LPSC: Wrapping up Tuesday: The Moon, Mars, Mercury, Vesta, and back to Mars
Posted by Emily Lakdawalla
12-03-2010 13:27 CST
Well, it's already mid-day on the Friday a week after the Lunar and Planetary Science Conference ended and I'm STILL not done writing up my notes. In the interest of moving on to other things, I am now posting pretty much everything I have left, with minimal editing to make it readable and few of my usual comments; any comments I have will be in italics and/or brackets. I apologize for these notes being less interpreted than previous ones and therefore occasionally (or even frequently) cryptic, but it was either do it this way or not at all!
I started Tuesday in the LCROSS session. Abstracts can be found here.
Tony Colaprete: About the choice of target: We luckily had the opportunity to be indecisive until very end with respect to target choice. We could not be confident in hydrogen association with Cabeus A. So we had to go to Cabeus where we were certain about hydrogen. Ejecta was illuminated by the time it got to 830 meters above the surface. But a large wall along Cabeus' rim really blocked view from Earth. "Space navigation just blows my mind." LCROSS Centaur hit within 88 ± 66 meters of its intended target.
The sustained ejecta cloud was the biggest surprise of the mission. [Funny, a sustained ejecta cloud was also the biggest surprise of the Deep Impact mission!] Late rise time of near infrared flash suggests impact into soft, fluffy, low-strength target. In flash radiometer we saw no flash at all, which suggests visible flash was completely quelched. High ejecta cloud may have attenuated it, or maybe energy went into vaporization of volatiles. At impact plus 1 sec we had 6 km across low angle [that is, near horizontal] blast wave, which lasted about 15 sec. This may relate to hot vapor cloud.
The types and flavors of things we see suggest there are probably multiple [water] sources. Surprising.His talk ended late, so the session chair said there was no time for questions. But Sasha Basilevsky had run to the mike and demanded to be heard, saying "It is CRUCIAL qvestion!" He asked -- demanded -- to know how they could be so confident that the hydrogen-rich materials they saw in the ejecta were not from propellant left in the Centaur. Colaprete said that he couldn't directly answer the question because the DoD was involved, but what he could say was that "This Centaur was not your typical Centaur."
Jennifer Heldmann on camera results followed. Long duration ejecta cloud lasting 200-plus seconds requires a high-angle plume; LCROSS had to have combination of both low and high-angle ejecta plumes. Images show that LCROSS plume was not a donut. It is not an annular ring; it is filled in in the middle. Corresponds with what we see in visible camera. This too suggests there were two ejecta plumes, one low angle (the donut) and one high angle (filling in the donut hole).
Hydroxyl was observed with UV/VIS spectroscopy. The hydroxyl was produced from ice particles that have sublimated and then photodissociated, in first 50 seconds. Ice particles were lofted into sunlight, then sublimated and photodissociated. Color changes with time.
mall particles get into sun first, large particles take longer. Small particles leave FOV first as plume expands. Larger particles contribute to plume at later times.
Post-impact spectral extraction: first divide by photometry to assess changes in spectral shape; divide by a pre-impact spectrum (which is dominated by scattered sunlight). Sum time-adjacent spectra to obtain high signal-to-noise ratio, then oversample in time to reveal evolution of features. Next fit a continuum scaled to solar spectrum then a gray body to represent thermal contribution. After all this, the spectra reveal SO2; possible features of C2H2; surprise to me [Jennifer, that is] because I do comets and I wasn't familiar with that species, but I am very familiar with methanol.
Hong summarized Earth-based observations using the Subaru telescope. They basically detected nothing, putting an upper limit of 30 kilograms on the mass of water in the ejecta. Why the discrepancy with LCROSS results? It's possible there was a small amount of high-speed ejecta; a dry surface layer; or low-angle ejecta [which would not have got into sunlight so was invisible to Subaru].
Paul Hayne gave Diviner results. We observed emission from LCROSS impact at about 2 hour intervals both before and after impact, including 90 seconds after impact. We saw thermal emission from the impact at 90 seconds in all 7 thermal channels. We can constrain temperature to about 600 K. Our solar channel also recorded scattered sunlight from ejecta cloud, allows us to constrain mass of material to be between 800 and 1800 kg. We can derive a cooling curve which puts initial temperature for hottest part of impact site at 1100 Kelvin. At 90 seconds after impact we also observed a cooler ejecta blanket around impact site, warmer than background material. We can't constrain volatiles but cooling curves from Diviner are consistent with presence of sublimating volatiles. We can't confirm or disprove their presence.
We weren't really sure what we'd get. We wanted to estimate impact temperature to constrain thermophsyical properties near surface. We thought we might see the ejecta cloud. We did not see anything at the limb viewing geometry. One issue we have to deal with constantly is presence of more than one temperature in a scene. In any given scene, even persistently shadowed areas, you get scattered light that illuminates rough boulders on surface, so you have to model sub-pixel temperature variations. You have to take into account different Planck curves within the scene. All these effects are well understood and we were able to accurately model sub-pixel contributions.
We derived cooling curves over the 4 hours that we observed thermal emission. You can see that we can only accurately reproduce observations using a regolith (not rock) model. Consistent with very porous, unconsolidated regolith.
At impact+90s: a very dynamic situation with hot 600K inner region surrounded by a 115K ejecta blanket and a warm sunlit ejecta plume of 800-1800 kg. Because thermal wave was propagating downward with time, additional volatiles should have been sublimating during 4 hours of our observation, from 5 to 50 kg of water.
Also we showed we can accurately target any spot on moon to sub-km accuracy. We also found that the Diviner channels can all see in the dark, seeing in permanently shadowed areas to lift the veil on these permanently shadowed regions.
Pete Schultz talked about some modeling work to give context to the LCROSS impact. This was an impact that had to send ejecta into the sunlight. We had to use the Sun to see species. Pre-encounter predictions: used NASA Ames Vertical Gun range to answer questions; can we get ejecta into sun? Can we get enough ejecta into sun? Can we see the cratering process, and watch how it evolves? Experiments show earliest ejecta had higher speed and lower ejection angles than nominal model, so not much ejecta may get into sunlight. Estimates ranged from 20,000 kg to below 2000 kg. We were very concerned about impact. [That is, they were concerned whether the impact would produce anything visible.]
The impact flash was a really rich spectral record. It was a slow riser; took 700 ms to rise [to maximum brightness]. We are looking for emission lines of gas species. Subtracting solar spectrum works really well. In first .8 sec, little ejecta in sunlight but some emission lines. By 3 sec, some ejecta in sunlight, emission lines persist. By 16-18 seconds, strong emission lines fade, still more ejecta emerges. Emission is from sodium, ammonia, H2O+ as well as O. Emission lines develop and appear over time. Very strong emission lines from elemental silver, Ag. CO2O+, CN, OH as prompt emission by dissociation of water, CS also.
Hydrogen is present in multiple forms. Could it be due to residual hydrazine? Estimates from project engineers conclude much less than 1 kg of hydrazine was left in the Centaur, 10-9 [that's one billionth] less than observed water vapor abundance.
Schultz points out interesting issue with interpreting images: surface is in the dark but receives illumination reflected from opposite crater wall, so solar illumination appears to have opposite sense from what is expected.
Showed a neat animation with a "ghost" or puff moving across view of Centaur impact site. This is high-angle ejecta cloud returning to surface; it changes apparent position as spacecraft position and point of view changes.
Why multiple impact ejecta angles? Was not a simple impact. This really stretched our understanding of scaling relationships. Lots of compression early which affects mass of ejecta launched to high altitudes. Also was a hollow impactor; experiment shows hollow produces significant high-angle component. From cavitation or inward collapse of impactor.
This was an incredibly interesting experiment. Here Pete showed an incredible animation of a hollow ball being fired at a target at the NASA Ames Vertical Gun Range. It produced two distinct ejecta curtains, one low angle and one spraying nearly vertically out of the impact point, just as was apparently observed by LCROSS. Even though I knew the answer would be "no," I asked Pete later if that animation was going to be posted on the Web anywhere. I got the expected response. Pete is fanatically devoted to not sharing any data he produces; it's a running joke among Brown students that everything he does is SECRET!!! The things his students do are, they are told, to be kept in the secret part of their secret brains.
At this point I moved on to the concurrent session on Terrestrial Planet Cryospheres for one talk, by Candy Hansen on observing spring coming to Mars' northern hemisphere.
In southern hemisphere, spring brings bizarre phenomena -- cryptic terrain, spiders, sublimation from bottom of CO2 ice slab, fans of fine surface material. Does the northern cap experience the same thing? If not, why not?
The dunes in the north polar erg exhibit a wealth of seasonal phenomena. Kieffer's model, developed for south polar cryptic terrain: idea is that seasonal CO2 ice forms a somewhat transparent layer, translucent enough to get sunshine all the way through the ice to warm surface below. If that happens, surface warms, get gas pressure, gas tries to escape through weak spots, dust entrained within gas blasts is carried downwind and deposited as fan.
In the north we do see bright fans, but no araneiform ["spider"] terrain. We do see dark fans and cracks in seasonal ice. Dunes are where the action is in the north. There seems to be a big difference in expression of sublimation activity depending on nature of surface material.
nly dune materials are loose enough to be mobilized. PSP_007043_2650 shows cool barchans with dark fans.PSP_007043_7043, PSP_007676_2650, PSP_007887_2650, PSP_008032_2650, PSP_008230_2650 an awesome time series on a defrosting barchan. Bright/dark banding at interface between dune and substrate...it's really hard to explain. Does it result from movement of material? Tricks of lighting? Is it enhanced sublimation on little ripples? Are there different ice transparencies? PSP_008968_2650 shows completely defrosted; streaks are gone.
Another series showing lengthening streaks: PSP_007404_2640, PSP_007905_2640. Propose streaks are dry mass wasting driven by the volatility of the sublimation process.
ne image catches cloud of dust -- unfortunately number cut off at top of slide. Catching it in action proves it is a dry process. Kieffer model applied to dunes: sub-ice gas flow may dislodge loose material on slip face (volatiles travel up dune crest to top).
Now to Tuesday afternoon. I started in the session on MESSENGER at Mercury.
Olivier Barnouin spoke on re-evaluating the shape of craters on Mercury. Why is the transition diameter from simple to complex craters different on Mercury than Mars? Might be Mars' crust was weak, sediment-rich, while Mercury's was stronger, dynamic. It might be difference in impact velocity: Mars is slower, 10-13 km/sec, Mercury is 26-43. Also differences in projectile density could play a role. But work by Strom has suggested that most impactors hitting Mercury are asteroids.
Current work: measure depth to flat floor of crater with MLA [laser altimeter], measure diameter from imaging. Use Trask's classification scheme for craters: from least to most degraded, class from 1 to 5. In general, all craters measured from MESSENGER data are shallow relative to Mariner 10 data. But few class 4 and no class 5 craters encountered yet. [There are only two MLA passes, so the sampling isn't too complete!] For simple craters, we find that they are shallow relative to Mariner 10. Complex craters typically lie on lower bounds established with Mariner 10. An exception is near transition diameter [diameter at which craters transition from simple bowl shape to complex peak or peak ring craters] around 10 -12 km.
On Mars you expect little melt in an impact, possibly deeper transient craters, so deeper simple craters. On Mercury you should get significant melt and possibly shallow transient craters, so shallower simple craters and a large transition diameter.
This isn't what's seen. So instead consider Mars to have lower crustal strength, little melt, greater collapse of crater. So propose that difference in transition diameter between Mars and Mercury is result of combined effect of greater strength of Mercury and higher impact velocities.
Louise Prockter spoke on evidence for recent volcanism on Mercury. For three decades since Mariner 10, the origin of smooth plains has remained uncertain. We've now established that volcanism was a ubiquitous process early in Mercury's history. But how long did it persist?
ven though third flyby didn't go quite the way we planned, we found evidence that volcanic activity persisted for quite a long time. Looking at "Manet" crater from M3 flyby [not yet an officially approved name; refers to crater visible in northern hemisphere on M3 inbound crescent image that has a prominent dark ring in color images]. Central flat floor has much higher reflectance than outer ring. Higher reflectance material has spilled out to south. We believe this is a volcanic deposition. This is the first time we've seen this on Mercury -- a flood of lava over the lower rim of a peak ring. We see a lot of these high reflectance plains completely within basins -- Manet, Renoir, Raditladi.
Nearby Manet is a bright, diffuse spot, is probably a pyroclastic vent. Both Raditladi and Manet have interior extensional troughs, but they have differing patterns. Extensional tectonic features on Mercury are rare.
Manet is resolvably older than Raditladi. Both are young however. The inner plains of Manet may be one of the youngest stratigraphic markers for volcanic activity on Mercury. We think Raditladi could be as young as 1 billion years old. Even though Manet crater is older than Raditladi crater, the volcanism inside Manet happened later than Raditladi.
Massironi spoke on deriving new crater counting curves for Mercury from main belt production, rather than by relating to Moon's impact history; also assumes rheological profile for Mercury's surface of 10 km regolith, underneath which is 40 km hard crust, then mantle. With these assumptions, he gets kink in crater production curve at right diameter as a result of rheological transition. But not so for younger impact basins; need to adjust rheological model, a constant feedback between observation and geological interpretation. At end of analysis, get age determination for Rembrandt around 3.6(?) billion years, other basins as young as 1.2. Modeling the production function is a flexible tool for obtaining ages because it can consider a variable flux with time. Rembrandt had volcanism soon after basin formation and with age similar to Caloris inner plains, 3.65 vs 3.7 billion years. Raditladi has an age of 1.20 billion years, as do internal deposits; smooth plains mainly of impact melt origin.
On to the session on Vesta before Dawn.
Vishnu Reddy spoke on using IRTF to study Vesta spectroscopically. He remarks on asking a fellow researcher to help dig up some older data that "was taken in 80s, when I was 2 years old" which got a bit of a groan from the audience. Vesta has at least 2 distinct compositional units. Is there olivine? Possible olivine-rich region between 90 and 150 longitude. 2007 data support a diogenite-rich region rather than pure olivine. Can't rule out the possibility of olivine presence because of incomplete rotational coverage. Southern hemisphere composition seems similar to northern hemisphere. Findings are consistent with HST observations.This last item was particularly cool. We have already been observing Vesta thru Dawn's spare filters, stuck the spare filters on a telescope in Hawaii which I can operate from my cell phone. Some data in his presentation was only 11 hours old.
Back to the session on Mars aeolian processes.
Matthew Chojnacki on recent changes in dunes in field at southeast side of Endeavour. Since early Mars telescopic observations, surficial albedo variations have been observed. But no observable dune migration was observed from Viking to MGS. At the same time, both rover sites have observed minor bedform modification. Spirit rover has found sand-sized grains atop 66-cm-high rover deck, indicating Mars wind can saltate sand effectively. Using MOC NAC, HiRISE, and CTX images, shows [if I remember correctly, I didn't write the number down] five spots in Endeavour where dunes disappeared or shrank with time. Dunes that showed most change are on edge of dune field. Have not observed any increase in area, only decrease.
Silvestro spoke on movement observed in dark dunes in Nili Patera. Three kinds of modification were observed: changes in ripple pattern, [something I missed,] and defect migration of Y junctions. Sense of motion is interesting: ripples move downwind, but ripple junctions appear to move upwind because the ends of dunes move slower than their middles. Also see change in shape of dune field edges. [I wrote an editorial comment in my notes here, that all of Silvestro's work was on one pair of HiRISE images spaced only 3 months apart, so there should be LOTS more dune motion waiting to be discovered in other HiRISE images.] Migration of ripple terminations about 2 meters implies high-stress winds blowing from east-northeast. Saltation events are frequent enough to prevent the formation of a stabilizing crust on top of the dunes.