deep impact: excavating comet tempel 1 michael f. a’hearn and the deep impact team

Deep Impact: Excavating Comet Tempel 1

Michael F. A’Hearnand

The Deep Impact Team

26 July 2007 COSPAR Workshop mfa - 2

Outline

• Review of Mission and Summary of Data• Density, porosity, & strength (energy balance)• Layering• Where is the ice?• Natural Outgassing• Activity & Outbursts• Dust/Ice Ratio• Craters


Deep Impact

• Mating of flyby with impactor, April 2004

• Last step prior to system environmental testing

• Impactor– 1/3 ton – 50% copper– Impactor Camera

• 10 rad/pixel• White light

• Flyby– 2/3 ton– Medium Res camera

• 10 rad/pixel• 8 filters

– High Res Camera• 2 rad/pixel• 8 filters

– Near-IR Spectrometer• 10 rad/pixels & slit• 1.05 < < 4.8 m• 230 < < 700


Interplanetary Trajectory

Launch Jan. 12, 2005

S/C

EarthOrbit

X

Earth at Encounter

Impact!July 4, 2005

Tempel 1 Orbit (5.5 yr Period)

Sun


Encounter Schematic

Tempel-1Nucleus

Shield ModeAttitude through

Inner Coma

Science and Autonav Imaging toImpact + 800 sec

ITM-1 StartE-88 min

ITM-2E-48 min

ITM-3E-15 min

Impactor ReleaseE-24 hours

TCA +TBD sec

AutoNav EnabledE-2 hr

Flyby S/CDeflection Maneuver

E-23.5 hr

2-wayS-band

Crosslink

Look-backImaging

500 km

Flyby S/C Science Data Playback at 175 kbps*

to 70-meter DSS

Flyby Science Realtime Dataat 175 kbps*

* data rates without Reed-Solomon encoding

Flyby S/C Science And Impactor Data

at 175 kbps*

64kbps


Key Results

• Approach held as many surprises as the impact event• Results from approach

– Dust has normal size distribution, overall similar to but slightly less than predicted - contrast with dust after impact

– Topographic features are very puzzling• Smooth areas that look like flows, but with puzzling stratigraphy• Many layers, possibly related to origin from cometesimals• Very non uniform distribution of “craters” in different stratigraphic units• Only two features are different in photometric behavior

– Remarkably frequent natural outbursts - rule out exogenic theories of origin

– Ice on surface, but not responsible for ambient outgassing, more likely frost from subsurface sublimation during “night”

– Nucleus is chemically heterogeneous - evolutionary or primordial?– Must rethink layering and active areas– Must rethink concept of active areas

• Results from impact– Ultra low strength (200 Pa) and very low gravity => high porosity– Very different size distribution

• Peaked at few-micron sizes• Led to obscuration of final crater• Implies surface particles, both ice and silicate, are weak aggregate particles (as

predicted by Mayo Greenberg and others)

– Ice is very near the surface (H2O <~ 20 cm; CO2 <~ 2 m)• CO2 and organics are enhanced relative to water below the surface

• Recall Harwit’s book, Cosmic Discovery• Data are all public at PDS-SBN (as of January 2006)

– Vis images reasonably well calibrated– IR spectra not well calibrated


Impactor Approach

• Original movie (not registered) to show pointing jitter

• Note one big jitter early due to ITCM. Note big jitters in last 30 seconds due presumably to dust hits

• Orientation is “upside down” mirror image of “sky” to visualize landing on oblique surface (~35° from horizontal). Ecliptic north is roughly near the bottom

QuickTime™ and aYUV420 codec decompressor

are needed to see this picture.


HRI Movie

• Much slower frame speed than with MRI

• Longer period included in movie

• “Vertical bar” immediately after impact is bleeding of the saturated CCD, not real ejecta

• Note shadow cast by optically thick ejecta




MRI Movie

• Frames every 62 msec• Initial stages of

excavation only• Small “poof” that goes

rapidly to left at onset is – hot, self-luminous plume– vapor + liquid or solid

particles– Moves @ 5 km/s

• Later ejecta are cold– Water ice survives the

ejection– Speeds start at few x

100 m/s and drop to below escape velocity as excavation continues




Volatile Composition

0.00

0.50

1.00

1.50

2.00

2.0 2.5 3.0 3.5 4.0 4.5

Pre-Impact Nucleus

Model

Impact

Wavelength (µm)

H2O

CH-X

CO2

HCN

First 0.2 sec


ITS Sequence

• ITS images - impact site indicated by arrows (now right side up - ecliptic north in upper right quadrant, sun to right)

• Sense of rotation - top is approaching (P ~ 40 hrs)• Oblique impact - 36° from horizontal by shape model but 20

to 35° from assuming circular craters


ITS Composite Image

• Note geological features– Large, smooth surfaces– Round features = craters?

(size-freq plot consistent)

– Stripped terrain (old)– Scarps– Evidence of layers

• Overall Shape– Effective radius 3.0±0.1

km– Max-min diameters 7.6 and

4.9 km but still uncertain– Well-mapped surface is

mostly in 3 large, more-or-less planar areas, i.e. the shape is as close to pyramidal as to ellipsoidal

• Impact site is between two craters near bottom of image.


Volume well determined from back-illuminated limb even though topography is determined only on one side

g is a global measurement

Our g is consistent with non-grav acceleration models

Uncertainties

Gas acceleration

early stage only!

Angle of ejection

Neither likely to be a

very large effect

If dust/ice by mass ~1

porosity > 80%

Shape Model


Why Are Shapes So Different?

• Nuclei are obviously different– Overall shape, topographic features, amount of activity– Nothing is obviously correlated with dynamics or age

except that the only Oort cloud comet is more active than the others

• What is common?– Source of activity? Depth of ice? Heterogeneity?

Cratering? Other topographic features? Basic physics of how comets work? Origin?

• What is the pattern that will tell us about origin of solar system?

Basilevsky & Keller 2007 Sol.Sys.Res.41, 109


Fallback of Ejecta Yields g

• Assuming ballistic trajectories and gravity scaling!!!

• Limit of lift-off of cone gives upper limit to strength – Orig <= 200±100 Pa– now < 12 kPa– Could --> 0

• Measure width of base of plume

• Yields local gravity about 50 mgal at impact site (g = 0.05 cm/s2) & escape velocity ~1.3 m/sec

• Shape model yields mass, 2x1016 g, assuming uniform density.

• Density = 0.4 g/cc (if uniform; was 0.6±0.3 in Science paper Richardson et al. 2007, Icarus, submitted


Does Gas Acceleration Matter?

• Our model assumes ballistic trajectories– Adding gas acceleration adds many free parameters

• Total gas ~4x1032 molecules, mostly H2O• Much H2O is still ice at I+45 minutes

– No quantitative model yet for amount of ice– Pure ice grains should last ~5 hours at 1.5 AU

(Patashnick & Rupprecht 1977)– Grains come out in a hollow cone while subsurface

volatiles (immediate sublimation) come out in center of cone dragging extra grains (Schultz experiments)

– Cone still optically thick up to ~7 km at I+45 (Holsapple & Housen, in press)

• Assume total gas over 5 hours from sphere of R~3 km– Average flux ~< 0.1 sub-solar free sublimation at

1.5 AU, i.e., smaller drag than normal outgassing• Gas drag matters but probably not a major change

– Ignoring lateral acceleration probably overestimates g


Strength

• Lack of detachment of ejecta means upper limit only!!! Strength itself not measurable

• What kind of strength– Tensile mostly, rather than shear or compressive

(compressive >> shear ~ tensile)– Dynamic, post-shock strength rather than static

• How Measured?– Max height of detachment (few 100m) yields upper

limit on minimum ejection velocity– Equate upper limit on KE/m to upper limit on

dynamic strength (using bulk density) --> 200 Pa (talcum powder)

• Issues– Static strength > dynamic strength– Data consistent with 12 kPa static strength

(Holsapple & Housen)– Implies less mass in ejecta, still strength <<

ice or rock


Energy & Momentum

• K.E. of impactor: 19 GJ• Orbital Change

– < 1 GJ from change in orbital energy– Momentum transfer efficiency perhaps 2x-3x (model

dependent)– Depends on obliquity of impact (ejecta momentum

not anti-parallel to impactor momentum)

• Hot Plume (~100 ton)– K.E. of plume has most of the impact energy

• May need additional internal energy source

– Sublimation and melting has 10% or less of impact energy

• Excavated material (~104 ton)– K.E. << 1% of impact energy, but momentum exceeds

input momentum– Sublimation of water MUST be due to sunlight

evaporating excavated ice; total energy of sublimation >> impact energy


ITS Composite Image

• Note geological features– Large, smooth surfaces– Round features = craters?

(size-freq plot consistent)

– Stripped terrain (old)– Scarps– Evidence of layers

• Overall Shape– Effective radius 3.0±0.1

km– Max-min diameters 7.6 and

4.9 km but still uncertain– Well-mapped surface is

mostly in 3 large, more-or-less planar areas, i.e. the shape is as close to pyramidal as to ellipsoidal

• Impact site is between two craters near bottom of image.


Deep layers wrap around,roughly parallel to “top”

facet.Regional differences in

erosion and array oftopographic forms

Why is the boundarybetween regions parallel to

body layers?

Layers


Other Features

Thomas et al. 2007 Icarus 187, 4


Layering on east facet

• Red-blue anaglyph for 3-D view shows that layer wraps around to dark side

• Fringes on right are an artifcact of deconvolution of the HRI image responding to the boundary of the image


TALPS Model

• Observationally, layering is not onion skin, but rather randomly oriented but extending over large areas

• Accretion regime allows flattening and ejecta blanket rather than fragmentation during collisions

• Based on ideas of Donn (interpenetration + compaction of fractals) and Sirono & Greenberg (deformation + compaction of fractals)

• Talps are primordial, not due to geologic evolution

• Expect compositional differences in talps

Belton et al. 2007 Icarus, in press


Thermal Map of Nucleus

• First real thermal map of a nucleus

• Consistent with STM plus roughness to warm areas near terminator; I~<20 W K-1 m2 s0.5

• No locations as cold as sublimation temperature of H2O ice

• Therefore ice must be below the surface but “not far” below

• Diurnal skin depth 3 cm, annual skin depth 0.9m for plausible separation of components of I

Groussin et al. 2006 Icarus, submitted


Where Is the Ice?

• No bulk ice on surface to explain active areas - all the major sources of water are below the surface

• There is a lot of H2O ice <10 cm below the surface over large areas

• There is a lot of CO2 within 1 m of surface in some reasonably large areas

• There is considerable heterogeneity in chemical composition as well as surface topography


Anomalously Colored Regions

Deconvolved High Resolution

Color Images

Sunshine et al. 2006, Science 311, 1453


0.4

0.8

1.2

1.6

1.2 1.6 2.0 2.4 2.8 3.2

Ice-Rich Nucleus

4% 20 µm Ice + 96% Non-Ice Nucleus

Non-Ice Nucleus (offset)

Wavelength (in µm)

Water Ice Absorptions

0.6

0.8

1.0

1.2

1.4

1.6

1.2 1.6 2.0 2.4 2.8 3.2

Ice-Rich Nucleus3% 70 µm Ice +97% Non-Ice Nucleus9% 5 µm Ice +91% Non-Ice Nucleus

Wavelength (in µm)

Water Ice Absorptions

Modeling Surface Water Ice

• Nominal (non-ice) nucleus + laboratory water ice– 3-6% water ice– 30 ± 10 µm size particles

• Not enough surface to be significant in overall outgassing• Frost from source of outbursts on shoulder?

Sunshine et al. 2006 Science 311, 1453


Activity off Limb


Ice in Early Ejecta

• Each row is a spectrum with slit fixed in position

• Early rows are 0.7-sec exposures

• Later rows (longer slit lengths) are 1.4- and 2.8-sec exposures

• Ice is present by fourth spectrum, i.e., by <3 seconds after impact

• Ice is concentrated in the down-range direction

Sunshine et al. 2007. Icarus, in press.


Ejecta - Icy and not

Sunshine et al. 2007.Icarus, in press


Structure Summary

• Fine-grained material– No boulders– No hard crust

• Grains are fragile aggregates – fragment during

excavation– Fragements ~1-3 m

• Layers within 1 impactor diameter of surface at impact site– Topmost layer (few cm?)

devoid of ice (see later slides)

• Layers are ubiquitous– Varying thickness– Some may be primordial– Smooth layers not yet

explained

Schultz et al 2007, Icarus, submitted


Detection of Asymmetric Inner Coma

• 1 hour before impact• ~440 m/pixel resolution• Northern and southern regions

examined• Spectra show comparable H2O but

factor of 2 increase in CO2 relative to H2O in the south

• 1 hour before impact• ~440 m/pixel resolution• Northern and southern regions

examined

Sun

Ecliptic North


Spatial Distributions Vary by Species

P = positive rotational poleE = Ecliptic northS = Sunward

Feaga et al. 2007. Icarus, in press

Farnham et al. 2007. Icarus, 187, 26

Dust is better correlated with CO2 than with H2O, but not perfectly with either


Results from Spectral Maps

• Optically thin at edges of field

• Ambient Q(H2O) ~ 4x1027 s-1

– Good agreement with ground-based measurement in FOV 1000x larger

• Q(CO2) ~ .07 Q(H2O)– Few x variation with azimuth

• Dust not well correlated with either H2O or CO2

– Better with CO2

– How does dust get lifted?– Models for dust ejection need to be rethought


Approach Photometry

A’Hearn et al. 2005 Science 310, 258

Outbursts common - typically 2 per weekOutbursts correlated with rotational phase (2 phases with at least 3 outbursts each)Thus, outbursts are endogenic and related to surface insolationProbably super-volatiles close below surface but Dina Prialnik can make them with water and Stochastic nature of outbursts not understood

2 July Outburst by D. Lindler

08:51

11:57 10:47

09:52


Monitoring OH

Küppers et al., 2005 Nature 437, 987Observations with OSIRIS on Rosetta

Enables determination of total water released in impact ~ 4000 tons original estimate, revised upward (4500 - 9000 tons) with better calibration (Keller et al. 2007 Icarus 187, 87)

Many other observers (incl. ODIN & ground-based OH) find 4 to 10x103 tons of water

Other species (CO & CO2 best determined) of order 5-10% of water

Total volatiles - est 8x103 tons ±3×103


Refractory Ejecta

• Schleicher et al. need predominantly 0.5 to 2.5 m grains– Avoid estimating total mass, but no particles with < .13– Similar analysis of OSIRIS data by Jorda et al. finds no need

for an unusual size distribution– Optical data are not sensitive to particles >> few microns

• Most distributions have total mass sensitive to largest particle

• Very hard to estimate total mass of refractories!!!

Schleicher et al. 2006 AJ 131, 1130

I - preimpact

Above - a/

Model - a/


Dust/Ice Ratio

• Fallback– 90% of solid ejecta fall back to surface in first

hour– Does the ice get buried deeply enough to not

vaporize?– Or does the H2O correspond to all ejecta while

refractory mass from remote sensing corresponds only to escaping fraction?

• Size cutoff– Radiation pressure effects do not require an unusual

distribution to explain optical data (e.g., Jorda et al.) but optical data not sensitive to large particles

– There is IR evidence for changes in the size distribution after impact compared to before (Lisse et al. Science, Subaru & Gemini data - Diane Wooden’s talk)

• Quoted dust masses vary dramatically– Probable range 103 to 104 tons– I.e. Dust/Ice ratio 0.1 to 1– If fallback of ice grains still leads to excess

sublimation, Dust/Ice ratio could go up to 10


Dust Before and After

4 Dust hits measured by ADCS correlated with image pointing between -21 and -3 sLast hit was at -3 s, after last transmitted image and had largest massExpectation values of mass

3 between 1 - 10 mg, 1 (last hit) 100-1000 mgImpactor cross-section about 1.1 m2

• Size distribution needed to successfully model SST spectra

• Distribution of surface area per unit mass - before (ambient release) and after (mechanically excavated grains) impact

• While largest particles still dominate total mass, they no longer dominate total cross-section

• Interpretation is that surface materials are all weak, large aggregates of smaller pieces with typical size of a few m

Best-Fit Model

PAHs

Pyroxenes

Spitzer IRS I+45 Min

FeMgS2

Olivines

Carbonates

(Pre

-Pos

t)/P

re =

Eje

cta/

Pre

-Im

pact

Com

a

(95% C.L. = 1.13)

Lisse et al. 2006 Science in press

Amorph Carbon

Water IceSmectite (clay)


An Aside on Craters

• There are ~ 60 circular features on Tempel 1• Size distribution matches impact craters on

Ida, Gaspra, etc. (Thomas et al. in press)• Size distribution on Wild 2 is bimodal and

nothing like the size distribution other places in the solar system (Thomas, priv. comm.)

• Circular features appear in many different layers on Tempel 1

• The two largest “craters” appear to be exhumed - when did they form & how did they get filled?


The Bottom Line

DI will produce new results for a long timeOnly need is scientists to use the data

Comets are still fascinating and full of surprises

More missions to comets are needed


More Missions

• EPOXI– Uses the DI flyby

spacecraft– Flies to 85P/Boethin in

Dec 2008– On the way observes

transits of extrasolar planets

• Stardust NExT– Uses the Stardust

spacecraft– Flies to 9P/Tempel 1 on

14 Feb 2011

Backup Slides


Approach Photometry

A’Hearn et al. 2005 Science 310, 258

Outbursts common - typically 2 per weekOutbursts correlated with rotational phase (2 phases with at least 3 outbursts each)Thus, outbursts are endogenic and related to surface insolationProbably super-volatiles close below surface but stochastic nature of outbursts not understood


Ejecta over Time

Dust(2 µm reflectance)

H2O Ice Absorption

H2O Gas Emission

TimeImpact

Sunshine et al. 2007 Icarus, in prep.


Possible Scenarios

• Crater formation on an intact nucleus– Gravity controlled crater– Compression controlled crater

• Split nucleus• Crater formation on an intact nucleus

– Strength controlled crater• Aerogel-like capture of the impactor• Shattered nucleus• Transit through the nucleus

• Above are roughly in order of decreasing probability (as guessed by the PI)

• N.B.: K.E. of Impactor << Gravitational Binding Energy of Cometary Nucleus

• N.B.: v2/2 > maximum energy per mass of any chemical explosive

D. K. YeomansCSR page 1-12


Deconvolved HRI Image

• IR + green + violet• Forced to average gray• Note very localized

“bluish” areas• Note curvature of ejecta

in up-range direction– Consistent with lab

experiments– Later (I+195s) detachment

of these rays from crater suggests layering

– Layering also suggested by hot plume in previous movie

– Schultz et al., in prep.

• Note smoothness of ejecta in radial direction– Primarily small particles– Rays from initial

conditions


Spectral Mapping

• Slit runs right -to-left in normally displayed images– One end of slit points to

sun when solar panels are normal to sun direction

• Spatial scans are made in orthogonal direction

• Get a spectrum for each point in slit at each point in scan.– Timing is complicated

• Slit read out one end to the other (r to l in this image)

• Spatial scans continuous, not step & integrate

• Net projected slit is tilted depending on scan rate

• Spectral resolving power >200 everywhere, ~450 at 1.8 m, ~700 at 1.05 m

• Currently believe calibration only 2.0 to 4.5 m


Spectra off Southern Limb

Post-impact spectrum obviously different from pre-impact spectrum

Initial ejecta are hot but after few seconds ejecta leave nucleus cool

At end of observations, composition seems to be asymptotic suggesting we excavated “deep enough”!

CO (4.7 m) and minerals in reflection (1.05 to 2 m) waiting on improved reduction of spectra at ends

Most species have optically thick lines so that column densities are not easy to determine

Rad

ian

ce (

W/[

m2 s

r m

])

Wavelength (m)

H2O

CO2

CH-X

Pre-Impact Post-Impact

CO2

CH-X


0.6

0.8

1.0

1.2

1.4

2.0 2.5 3.0 3.5 4.0

Wavelength (µm)

Thermal

Solar

Continuum Removed

Average Early Ejecta

3 µm Absorption– Present even without thermal component

– Consistent with microscopic crystalline H2O ice particles

• Near-Surface Water Ice• Ejecta are cold, thus unprocessed!!

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

2.0

2.0 2.5 3.0 3.5 4.0 4.5

Wavelength (µm)


Dust on Impactor

4 Dust hits measured by ADCS correlated with image pointing between -21 and -3 sLast hit was at -3 s, after last transmitted image and had largest massExpectation values of mass

3 between 1 - 10 mg, 1 (last hit) 100-1000 mgImpactor cross-section about 1.1 m2


Size Distribution at E+45m

• Size distribution needed to successfully model SST spectra

• Distribution of surface area per unit mass - before (ambient release) and after (mechanically excavated grains) impact

• While largest particles still dominate total mass, they no longer dominate total cross-section

• Interpretation is that surface materials are all weak, large aggregates of smaller pieces with typical size of a few m

Lisse et al. 2006. Science, in press


Nature of Ejecta

• Total ejecta from scaling laws - 1-2x107 kg

• Total volatiles (Küppers H2O, Biver H20, Feldman CO) - 5x106 kg, but with Schleicher’s H2O, 15x106 kg

• Mass of refractory dust sensitive to large end of size distribution ~104 kg

• Dust/Ice ratio 100±0.5 from considering various sources• Particle size - mostly < 10 m

• Superheat• Mass per cross-section• Radiation pressure model (Schleicher)

– But see possible contrary view by Keller

• Strong mid-IR emission features (Lisse et al.)• Implies pre-existing 1m grains or very weak aggregates thereof to depths of tens of

meters

• Particle composition– Mix of silicates, organics, and volatile solids

• Ice detected in situ in ejecta

– Additional refractories seen with Spitzer (carbonates?)

• Gaseous Ejecta

• In situ large increase in CO2/H2O), larger increase in organics/H2O

• Remote sensing - less variation relative to water because water came out as ice & then sublimed

• Some species large variation


Key Results

• Approach held as many surprises as the impact event• Results from approach

– Dust has normal size distribution, overall similar to but slightly less than predicted - contrast with dust after impact

– Topographic features are very puzzling• Smooth areas that look like flows, but with puzzling stratigraphy• Many layers, possibly related to origin from cometesimals• Very non uniform distribution of “craters” in different stratigraphic units• Only two features are different in photometric behavior

– Remarkably frequent natural outbursts - rule out exogenic theories of origin– Ice on surface, but not responsible for ambient outgassing, more likely frost from

subsurface sublimation during “night”– Nucleus is chemically heterogeneous - evolutionary or primordial?– Must rethink layering and active areas– Must rethink concept of active areas

• Results from impact– Ultra low strength (200 Pa) and very low gravity => high porosity– Very different size distribution

• Peaked at few-micron sizes• Led to obscuration of final crater• Implies surface particles, both ice and silicate, are weak aggregate particles (as predicted

by Mayo Greenberg and others)

– Ice is very near the surface (H2O <~ 20 cm; CO2 <~ 2 m)• CO2 and organics are enhanced relative to water below the surface

• Recall Harwit’s book, Cosmic Discovery


Bigger Implications & Issues

Comets formed gently to preserve bulk porosity– How does this happen?– TALPS scenario proposed by Belton et al. submitted to

Icarus– Does scenario also require radial mixing in protoplanetary

disk?• Sample return easier than previously thought

– Penetrating to the water ice is easy if one can hold the spacecraft down

– E.g., 10N thruster may be enough for coring a sample– Large (1 gram) particles are fragile

• “Active areas” are not evident on the surface– Bulk of H2O sublimation is from sub-surface but close

enough to surface to see diurnal variation (~< 10 cm)– CO2 remembers seasonal thermal wave at negative pole (depth

~1m)– Dust jets better correlated with CO2 than with H2O, but not

well with either– Possible (probable?) primordial chemical heterogeneity in

volatiles would require radial mixing in proto-planetary disk

– Outbursts are endogenous & deposit frost of ice very locally between outbursts


Volume well determined from back-illuminated limb even though topography is determined only on one side

g is a global measurement

Our g is consistent with non-grav acceleration models

Uncertainties

Gas acceleration

early stage only!

Angle of ejection

Neither likely to be a

very large effect

If dust/ice by mass ~1

porosity > 80%

Shape Model


The darker area with close round depressions is thepart of Tempel 1 that most closely resembles Wild-2.


Fallback of Ejecta Yields g

• Simulation assumes ballistic trajectories and gravity scaling!!!

• Measure width of base of plume

• Yields local gravity about 50 mgal at impact site (g = 0.05 cm/s2) & escape velocity ~1.3 m/sec

• Shape model yields mass, 2x1016 g, assuming uniform density.

• Density = 0.35±0.12 g/cc (if uniform; was 0.6±0.3 in Science paper; this error 1- Icarus paper will quote 2- errors)

Richardson et al. 2006, Icarus, submitted


Thermal Map of Nucleus

• First real thermal map of a nucleus

• Consistent with STM plus roughness to warm areas near terminator; I~<20 W K-1 m2 s0.5

• No locations as cold as sublimation temperature of H2O ice

• Therefore ice must be below the surface but “not far” below

• Diurnal skin depth 3 cm, annual skin depth 0.9m for plausible separation of components of I

Groussin et al. 2007 Icarus, 187, 16


The darker area with close round depressions is thepart of Tempel 1 that most closely resembles Wild-2.


Where is ice?

• Uprange rays have no ice and come from a layer much less than size of impactor (<few 10s of cm)

• Downrange rays have ice and come from a layer comparable to impactor size (~ 1m)

• Ice appears downrange within 2 seconds

Dust

Ice

Sunshine et al. 2006 LPSC


EPOXI -

deep impact: excavating comet tempel 1 michael f. a’hearn and the deep impact team

Documents

impact eventresults

impactor dataat

oblique surface

coldwater ice

impactultra low strength

normal size distribution

low gravity

impacttopographic features