Hmmh you may have found that I made a fairly basic error in my calculations. Seems I worked out terminal velocities for 0.5mm spherules. If so then good catch, if not well it is clearly a good idea to ask for things to be worked out openly. Anyway to answer the questions and rework the hematite spherule calculation.

Some anecdotal web research gives 90mph for 3" diameter hail.

Let's see if that makes sense.

My calculations for this us the following values; Coefficient of drag = 0.4, Density of (terrestrial air) = 1.2kg / m^3 and Density of Hail = 900kg/m^3.
Using the following formulae:
Drag Force : Fd=(1/2)*Cd*rho*A*v^2 , where Cd=Coefficient of Drag, rho=density of the medium (atmosphere), A=cross sectional area of the object and v=velocity
Force due to gravity : Fg= g*M where g = local gravity and M = Mass of the object.

At terminal velocity by definition both of the above must be equal so:

Vt=sqrt ( 2*Mass*g / (Cd*rho*A) ).

This obviously assumes simple fluid dynamics ie relatively slow velocities with no shock waves\subsonic etc. Also the coefficient of drag could vary quite a bit but it's not going to be significantly far from 0.4 for objects that are fairly smooth spheres.

Anyway these terrestrial values yield the following for hailstones:
Tennisball (6.7cm) - 40.5m/sec (91mph)
Golfball (4.27 cm) - 32.3m/sec

For comparison the same items on Mars (12g/m^3 atmospheric density and 3.822m/s^2 gravity) give
Tennisball - 260 m/sec (584mph)
Golfball - 207 m/sec

And so on to my earlier error. Hematite (well something with a density of 3g/cc) spherule. Taking Cd=0.4 because I don't have any better number and using 12g/m^3 for atmospheric density, 3.822 for g.
5mm diameter - 126m/sec (284 mph)
On Earth it would be 20m/sec (45mph)

Apologies again for the errors in the earlier calculations. I don't know whether the change actually skews things in favour of the hypothesis or not. What it certainly would account for is the absence of large concretions - a tennisball sized hematite spherule with a density of 3g/cc has a terminal velocity of about 400m/sec.

That wasn't my intention (Honest!), I think that the hailstone-style formation model in an impact surge column\collapse seems like the easiest way to get the physical attributes the way they are. I'm just trying to see if the numbers that I'm able to hack together make it seem more or less likely. I haven't figured out which way it's leaning yet myself but I've got to admit I'd like to see it be possible - hematite hail has a nice ring to it.

Dburt
If I understand correctly you are proposing that that the spherules are hematite microkrystites that condensed out of the impact plume and then distributed through an extremely thick surge deposit rather than as a boundary layer. This stretches my imagination to a degree and there does not seem to be any evidence of tektites coincident with the microkrystites, or any evidence of splash forms or other melt products. The apparent thickness of the spherule rich layer does not fit a single impact layer scenario and the fact that the hematite rich area of Mars is limited to Meridiani indicates that hematite microkrystites are not a feature of impacts on Mars, or indeed to the best of my knowledge do they have a hematite analogue in Earth impact surge deposits. Or am I missing something?

One other interesting thing about the potential terminal velocity of the spherules is that as the size approaches 16mm in diameter the terminal velocity approaches Mach 1 at ground level (~223m/sec). Mach 1 at 50km on mars is around 180m/sec which would correspond to a spherule diameter of 10mm.

I'm not really sure if this is going anywhere but it seems to me that it would be highly unlikely that anything could form smoothly in the transonic regions so that there would be a hard limit somewhere between 5 and 20mm diameter for fairly dense (>3g/cc) accretionary spheres forming in either a vertical plume or laterally travelling surge on mars that approximated the density of the current martian atmosphere. That latter assumption seems unlikely to me though.

I'm wondering how one would go about modelling the dynamics of an impact's plume to try and explore what sort of vertical velocity\pressure\temperature profile would be seen post impact. I can imagine scenarios that would allow accretions to grow to a limit and eventually rain out as they get too big to continue to rise but I've no idea if they can actually happen.

Hydrocode modelling of impact plumes is an ongoing industry in Arizona (Google Scholar the publications of HJ Melosh, E Pierazzo, BA Ivanov, NA Artemieva), but microtektites are mainly seen as the condensation products of vaporized rocks in the distal ejecta of large impacts. I have not seen in the models the scenario of "hailstones" falling out of a horizontal surge cloud as they grow to a certain size. That doesn't say it's impossible, just that it doesn't seem to figure large in the models. The horizontal surges really contribute mostly to the proximal ejecta, forming the chaotic meter-scale clastic deposits such as those that surround Chicxulub in Mexico, Belize, Cuba, Texas etc.

Kye - More data was contained in a July, 2005 Nature paper on soils by Yen et al. - they pointed out (via a graph) that Fe was positively correlated with Ni in the Berry Bowl experiment - suggesting an enrichment in Ni in the berries. That all Meridiani rocks appear to be somewhat enriched in Ni (for unknown reasons) - adding possible support for an Fe,Ni-sulfide target or a Ni-rich impactor - was published by Yen et al. in JGR in 2006.

The most recent Ni discussion, by McLennan et al., is here:
http://www.lpi.usra.edu/meetings/7thmars2007/pdf/3231.pdf
This abstract, of course, misquotes our 2005 Nature article as implying that the Fe/Ni ratio in the berries must match that of an Fe,Ni meteorite and seemingly implies that the late Roger Burns (after whom the Burn Formation was named) was a complete fool for inferring that Fe,Ni sulfide deposits should be common on Mars. It compares the Ni/Fe ratio in the spherules with that in the rocks as a whole, clearly comparing apples to oranges, inasmuch as all of the Fe is presumably oxidized (3+) and the Ni is reduced (2+). It should instead have considered the Mg2+ content of the host rocks - that is where the Ni2+ would partition, if the rocks were ever soaked in a brine (really elementary crystal chemistry.)

In this regard, its second page states "At least four distinct groundwater (brine) recharge events...can be documented or inferred..." whereas its last page states "Nevertheless, the general textural integrity of the Burns formation suggests that the amount of fluid that has interacted with these rocks after deposition of diagenetic cements has likely been very small (Fig. 3)." Now if that isn't having your cake (soaking the wind-transported salts in multiple brines) and eating it too (maintaining textural integrity), I don't know what is, as I have brought up in many previous posts. Fig. 3 shows the drastically different solubilities of gypsum vs. some other salts (except jarosite, which was left off). This graph, BTW, provides a strong argument against the extant playa model - if there were a vanished playa with the wind blowing across it, gypsum-only dunes should have been produced, as in ALL terrestrial examples, while the far more soluble salts soaked into the mud or disappeared underground (or at least, stayed too damp for the wind to move).

The assertion that Ni2+ substituted in crystalline hematite (for Fe3+) via addition of a proton (H+) to provide charge balance, seems dubious too, but is technical enough that it probably deserves a separate discussion. (About as probable as a 4-foot tall wanna-be basketball player being allowed to join a championship-bound team of 6-footers because he is standing on the shoulders of a 2-foot tall "little person".)

The bottom line is that I can conceive of no reason why low temperature concretionary hematite, soaked in acid brine in the presence of abundant Mg-phases, should be enriched in Ni, even by adsorption. The Ni2+ should stay with the Mg2+ in the host rocks (as Mg-sulfates, Mg-clays, etc.) which has the same ionic size and charge, and does so in every terrestrial example with which I am familiar.

Incidentally, some adsorption of Ni into hematitic concretions is reported by Beitler et al. (2005) "Fingerprints of fluid flow..." for the Navajo Sandstone here:
http://jsedres.sepmonline.org/cgi/content/abstract/75/4/547
but that was a system without Mg-phases (i.e., the hematite had no competition for Ni) and was possible only because of the alkaline pH (see the calculated adsorption curves in their Fig. 9B). If you are able to look up that article, check out their Fig. 3 for examples of what the spherule and color distribution at Meridiani might look like if fluid flow and brine mixing had indeed been responsible (especially Fig. 3E).

A useful quote (p. 550-551): "Small concretions commonly coalesce to form larger clumped concretions (Fig. 3E, F). Concretionary iron oxides occur in a variety of morphologies including tabular subvertical mineralization filling joints or faults, vertical pipes, subhorizontal planar strata-bound pipes, tubes, sheets, and/or irregular bodies, Liesegang-type banding, and regional zones of organized spherical concretions (Chan et al. 2000, Chan et al., 2004). These range in color from dusky brown to dusky red..." Couldn't have said it better myself.

Speaking of berries, a recent summary of the hematite content of the berries by Joliff et al. (2007) is here:
http://www.lpi.usra.edu/meetings/7thmars2007/pdf/3374.pdf
This abstract concludes, after considerably hemming and hawing about coatings, dust, and other factors, that the spherules probably are not pure hematite, but leaves a very wide range of compositions open.

Another spherule abstract by Calvin et al. (2007) is here:
http://www.lpi.usra.edu/meetings/7thmars2007/pdf/3163.pdf
Predictably, it uses the same arguments I use to argue against their being concretions (size and shape, lack of any evident pathway controls, lack of clumping, etc.) to argue that they are concretions. Go figure. The blue-gray color issue is never addressed. At least I was pleased to see that the above two abstracts referred to the berries as "spherules" rather than "concretions". Little by little...

--HDP Don

tty - Thanks for your questions. I may already have answered them in previous posts (especially the postscript to #245), but impact accretionary lapilli are exclusively spherical, as far as I am aware. "Impact spherules" narrowly defined include glass condensates which could take on odd shapes before they congeal. However, even those should not be confused with very irregular to teardrop-shaped tektites - glassy splash droplets of impact melt.

If you think 5 mm is close to the maximum size ever seen in terran spherule deposit, I disagree, unless you are leaving out spherical impact accretionary lapilli, which is what we believe the Meridiani hematite spherules to be (impact spherules, sensu strictu, are direct condensates and are smaller). Also, 5 mm is only the maximum size seen for Meridiani spherules - most are smaller to much smaller.

Hope that helps.

--HDP Don

That paper was not geology as usually understood. It was a hydrological inferrence used solely to justify, completely after the fact, the MER team's "invisible playa" or "lost oasis" or "Navajo Sandstone on Mars" hypothesis for Meridiani. The only unique feature of Meridiani geology as seen from orbit was the huge aerial extent of the specular (blue-gray) hematite signature. Finely layered, probably sulfate-rich sediments occur in many, many other areas around the highlands, as first noted by Malin and Edgett (2000, Science).

--HDP Don

Other Don - Well, got your terrestrial blinders firmly fastened still? This applies to impacts too. This question was first asked by Shaka in his post #56 and answered by me in post #60, as well as in later posts. Two factors you possibly haven't considered: 1) Earth has had solid bedrock for impact targets, owing to plate tectonic processes, shallow marine sedimentology, and so on. On Mars at the end of the Late Heavy Bombardment and prior to most Tharsis volcanism, solid bedrock may have been a commodity in rather short supply. As I said in prior posts (as per Wm. K. Hartmann's "kablooey of dust and steam" quote), beat on Meridiani, and you'll just get more Meridiani. 2) Terrestrial impact studies are highly biased by the immediate removal and alteration of distant fines, such as those we see at Meridiani. They simply aren't available for study on Earth. That doesn't mean they were never there. Volcanic surges tend to be preserved by overlying lava flows or ignimbrites. No such luck for impact surges - you mainly preserve the coarse breccia (suevite) near the crater, or spherules in marine sediments. And I'm getting tired of repeating this - we never claimed that the Meridiani spherules were "melt droplets or tektites" - they appear to be impact accretionary lapilli of unusual composition.

--HDP Don

P.S. (added a day later). Last night I forgot to mention two other reasons, other than possible rarity of bedrock targets and likelihood of erosion/alteration, why impact processes on Mars are probably different from those on Earth (so please remove your terrestrial blinders). The first is that the Martian subsurface is frozen, probably to a depth of several kilometers (the so-called cryosphere), and probably has been since at least the end of the Late Heavy Bombardment (because erosional remnants of very old rampart craters are found, and these are widely accepted as evidence of the martian cryosphere). A cold, brittle, broken impact target, with up to several tens of percent ice (permafrost) cementing it, is probably going to break far more violently as its ice flashes into steam, than any terrestrial bedrock target. Therefore, many more fine particles will be produced. In this regard, so-called rampart craters appear unique to Mars, and are themselves probably preserved only as erosional remants (coarser and/or better cemented, near-crater materials).

The second, as mentioned in previous posts, is that the surface of Mars (unlike that of Earth) has probably always been covered nearly everywhere with a thin to thick veneer of drifting sand and dust (most of it probably impact-derived), because there usually is no liquid water to cement it into a rock (although ice cements it at the poles). An impact surge cloud will scour and transport all that sand and dust and incorporate it into the steam- and salt-cemented surge deposit - which therefore is going to look extremely sandy, compared to most terrestrial surge deposits (volcanic or impact). In an earlier post I already used this feature as an argument for why the sandy surge deposits at Home Plate are probably impact- rather than local volcano-related, despite their single ballistic sag. It may also help explain the sandy nature of the Meridiani deposits (along with simple impact reworking of earlier sandy layered deposits, mentioned in my original post).

Hope that helps remove the blinders. BTW, without mentioning names, some people who study terrestrial and lunar impacts have been EXTREMELY opposed to the idea that the Meridiani deposits could be impact related. Your reaction (and the original one of Shaka) have been comparatively mild. --HDP Don

Apologies, HDP Burt...missed that point in the argument, and I can only concur with that statement.

Your hailstone "eureka" does bring up an interesting aspect of this debate, though. Has anyone seen any evidence for accretion by whatever means in blueberries? Hailstones show layers, but the blueberries that Oppy has bisected within a matrix look pretty uniform. Of course, this may just reflect the behavior of the RAT (plus the fact that we can't blow off the residual dust), but this may also indicate that any accretional layering is very fine in scale--if it's there at all. Finer layers imply more gradual, perhaps even cyclical. formation processes.

Shaka - There are no Meridiani evaporites. There never were, except perhaps in press releases. Each of the first 3 landers (Viking 1 and 2, Pathfinder) detected about 10% sulfates in soil. Duricrust (a moisture-related process resulting from capillarity) was suggested. Oppy detected about 30% sulfates as an improbable mix of highly soluble and nearly insoluble salts (almost no chlorides, which might indicate a true evaporite). To account for this improbable mix the MER team had to back off their initial playa/sabkha claims and hypothesize a totally invisible playa whose wind erosion led to these mixed salts (despite the fact that all terrestrial analogs consist purely of the insoluble salt, gypsum).

I agree with you that impacts are awesome. I don't think that you or most other people here are considering what happens if you impact soft sand, probably with interstitial ice, rather than flinty bedrock. In his 2003 Mars book William K. Hartmann predicted that most energy would be absorbed, and that the only result would be "a kablooey of dust and steam" or what we are calling an impact surge deposit. If you don't think it's "violent" to deposit perhaps several meters of sediment in a few minutes, perhaps you need to expand your definition. Shoot bullets into a sandbank and see what happens. Distant impacts into bedrock are permitted too, of course.

--HDP Don

The ones that have been RATed show no layering, but some of the broken ones do (including at least one imaged in the past few weeks). Our 2005 Nature paper illustrates a broken berry with layering. Of course, layering could be present in either concretions or accretionary lapilli, so we have to look to other indicators for a diagnosis.

--HDP Don

No it doesn't, because then the berries, rather than being randomly distributed in the rock, should be concentrated just below the alleged water table (as they are in the Navajo and Page Sandstones - the putative analogs for Meridiani sedimentology). Diffusion through a stagnant, pore-fluid brine is extremely slow.

--HDP Don

Don, do you have a link to that image, please? Very interested...

ngunn - The "official" model doesn't have billions of years to work its magic - they have to do it while that part of Mars was somehow warm and wet long after all available other evidence indicates that the rest of Mars was cold and dry. Also, you can have as many impact surges as you want, spread out over as may billions of years as you want, so long as Meridiani was the target. Victoria itself probably generated a surge, which Oppy rolled right over without noticing anything amiss (I hope to address this in more detail in a future post).

If you understand, through the medium of the putative brine, likely Ni2+ partioning between good-fit Mg-phases and the misfit Fe3+ phase (hematite), the Ni argument is quite straightforward - there simply shouldn't be any in aqueously-crystallized hematite. You have to know something about about crystal chemistry and partition coefficients though.

--HDP Don

Helvick - Thanks for the clarification. Keep in mind that the impact makes its own atmosphere - it vaporizes everything, even silicate rocks (briefly). Most of the turbulent cloud from which we hypothesize that the blue-gray hematite might have crystallized and been accreted would then have consisted of steam, presumably derived from vaporization of near-surface ice or very deep brine (plus minor contributions from hydrated and hydrous minerals). So the terminal velocity through the much thinner Martian atmosphere would be largely irrelevant. The turbulent could would contain lots of suspended solids too, contributing to its ability to support spherules. We're not particularly happy with that "hematite hailstones" model either (post-depositional oxidation and/or leaching of some other type of accretionary lapilli would be easier conceptually), but it's the only way we can think of to explain the blue-gray color of the spherules (something the concretion model absolutely fails to do - and even forgets to mention).

--HDP Don