sptunik Planum, including Norgay Montes, Hillary Montes, Zheng He Montes and al-ldrisi Montes, were all likely formed this way. The fact that these chunky mountain ranges are all along the western side of Sputnik Planum implies that this was the first place where the crust was drawn down after the impact.

jumbled blocks in the upper left(Glastoner)The slab-like nature of these angular mountains is particularly obvious for the al-ldrisi Montes

cellular terrain outlined in yellow


The extent of the al-ldrisi Montes is outlined in yellow (thanks to Glastoner)




Crust Sunken Under Ice

There is a line of demarcation that arcs across Sputnik Planum. This may well mark the submerged edge of the Sputnik Planum dirt division north upintact but sunken crust. Inside the arc the ice is whiter and cleaner. Here the convection cells likely go deep. Outside the arc, to the north and west, the ice is darker. Here warmer ice might move across the top surface of the crust, picking up dark crust solids, before carrying it to the surface. The surface of the ice is even darker at the northernmost limit of Sputnik Planum. Here the ice is flooding over the crust so the ice is at its shallowest and it is passing over crust that has not been inundated before and so may have had a lot of loose regolith for the convective motions to dredge up.



Pluto’s Cycles of Glacial Flood and Retreat

Pluto enhanced colour- right ventricle extent

The darker tone of the northern third of Sputnik Planum, where the glacial front is flowing northward is clear in this false (greatly exaggerated) color view.

While the landscape is presently being flooded with a tsunami of ice in the north of Sputnik Planum, in the south ice appears to be draining off the highlands south of the Norgay Montes and leaving behind a landscape painted in a patina of ice that shows up as a light greenish-blue in enhanced color images.

The right ventricle of the ‘heart’, to the east of Sputnik Planum, shares this greenish-blue hue of remnant ice. Although, there is a subtle difference in hue between the two lobes, with the greenish-blue of the right ventricle being slightly more greenish than the color of the southern lobe. It is hard to say whether this difference is significant due to the different lighting angle of the 2 regions, but it does suggest a subtle difference in composition, perhaps due to differing ages of the icy veneer.

The obvious interpretation of these large regions, painted with ice, extending away from Sputnik Planum is that they are the result of previous episodes of icy flooding and that they are part of a sequence of major icy inundations. ice floods. In the highest resolution image of the edge of the right ventricle, so far released, ice is apparently flowing down from terraces and high ice lakes into the Sputnik Planum but the quantities of ice flowing are tiny compared to the broad expanse of ice apparently flowing north between the Norgay Montes and Krun Macula from the southern lobe. So it seems likely that southern lobe flood occurred more recently than right ventricle (eastern lobe) flood.

enhanced color global view of Pluto enhcrpThe al-Idrisi Montes, Bare Montes, Zheng He Montes, Hillary Montes and Norgay Montes all lay along the western side of Sputnik Planum. An even earlier flood, the first flood after the crust shattering impact that created the Sputnik Planum, must have occurred in this direction in order to deposit the fragments of crust that make up these mountain ranges.

The greenish-blue icy veneer left behind by the right ventricle flood shows that it extended perhaps 90º around Pluto’s equator. It’s noteable that compared to this none of the fragments of crust that now make up mountain ranges lies very far from the location under Sputnik Planum where they started. Yet that initial westward flood must have been big enough in scale to sweep all the sizable fragments of crust to the west since almost none seem to be still floating around or lodged on the shores of the later floods. So it seems likely that the initial westerly flood (#1) extended far past the present western shore of Sputnik Planum but that it was not deep enough to carry the detached chunks of crust with it. The C-shaped lake of light colored ice in Elliot crater and the greyish color of the land to its north is evidence that the ice flood got at least that far.

From Bare Montes to Norgay Montes in the south, the crust chunk mountain ranges all lie up against the heavily cratered ancient terrain of Cthulhu Regio. The depth of the craters there and the height of their rims shows that the crust is thick. The strength of the crust under Cthulhu Regio must have prevented it from buckling downwards under the weight of that initial ice tsunami (#1) and saved it from flooding too.

Pluto enhanced colour Sleipnir FossaGiven the 3.5 km height of peaks within the Norgay Montes and that they are embedded at odd angles, the depth of the flood that delivered these chunks of crust to this location was probably ~3 km. Judging by the greenish-blue residue, when the right ventricle flood got as far the Sleipnir Fossa (long, narrow depression) it was not deep enough cover the highlands on either side of the Fossa, but flowed around them. This pattern of icy residue in valleys but not on highlands is further confirmation that the veneer of ice marking out the right ventricle results from flooding rather than weather.

The Nature of Pluto’s Crust

The sequential character of Pluto’s ice floods constrains the our theoretical understanding. They cannot, for example result from a huge up-welling of ice out of Sputnik Planum. If the Sputnik Planum ice had ‘broken its banks’ and flooded the surrounding surface all of the floods would have happened simultaneously. The idea that the floods could be due to waves of ice sloshing around in Sputnik Planum and breaking out in one direction and then another is implausible due the vast extent of the floods, clearly much bigger than Sputnik Planum itself in the case of the ‘right ventricle’ flood.

The only way tenable way I can think of to generate the observed series of ice floods on Pluto is if the crust itself sinks to allow ice to flow over it in a series of locations. This means that the crust must be thin and it must float on an ocean of the methane/nitrogen/carbon monoxide ice we see in Sputnik Planum. The crust must also be plastic, not too brittle, so that it can deform and return to its initial position without breaking up. Long chain organic compounds, like tar or plastic, would fit the bill if they have low enough density to float. This tarry crust could have been built up from tholins formed in Pluto’s atmosphere and raining down on the ice over long periods of time.

Cold ice in the south may be flowing back
Pitting in this large southern extension of Sputnik Planum shows that convection has stopped. It connects to the region extending into the southern hemisphere covered in the icy patina that is likely the remnant of an ice flood.

Sputnik Planum south pitted texture detailThe flooding of the crust with ice would tend to be a runaway process, at least at first, since as warmish ice floods over the crust it increases Sputnik Planum’s surface area for cooling. More cooling means that the ice on top of the crust would become denser than that ice below the crust, tending to weigh it down and allow even more ice to flood over the crust. Perhaps it is this positive feedback that allowed the easterly flood to extend so far around the equator.

The flooding might eventually be halted when it reached a particularly thick section of crust, strong enough the resist the weight of flood ice. Or the flood might just have reached the point where it was so extended that the convection driven lateral flow supplying warm ice could not keep up with the rate of cooling. In this case, vertical convection would tend to stop where the ice flood was shallowest, particular at leading edge of the flood, but evaporation from the surface would continue. If enough ice eventually evaporated from near the front of the flood, it could lighten the load on the crust there enough for it to begin to rebound upwards. That could change the momentum of the flood so that it started to flow back towards Sputnik Planum. The final stages of such a reverse flow may be what we are now witnessing in the southernmost extension of Sputnik Planum, between the Norgay Montes and Krun Macula. Any shallow pools of CH4+N2+CO ice not flowing back into Sputnik Planum and left behind on the rising crust would rapidly evaporate, leaving only the patina of the still mysterious non-volatile impurity (H2O ?, CO2 ?) that we recognize by its greenish blue color in the false (highly exaggerated) color images from Ralph instrument on New Horizons. A significant fraction of flood ice might be dissipated through evaporation.

4 ice floods occuring sequentially around Sptnik Planum
A possible time sequence of the series of icy floods occurring around Sputnik Planum. Sagging of the thin tarry crust allows ice to flow over it. Eventual evaporation of most components of the flood ice allows the crust to return to its original height, leaving behind a bluish residue of non-volatile ice.

Crustal Healing – After the Floods

It is noteworthy that the bluish hue due to the residue of non-volatile ice left by floods to south and in the ‘right ventricle’ to the east of Sputnik planum is not visible on the site of the initial flood that created the ranges of mountains along the western edge of Sputnik Planum. This might mean that the ‘non-volatile residue’ is actually slightly volatile and does eventually evaporate or that enough time has passed for the bluish veneer of non-volatile ice to be covered by a layer of hydrocarbons raining down from the sky thick enough to mask its bluish hue.

The second possibility is more interesting because it points to a mechanism whereby the Sputnik Planum phenomenon can eventually come to an end and the surface of Pluto can be healed. If the convection in Sputnik Planum eventually stops, as it has done in the southern part of the plain, the continued evaporation of volatile ices without stirring would allow the accumulation of the bluish non-volatile residue into an insulating layer that would seal the volatile ices from the near vacuum of Pluto’s atmosphere, halting evaporation. Over time hydrocarbons raining down from Pluto’s haze layers would build up new crust over the plain.

Events such as we now see around Tombaugh Regio must have happened many times in Pluto’s past, when the conditions of a sufficient build up of internal heat and a sufficiently large impact on a sufficiently thin section of crust were met. So on Pluto today we should see the signs of other episodes where areas of shattered crust have healed over at various times in the past.


continue with “Pluto: The Big Picture”



5 thoughts on “Reading the story of Pluto’s surface

  1. Fantastic page, congratulations!
    The similiarity of Sputnik Planum “polygonal terrain” to Benard convection cells was also one of my first thoughts when I saw the first pictures from the NH mission. Well done on taking this idea so far. The pitting of the surface through evaporation – and deducing age out of them – is a very intriguing idea, especially that cometary nuclei have circular pits created in exactly the same way. So the freezing of convection could in fact do something like that.
    One thing that bothers me in this interpretation are the apparent ridged bottoms of the troughs separating the Sputnik Planum cells. This does not look like something related to convection, they are far too regular and linear to be floating remnants in the “lanes” between convective cells (a la black blobs that you interpret as light floaters). A large part of Sputnik Planum has “double troughs” between the cells. I personally interpret them as deposits at the bottoms of extensional cracks, where subsurface ices get released. Something like double ridges on Europa.
    I’m writing a bit about Pluto’s geology on my WordPress blog as well, I will try to translate as much as possible into English soon. If you’re interested, check it out: .


    1. Thank you very much for your praise for my blog.

      I too was troubled by the ridges that run along the troughs surrounding Sputnik Planum’s ‘irregular polygons’, giving the double trough appearance. They are not a feature of any laboratory demonstration of Rayleigh-Benard convection that I have seen. Yet, the evidence of the evidence for convection from the morphology and rapid renewal of the surface within the polygons is just too compelling.

      Ridges within troughs around convection cells The scale of this convection is far greater than anything done in the laboratory so it is not surprising that it has unfamiliar aspects. At first I wondered if convection on the Sun might provide a clue. On the Sun jets of gas called spicules shot up hundreds of kilometres above the surface (the photosphere) around the boundaries for convection cells. I thought that perhaps on Pluto the relative motion of two neighbouring descending streams of cold ice might create a vertical plane of dislocation, of fracture, between them that provides a shortcut for hot material (ice/liquid) at depth to rapidly reach the surface. In this case the ridges would appear hotter than the surrounding ice in an infrared photo.

      However, I came to the conclusion that the reason for the ridges is probably much simpler. The ridges likely result from turbulence where streams of cold ice from neighbouring convection cells meet as they start to fall into the interior. It is what happens when too much liquid tries to fall down a slot which is too narrow. I have seen video of something like this. It was in a factory setting where two fairly laminar streams of viscous liquid met from opposite sides and poured into a slot. A ridge of liquid that couldn’t immediately enter the slot formed above the point where the streams met. Unfortunately I can’t remember exactly what the material in the video was. It might have been liquid rubber, molten chocolate, molten glass or something else. I will try to find an example. The closest example I have been able to find that kind of illustrates the principle is this video of liquid rubber being squeezed between rollers. A ridge forms where liquid rubber is squeezed between rollers.Squeezing rubber between rollers on YouTube

      The ridges within the troughs around convection cells are most prominent in the center of Sputnik Planum, where the cells are the biggest and the convection should be most vigorous. This supports the idea that the ridges are a by-product of ice flow.

      On the topic of comparing the extent of pitting due to ice evaporation in the center of a convection cell with its periphery in order to gauge the rate of convection, since I wrote that I’ve noticed that there are actually convection cells near the shore of Sputnik Planum which have visible pitting around their edge but not in their centers. Convection within these cells must be slow enough that the time for ice to move from the cell’s center to its edge is a little more than it takes for evaporative pits to become visible in New Horizons’ images. So that idea works.

      Your blog looks quite impressive. I’m looking forward to reading your posts in detail as soon as I catch up on writing a couple of new Pluto posts.


      1. Sure, I know what you mean, I have seen videos of that kind. Trouble is that the central ridges in between the downgoing cells were usually pretty narrow in absolute terms and the force responsible for their smooth and uniform appearance was most likely surface tension. I can hardly imagine this working at the kilometer scale with solid/semisolid ice as the material, and the result being long, narrow, nearly linear and uniform rolls. Your rubber video kinda proves the point.

        Now that I’m thinking about it, my own initial model – the ridges being depositional features from the release of subsurface material – is rather weak as well. Such features are always localized, with discrete, separated deposition centers. So my bet is that a third option might be preferable.


      2. I find the whole phenomenon of Rayleigh-Benard convection totally amazing. The fact that virtually the same patterns occur on scales from a cup of tea to the atmospheres of stars is mind boggling. I guess that it is an example of order born out of chaos in self organising systems.

        The ice of Sputnik Planum is much softer on the kilometer scale of the width of the ridges within troughs than the rubber between the rollers is on its centimeter scale. As for the uniformity, the key words are ‘self organising’. Uniformity is in the nature of Rayleigh-Benard convection.


  2. Hey man, I see you’re not posting as frequently anymore, but I wanted you to know that I remembered this post from a year ago and sought it out again today. you’re making some really high quality science writing here! As a very impatient science enthusiast, I always suspect that the best scientists know the explanations almost immediately, but have to wait for the peer-review process to do its thing before everybody else gets those answers. For example, only recently have reports published that the polygonal structure was caused by convection currents. Here you are over a year ago saying the same thing! Even better, you’ve written about it in an interesting and approachable way, with plenty of documentation. I appreciate that.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s