Cassini radar observation of Punga Mare and environs: Bathymetry and composition

2018

Publication Year

2021-04-22T14:39:09Z

Acceptance in OA@INAF

Cassini radar observation of Punga Mare and environs: Bathymetry and

composition

Title

Mastrogiuseppe, M.; Poggiali, V.; Hayes, A. G.; Lunine, J. I.; Seu, R.; et al.

Authors

10.1016/j.epsl.2018.05.033

DOI

http://hdl.handle.net/20.500.12386/30859

Handle

EARTH AND PLANETARY SCIENCE LETTERS

Journal

496

Number

JID:EPSL AID:15086 /SCO [m5G; v1.236; Prn:23/05/2018; 14:31] P.1 (1-7)

Earth and Planetary Science Letters•••(••••)•••–•••

Contents lists available atScienceDirect

Earth

and

Planetary

Science

Letters

www.elsevier.com/locate/epsl 1Q1 67 2 68 3 69 4 70 5 71 6 72 7 73 8 74 9 75 10 76 11 77 12 78 13 79 14 80 15Q2 81 16Q3 82 17 83 18 84 19 85 20 86 21 87 22 88 23 89 24 90 25 91 26 92 27 93 28 94 29 95 30 96 31 97 32 98 33 99 34 100 35 101 36 102 37 103 38 104 39 105 40 106 41 107 42 Q4108 43 109 44 110 45 111 46 Q5112 47 113 48 114 49 115 50 Q6116 51 117 52 118 53 119 54 120 55 121 56 122 57 123 58 124 59 125 60 126 61 127 62 128 63 129 64 130 65 131 66 132

Cassini

radar

observation

of

Punga

Mare

and

environs:

Bathymetry

and

composition

M.

Mastrogiuseppe

a

,

b

,

V.

Poggiali

b

,

A.G.

Hayes

b

,

J.I.

Lunine

b

,

R.

Seu

a

,

G.

Di Achille

c

,

R.

Lorenz

d

a_{University of Rome “La Sapienza”, Via Eudossiana 18, 00184 Rome, Italy} b_{Department of Astronomy, Cornell University, 14853 Ithaca, NY, USA}

c_{INAF National Institute for Astrophysics, Astronomical Observatory of Abruzzo, Teramo, Italy} d_{Johns Hopkins University, Applied Physics Laboratory, 11100 Johns Hopkins Road, Laurel, MD, USA}

a

r

t

i

c

l

e

i

n

f

o

a

b

s

t

r

a

c

t

Article history:

Received23December2017 Receivedinrevisedform20April2018 Accepted20May2018 Availableonlinexxxx Editor: W.B.McKinnon Keywords: radar altimetry planets sounder Titan bathymetry

InJanuary2015(fly-byT108),theCassiniradarobservedPunga Mare,Titan’snorthernmostand third-largest sea,inaltimetrymode duringclosestapproach. The groundtrack interceptedasectionof the mareandasystemofchannelsandfloodedareasconnectingPungatoKrakenMare.Weuseaprocessing technique, successfullyadopted for LigeiaMareand Ontario Lacus, for detectingechoes fromthe sea floor andconstrainingthedepthandcompositionoftheseliquidbodies.Wefindthat,alongtheradar transect,PungaMarehasamaximummeasureddepthof110m.Therelativereductioninbackscatterof theseafloor,asafunctionofincreasingdepth,suggestsaliquidlosstangentof3±1×10−5_.Whilethis

value iswithintheformaluncertaintyofthelosstangentderivedforLigeiaMare,thebest-fitsolution islowerandisconsistentwithanearlypurebinarymethane-nitrogenliquidwithlittletonoethaneor higherordercomponents.Theindicationofverylowamountsofethanetowardthepolesuggeststhat atmosphericprocessesarecontrollingthesurfaceliquidcompositionofTitan’sseas.

©2018PublishedbyElsevierB.V.

1. Introduction

Titan’ssurface hasbeenwidelymappedby theCassini RADAR (2004–2017),amicrowaveremotesensinginstrumentableto pen-etratethedense atmosphereofthemoonat2.17cmwavelength. The Cassini radar was a multimode instrument capable to op-erate in active mode as a Synthetic Aperture Radar (SAR) for surface imaging, as a radar altimeter for topography measure-ments,asascatterometerforsurface composition and,inpassive mode, as a radiometer for brightness temperature (Elachi et al.,

2004). The instrument modeswere activated sequentially during each fly-by to Titan, from an altitude of 100,000 km down to a 1,000 km at the closest approach, pointing the antenna in a convenientway toaccomplish the targeted measurements.A de-tailed description of sequence planning and instrument perfor-mance is reported in West et al. (2009). A total number of 53 fly-bys dedicated to the radar observations, allowed Cassini to cover

∼

50%ofTitansurface at

<

1 kmresolutioninSARmode,as wellastoacquire40topographicprofilesinaltimetrymode.This datasetenabledtheidentificationandcharacterization ofaseries

E-mail address:marco.mastrogiuseppe@uniroma1.it(M. Mastrogiuseppe).

ofgeomorphologicfeatures, includinghundredmetershighdunes (Mastrogiuseppe et al., 2014b), fluvial network of channels and canyons(Poggialietal.,2016),mountains(Radebaughetal.,2007; Mitrietal.,2010),craters(Woodetal.,2010),possiblecryovolcanic features (Lopesetal.,2013) andlargedepositsofliquid hydrocar-bonsinlakesandseas.Adetaileddescriptionandmappingofthe Titan’spolarterrainsisreportedinBirchetal. (2017).

The presence of standing hydrocarbons liquidbodies on Titan wasrevealedbyCassinionJuly22nd,2006,duringthefly-byT16, whentheradarmappedacollectionof10–100 kmdiameterlakes present in the Northern hemisphere (Stofan, 2007). Later obser-vations revealed the existence of three northern seas, or maria: KrakenMare,LigeiaMareandPungaMare(Hayesetal.,2008).

The altimetric observation acquired in May 2013 (flyby T91) over Ligeia Mare demonstrated that the Cassini RADAR can also operate as a sounder, capable of probing Titan’s seas down to

∼

200 m,dependingmainlyonliquidcomposition(Mastrogiuseppe

etal.,2014a).Thiswaspossiblebecauseoftheverylowmicrowave

absorptivity of liquid hydrocarbon (methane and ethane), which has a microwave loss tangent (defined as the ratio between the imaginary andrealcomponentsofthedielectric constant)that is approximatelyfiveordersofmagnitudelowerthan seawater.This successful experimentsuggesteda re-design ofthe final targeted https://doi.org/10.1016/j.epsl.2018.05.033

JID:EPSL AID:15086 /SCO [m5G; v1.236; Prn:23/05/2018; 14:31] P.2 (1-7) 2 M. Mastrogiuseppe et al. / Earth and Planetary Science Letters•••(••••)•••–•••

1 67 2 68 3 69 4 70 5 71 6 72 7 73 8 74 9 75 10 76 11 77 12 78 13 79 14 80 15 81 16 82 17 83 18 84 19 85 20 86 21 87 22 88 23 89 24 90 25 91 26 92 27 93 28 94 29 95 30 96 31 97 32 98 33 99 34 100 35 101 36 102 37 103 38 104 39 105 40 106 41 107 42 108 43 109 44 110 45 111 46 112 47 113 48Q13 114 49 115 50 116 51 117 52 118 53 119 54 120 55 121 56 122 57 123 58 124 59 125 60 126 61 127 62 128 63 129 64 130 65 131 66 132

Fig. 1. (Upperpanel)SARmosaicandrelativeoutboundaltimetric700-km-longtrackofflybyT108withredcirclesindicatinghalfpowerfootprintsand,inwhite,thetwo selectedregionsshowninFig.2.(Middlepanel)Radargramandrelativealtimetryobtainedusingdifferenttrackingmethods.(Lowerpanel)Radarcrosssectionobtainedfrom altimetry.Notethattheabruptchangestoveryhighvaluesofradarcrosssectionindicatethepresenceofexposedliquidinterceptedbytheradar.(Forinterpretationofthe colorsinthefigure(s),thereaderisreferredtothewebversionofthisarticle.)

observationsofTitanaimedatacquiringsimilaraltimetrydatasets over Titan’s other seas in order to investigate their depth and composition. Herein, we discussthe January 11th, 2015 observa-tion (flyby T108) of the Cassini RADAR, that acquired data over a

∼

100 km long transect through Punga Mare anda portion of thefloodedterrainconnectingPungaandKrakenMareduring the spacecraft’s closest approach to Titan, witha favorable geometry forsounding(Fig.1andSupplementarymaterial).

PungaMarehasbeenrepeatedlyobservedbytheCassiniRADAR in its high-resolution SAR mode, notably in October 2006(T19), April2007(T29) andDecember 2009(T64). Itisthethirdlargest (6

.

1

×

104 km2, Hayes, 2016) and most poleward sea on Ti-tan (85◦N, 342◦W). At the time of the T108 altimetry observa-tion, PungaMare’s surface was very smooth atthe Cassini radar wavelength(

λ

=

2

.

17 cm).Investigation ofthe surface roughness from T108 altimetry resulted in an estimated effective

σ

h

(stan-dard deviation of the surface height) ranging between 2.3 and

2.5 mm(Grimaetal.,2017),consistentwithalackofwind-waves (Hayes et al.,2013).InJuly2012(T85),theCassiniVisualInfrared MappingSpectrometer(VIMS)observedoffsetsunglintsthatwere characterized as isolated patches of increased roughness consis-tent with wind-waves (Barnes et al., 2014). In order to produce the observed glint magnitudes, the wind-waves would have re-quired“SignificantWaveHeights”(SWH

=

4

σ

h)of2+₋21cm(Barnes

etal.,2014),consistentwithwave fieldsgeneratedbylightwinds of0.4–0.7 m/s nearthe threshold forwave generation (Hayes et al.,2013).

In this paper, we adopt the dedicated radar processing tech-niquedescribed inMastrogiuseppe etal. (2016) to quantitatively investigate the seafloor topography and composition of the liq-uidbasins observed during fly-by T108. Thistechnique hasbeen successfully used to characterize the depth and composition of LigeiaMare(Mastrogiuseppeetal.,2014a,2016)andOntarioLacus (Mastrogiuseppeetal.,2018) asreportedinTable1.Asimilar

anal-JID:EPSL AID:15086 /SCO [m5G; v1.236; Prn:23/05/2018; 14:31] P.3 (1-7)

M. Mastrogiuseppe et al. / Earth and Planetary Science Letters•••(••••)•••–••• 3

1 67 2 68 3 69 4 70 5 71 6 72 7 73 8 74 9 75 10 76 11 77 12 78 13 79 14 80 15 81 16 82 17 83 18 84 19 85 20 86 21 87 22 88 23 89 24Q7 90 25 91 26 92 27 93 28 94 29 95 30 96 31 97 32 98 33 99 34 100 35 101 36 102 37 103 38 104 39 105 40 106 41 107 42 108 43 109 44 110 45 111 46 112 47 113 48 114 49 115 50 116 51 117 52 118 53 119 54 120 55 121 56 122 57 123 58 124 59 125 60 126 61 127 62 Q8128 63 129 64 130 65 131 66 132 Table 1

DepthandcompositionvaluesfromCassiniradarassumingaternarymethane–ethane–nitrogencomposition.

Mare Best fit / Limits (1σ)

Methane [%] Nitrogen [%] Ethane [%]

Mean value / (1-sigma) Mean value / (1-sigma) Mean value / (1-sigma)

Ligeia 71 / (63-78) 17 / (14-20) 12 / (2-22)

Punga and Baffin Sinus 80 / (74-80) 20 / (18-20) 0 / (0-8)

Ontario Lacus 51 / (21-74) 11 / (05-18) 38 / (08-74)

ysis,whichadoptsawaveformapproach,hasalsobeenusedto de-riveTitan’ssolidbodytopography(Mastrogiuseppeetal.,2014b).

2. T108dataanalysisandseafloordetection

Inaltimetrymode,the CassiniRADARcollected databy point-ingtheantennaboresighttowardthecenterofmassofthemoon, transmitting/receiving burstsofchirped signalswitha pulse rep-etition frequency (PRF) of 5 kHz and a bandwidth of 4.25 MHz allowing a rangeresolution equalto 35m invacuum. When the radar altimeter is used as a sounder (bathymetric investigation), therangeresolutionisdegradedto50–60 mduetotheapplication ofcustom taperfunctionsforsuppressingsidelobesassociatedto thestrong sea-surfacereflection(see Supplementary material).In thiswork weadoptsuper-resolutiontechniques(Cuomo,1992) to extendthestandardCassini radarbandwidthbya factorofthree, allowingroughly20 mresolutionforshallowdepthinvestigations ofTitanseasandlakes.

Thededicatedoutboundradaraltimetryobservationconducted duringtheT108fly-by,enabledcollectionofdataalonga700km groundtrack with spacecraft altitudes ranging from 1000km to 1200kmaboveTitan’snorthpolarterrain(Fig.1and Supplemen-tarymaterial).

Duringthisfly-by,the antennawas pointedclosetothenadir direction(

θ <

0

.

035 deg,seeSupplementaryFig. S1)andits foot-printinterceptedthesoutherneasternpartofPungaMare,aswell asafloodedareathatconnectsPungawithKrakenMare.Overthe predominantlyland surface ofthe swath, nadirglintsfromsmall patchesofliquids suchassmalllakesandrivers,similar tothose detected during the fly-by T91 (Poggiali et al., 2016), were also observed.TheprocessingresultsarereportedinFig.1,wherethe middle panel shows the radargram and relative altimetry of the observationandthe upperpanelthe SARmosaicoftheregion of interest. The altimetric profile obtained using conventional pro-cessing (Alberti et al., 2009), shows the presence of a complex topography whereflat regions representing liquidbodies are fol-lowedbyhillyregions risinghundredsofmetersabove theliquid surface along a slope ofa few degrees. Twomajor liquid bodies areobservedduring theflyby.First,locatedfrom

∼

200to350km along track (Fig. 1), is an estuary connected to northern Kraken Mare,hereaftersimplyreferredtoasBaffinSinus.Thesecond,from 470to 570km,is thesouthern part ofPungaMare. The smooth surfaceofbothliquidbodiesresultedinaverystrongspecular re-turn(Fig.1,bottompanel),permittingveryprecisemeasurements (

∼

30–50 cm) of the distance between the spacecraft and liquid surfaces.Acrossthe

∼

350kmoftrackoverPungaMare and Baf-finSinus, theliquidsurfacesareobserved tosmoothly changeby

∼

11 m (Hayes et al., 2017), consistent with estimate of Titan’s geoid variabilityover the same trackby Iess etal. (2012). Along with elevation measurements of Ligeia Mare and Kraken Mare, Hayes et al. (2017) used this observation to argue that Titan’s Maria are connected and share a common equipotential surface (likeEarth’soceans).

The results of the standard altimetric processing of the T108 radar product is shown in Supplementary Fig. S2. Radargrams showsignalsreflectedbythesmoothsurfaceoftheliquidin addi-tion to side lobes resulting from the processing (Supplementary

Fig. S2, upper panels). Reprocessing data using a custom taper function (Blackman window,BlackmanandTukey, 1959) resulted ina mitigationofside lobes(SupplementaryFig.S2,middle pan-els).Subsequently,weapplysuper-resolutiontechniquetoenhance range resolution and improve the detection of shallow (

<

35 m deep)subsurfaceechoes(seeFig.2,bottompanel).Thesame tech-niquewasimplementedontheT91radardataacquiredoverLigeia Mare fordiscriminatingsubsurfaceechoesattheshoreline ofthe sea (Mastrogiuseppe et al., 2014a). The result of the processing is shown in Fig. 2 (central panel) and the radargrams show the presenceofechoesreflected fromtheseafloorofPungaMareand nearbyliquidbodies.

InFig.2,thelower-rightpanelshowsadetectionofliquidover the relatively small Dingle Sinus, a floodedarea which connects PungaandKrakenmaria.Theresultingwaveformshowsaspecular reflection fromthe surface followedby a weaker reflectionfrom the subsurface. Althoughwe applied super-resolutionalgorithms, thesubsurfacereturncannotbeseparatedfromtheechoesatthe surface, suggestingthat the depth is shallowerthan the limit of detection(15–20 m)aftersuper-resolutionprocessing.

3. BathymetryandcompositionofPungaMareandBaffinSinus

We investigate thedepth andcomposition of PungaMare us-ing the MonteCarloapproach described inMastrogiuseppe etal. (2016).Thistechnique,alreadyappliedtothedataacquiredduring T91(Mastrogiuseppeetal.,2016) andT49flybys(Mastrogiuseppe etal., 2018),estimatesthemostprobablevaluesandrelative un-certaintiesforthethreeparametersusedforderivingtheloss tan-gent of the liquidand bathymetry: surface to subsurface relative peak power ( Ps/ Pss), two-way travel time of the echo(



τ

) and

subsurfaceroughness(

σ

h).

Subsequently, the estimated values( Ps/ Pss and



τ

) are used

alongwithaparametricmodelfordeterminingthespecificsignal attenuation K of theliquid,writtenaslogarithmic losspertravel timeasdescribedinMastrogiuseppeetal. (2016).

Alinearregressionisappliedtoasubsetofdata,includingonly selectedburstswherethe

−

3 dB footprintinterceptedexclusively theliquid,thusnotselectingtheburstswherethefootprints over-lapthecoastlines.Theselectionresultedin21footprintsandina mean value of specific attenuation K

=

13+₋5₆ dB/μs (1-sigma un-certainty)forPungaMare,and17footprintswithameanvalueof K

=

10+₋6₈ dB/μs forBaffin Sinus (Supplementary Fig. S3 and Ta-ble S1).

The estimated attenuation when converted into loss tangent givesavalueoftan

δ

=

3₋+1₁

×

10−5(1-sigmauncertainty)forPunga and3+₋1₂

×

10−5forBaffinSinus.

Assuming that the liquid is composed of a ternary mixture of methane, ethane, and nitrogen, we determined the composi-tion of the Punga Mare and Baffin Sinus using the component dielectric properties measured by Mitchell et al. (2015) and the same assumptions adopted as forLigeia Mare and OntarioLacus (Mastrogiuseppeetal.,2016,2018):

1) Ethane and methane are the most abundant materials on Titan that have loss tangents low enough to matchour inferred value.TheyareliquidunderTitanconditions.(Nosolidcompounds havetherequisitelowlosstangent.)

JID:EPSL AID:15086 /SCO [m5G; v1.236; Prn:23/05/2018; 14:31] P.4 (1-7) 4 M. Mastrogiuseppe et al. / Earth and Planetary Science Letters•••(••••)•••–•••

1 67 2 68 3 69 4 70 5 71 6 72 7 73 8 74 9 75 10 76 11 77 12 78 13 79 14 80 15 81 16 82 17 83 18 84 19 85 20 86 21 87 22 88 23 89 24 90 25 91 26 92 27 93 28 94 29 95 30 96 31 97 32 98 33 99 34 100 35 101 36 102 37 103 38 104 39 105 40 106 41 107 42 108 43 109 44 Q9110 45 111 46 112 47Q14 Q10113 48 114 49 115 50 Q11116 51 117 52 118 53 119 54 120 55 121 56 122 57 123 58 124 59 125 60 126 61 127 62 128 63 129 64 130 65 131 66 132

Fig. 2. (Upperpanels)SARmosaicshowingthelocationofthe−3dBbeam-limitedaltimeterfootprintsoverliquid(redcircles)andsolidsurfaces(bluecircles).Notethat thefootprintsbarelyoverlap.(Middlepanels)Super-resolvedradargram:Pungamare(left)andnearbyliquidbodies(right).(Bottompanels)Threewaveformsreceivedover PungaMare(left),BaffinSinus(middle)andDingleSinus(right):locationsofrelativefootprintsareindicatedbywhitearrows.

2) Nitrogen is a dissolved component whose molefraction in the liquid isdetermined by the partial pressure (1.5 bars)of ni-trogen in contact with the open seas, the solubility of nitrogen measuredinpureethaneandmethane(CheungandWang,?),and inthebinaryethane–methanesystem(Malaskaetal.,2017).

3)The2.18cmwavelengthmicrowaveabsorptioncoefficientof themixture isacombinationofits individualcomponentsas de-terminedbyMitchelletal. (2015) usingtheLorentz–Lorenzmixing rule,fornitrogenmixingratiosinmethane–ethanedeterminedby Malaskaetal. (2017) assumingatemperatureof91 K.Fora tem-peratureof95 Ktheethanemolefractionwillincreaseslightlyand thenitrogendecrease.

Wenotethatthebest-fitlosstangentforPungaMareandBaffin Sinusisclosetothe3

.

3

×

10−5_{,valuegiveninMitchelletal. (2015)}

for a purely binary mixture of methane and molecular nitrogen, withthelatter’smixingratiodeterminedbyitssolubilityinliquid methaneunderthe 1.4bar atmosphericpressure(see Malaska et al.,2017).Ourbest-fitcompositionfortheobservedlosstangentis thereforeabinarymixtureof80%CH4 and20%N2,withonlytrace

amounts of ethane and/or high-order components which would causehigherabsorption.Wenotethattheuncertaintiesoftheloss

tangentsofLigeiaMare(Mastrogiuseppeetal.,2018),PungaMare, andtheKrakenMare’sBaffinSinussignificantlyoverlap.

Consideringthe estimatedcomposition ofPungaMareand us-ing the Lorentz Lorenz formula (Born and Wolf, 1999), we de-rivedthe realpartofthe dielectriccomponentof themixtureas 1

.

67+₋0.05_0.05. This value is consistent withthe low emissivity mea-suredbytheradiometry(LeGalletal.,2016; Janssen etal.,2016) andisusedasinput togeneratethebathymetryofthemare and constraintheseafloorroughness.

ResultsoftheMonteCarloestimatesandrelative uncertainties arereportedinFig. 3,whichshowsthatthe maximumdepthfor Punga Mare measured along the track is about 110 m. The de-rivedalong-trackbathymetry isconsistent withthe observational evidenceobtainedfromSAR imagesofboth PungaMareand Baf-finSinus(Figs.1and2). Theseafloor ofPungaMare showsa50 meterelevationriseduetothepresenceofthecentralisland visi-blefromSARimages(seeFig.2,upper-leftpanel).Moreover,there isan overall agreement betweenthe derived seafloorprofile and thebathymetrythatwouldbequalitativelyinferredfromthe mor-phology(e.g. shapeofcoastal features,presenceofbays, distance fromshorelines) of the mare and the brightnessof the SAR im-ages,ifinterpretedasaproxyfortheseafloordepth.Basedonthe

JID:EPSL AID:15086 /SCO [m5G; v1.236; Prn:23/05/2018; 14:31] P.5 (1-7)

M. Mastrogiuseppe et al. / Earth and Planetary Science Letters•••(••••)•••–••• 5

1 67 2 68 3 69 4 70 5 71 6 72 7 73 8 74 9 75 10 76 11 77 12 78 13 79 14 80 15 81 16 82 17 83 18 84 19 85 20 86 21 87 22 88 23 89 24 90 25 91 26 92 27 93 28 94 29 95 30 96 31 97 32 98 33 99 34Q15 100 35 101 36 102 37 103 38 104 39 105 40 106 41 107 42 108 43 109 44 110 45 111 46 112 47 113 48 114 49 115 50 116 51 117 52 118 53 119 54 120 55 121 56 122 57 123 58 124 59 125 60 126 61 127 62 128 63 129 64 130 65 131 66 132

Fig. 3. Estimateddepths(upperpanels),surface-subsurfacerelativepower( Ps/ Pss)(middlepanels),andsubsurfaceroughness(lowerpanels)forPungaMare(left)andBaffin

Sinus(right).Errorbarsarerelativeto1-sigmauncertainty.

samequalitativeconsiderations,theshallowerbathymetryofNorth KrakenMareatBaffinSinusisalsoconsistentwiththemorphology andbrightnessofthisliquidbodyasobservedfromSARimages.

The depth of 110 m was measured at about20–30 km from thenearestshoreline.Thissuggestsaseabedprofilethatisrather steep, similar to that of the southern margin of Ligeia Mare as measuredonT91(Mastrogiuseppeetal.,2014a,2014b),andrather steeperthanthenorthernpartofthattransect(160mover

∼

150 km). We note similarly that our bathymetric transect is at the southernmarginofPungaMare.

4. Seafloorbackscattering:comparisonofPungaandLigeiaMare

In addition to the liquid composition, we use radar data for investigating dielectric properties of the seafloor. We compare seafloorbackscatteringofPungaMare, observedduring theT108, alongwiththebackscatteringofLigeiaMare,measuredusingT91 flyby data. This comparison is particularly reliable since geome-try and radar system parameters during the acquisition of the two observations were similar. We noteda remarkablesimilarity among echoes backscattered at the seafloor of the two seas, as shownwithanexampleinFig. 4. Radarwaveformsindicate rela-tivesmoothtopography(i.e.subsurfacepulseshapeapproachesto a Gaussian single peak) andhigh transparency of the liquid(i.e. intensityratioofsurfaceandseafloorechoesislowerthan30dB). Considering the estimated specific attenuation K values of 13 dB/μsand16 dB/μsforthePungaandLigeiamare,respectively, weuseformula(1) foraquantitativeinvestigationoftheseafloor compositionofthetwoliquidbodies.

Formula (1) is used to relate the Ps/ Pss measurements with

seafloorreflectivity R12,surfacereflectivity

Γ

s,andattenuationK ,

assumingthreepossible valuesforthescatteringterm fss,which

accountsforsubsurfaceroughnessscatteringeffectsandcanbe ex-pressed by rms slope. Such assumptions are necessary since the inversionproblem isillposed. Forourstudywe derive the scat-tering terms fss using the Geometrical Optics (GO) formulation

(Ulabyetal.,1982) alongwiththreepossiblevaluesofrmsslopes rangingfrom1◦to5◦.Notethatthermsslopevaluesadoptedhere

Fig. 4. Waveformsacquiredduringfly-byT108overPunga(upperpanel)andduring fly-byT91overLigeiaMare(lowerpanel).Echoesbackscatteredfromtheseafloor areindicativeofrelativesmooth topographyatthe resolutionscaleofthe radar (35 m)andhightransparencyoftheliquid.

arearbitraryandalthoughtheycouldrepresenttherealvalueson Titan,theyareonlyusedinthisstudyasreferencevaluesfor com-paringthedielectricpropertiesofthetwoseafloors.

Ps Pss





dB

= Γ

s

|

dB

− (

1

− Γ

s

)

2



_dB

−

R12

|

dB

+

K

|

dB

+

fs fss





dB (1) Foreachofthewaveformsacquiredoverthetwomaria, invert-ing formula(1), we calculateda value ofeffectivepermittivity of theseafloorfordifferentvaluesofrmsslope.The resultisshown in Fig. 5 where the permittivity of the two seafloors is plotted alongtheradartransectofT91andT108.WenotethatLigeiaMare andPungaMareseafloorsshowsimilarvaluesofeffective permit-tivity for similar values of small scale roughness. Moreover, the shape ofradar waveforms suggeststhey havesimilar seafloor to-pography.Ourconclusionisthattheseafloorsofboththesemaria are relatively smooth at the resolutionscale of theradar (35 m)

JID:EPSL AID:15086 /SCO [m5G; v1.236; Prn:23/05/2018; 14:31] P.6 (1-7) 6 M. Mastrogiuseppe et al. / Earth and Planetary Science Letters•••(••••)•••–•••

1 67 2 68 3 69 4 70 5 71 6 72 7 73 8 74 9 75 10 76 11 77 12 78 13 79 14 80 15 81 16 82 17 83 18 84 19 85 20 86 21 87 22 88 23 89 24 90 25 91 26 92 27 93 28 94 29 95 30 96 31 97 32 98 33 99 34 100 35 101 36 102 37 103 38 104 39 105 40 106 41 107 42 108 43 109 44 110 45 111 46 112 47 113 48 114 49 115 50 116 51 117 52 118 53 119 54 120 55 121 56 122 57 123 58 124 59 125 60 126 61 127 62 128 63 129 64 130 65 131 66 132

Fig. 5. ComparisonbetweentheseafloorreflectivityandtherelativedielectricconstantforPungaandLigeiaMareforthreepossiblevaluesofrmsslope.(Leftcolumn)Punga Mare.(Rightcolumn)LigeiaMare.

andconsideringthe adoptedvaluesforthemicroscaleroughness, the composition of the two sea floors is consistent with several possible materials present on Titan, such as solid organics, wa-tericeora mixtureof thesematerials.The roughnessofthe sea floors is smoother than that of terrain immediately surrounding themaria,consistentwithsedimentdepositionthroughoutthe ob-servedseafloor.

5. Summaryandimplications

WeanalyzedtheflybyT108 radaraltimetrydataacquiredover Punga Mare and Baffin Sinus. The detection of the seafloor al-lowed us to investigate the liquid composition, the bathymetry, andseafloortopographyoftheseliquidbodies.Wefoundthat the loss tangents of PungaMare andBaffin Sinusare similar to that of Ligeia Mare within the 1-sigma uncertainty of the estimates (tan

δ

=

4

.

4+₋0.9_0.9

×

10−5_{forLigeiaMare,3}+1

−1

×

10−5 forPungaMare

and 3+₋1₂

×

10−5 _{for Baffin Sinus. See Supplementary Table 1 for}

relative compositionvalues.However,the bestfitvaluesobtained heresuggestthatthecompositionforPungaMareandBaffinSinus mightbemostconsistentwithabinarymethane-nitrogensystem, withnitrogenapproachingitssolubilitylimitof20%molefraction inliquidmethaneandwithlittleornoethaneorhigherorder hy-drocarbons.

Several scenarios have been proposed which could permit ethane-poorseas.Mousisetal. (2016) andChoukrounetal. (2010) suggestedthat ethane mightbe sequesteredby clathration ifthe crust ofwatericeormethanehydratesisexposedattheseabed, resultinginmethane-dominatedliquids.However, thiseffectdoes nothaveanyapriorilatitudedependence.

The hydrological scenario proposed by Lorenz (2014) predicts an equator-to-polegradation inmethaneandnitrogenabundance intheseas,muchassalinityvaries,e.g.,betweentheBlackSeaand theMediterranean(or theBalticandNorthSea)on Earth.Higher inputsof‘fresh’liquid(onTitan,methane),eitherbyhigherrainfall rates and/or by largerland catchment areas,can displace solutes (salt,orethane)tootherbasinsinasystemofconnectedseas. Al-thoughthesimplemodelinLorenz (2014) didnotconsiderPunga specifically(atthetimeits hydraulicconnectiontotheotherseas was not obvious), the paradigm in that paper qualitatively sug-geststhatthenear-polarPungaMareshouldbemostethane-poor,

sincecirculationmodelssuggestmethaneprecipitationonTitan in-creases with latitude, and asthe smallest sea, Punga Mare may haveaproportionatelylargercatchmentarea,bothfactorsfavoring efficient flushing of solutes (the ethane and propane etc. which wouldpredicthigherlosstangents)intoKraken.

Alternatively, the latitudinal difference in composition could be explained as the results of the gradient in the surface tem-perature, assuming thermodynamic equilibrium with an atmo-sphere with constant methane mixing ratio (Tan et al., 2015; TanandKargel,2018).

Particularly, using models of cryogenic chemical systems, Tan etal. (2015) suggestedthatlatitudinal (e.g.two Kelvin difference intemperaturefrompoleto20degreesof latitude)andseasonal variationofsurfacetemperaturecoulddeterminevariationsin va-por density and equilibrium phase compositions on Titan, thus affecting the atmospheric dynamics as well as the global fluid circulation in the surface and upper crust. More specifically, the authorspredictmoreabundantbutlessdenseliquidinlargeseas at higher latitude. These methane-rich liquids will tend to flow toward lower, warmer latitudes, re-equilibrating on the way to the equator with the lower atmosphere through evaporation of methaneandthusaprogressiveenrichmentofethaneatthelower latitude.Finally,ethane-richer,denserliquidwouldsinktothe bot-tom of the sea and then flow poleward in a density-controlled cycleanalogoustothermohalinecirculationsystemsonEarth(Tan etal., 2015). However thisthermally-drivenlocal-thermodynamic equilibriumcirculationscenario, failsto predict thecompositions we have observed, but rather is too ethane-rich in the north-ern seas. These models also predict the presence of higher or-der components, such as propane, which wouldfurther increase absorptivity and likely make the models incompatible with ob-servedlosstangents. Basedontheseobservations,itappears that non-equilibriumprocesses are key to understanding Titan, ason Earth,wheretheatmospherichumidityisnot100%despitebeing alargelywater-coveredplanet.

Giventhat thelosstangentsobtainedsofar arethesameina formal(onesigma)sense,itmayinfactbechallengingtoproperly infer any compositional variability of seas until the composition anddepthofKrakenMare, by farthe largestliquidreservoir, are determined.

Tobetter constrain theuncertainties regarding thisvariability, neworbital missions equipped with multi-frequency radar

dedi-JID:EPSL AID:15086 /SCO [m5G; v1.236; Prn:23/05/2018; 14:31] P.7 (1-7)

M. Mastrogiuseppe et al. / Earth and Planetary Science Letters•••(••••)•••–••• 7

1 67 2 68 3Q12 69 4 70 5 71 6 72 7 73 8 Q1674 9 75 10 76 11 77 12 78 13 79 14 80 15 81 16 82 17 83 18 84 19 85 20 86 21 87 22 88 23 89 24 90 25 91 26 92 27 93 28 94 29 95 30 96 31 97 32 98 33 99 34 100 35 101 36 102 37 103 38 104 39 105 40 106 41 107 42 108 43 109 44 110 45 Q17111 46 112 47 113 48 114 49 115 50 116 51 Q18117 52 118 53 119 54 120 55 Q19121 56 122 57 123 58 124 59 125 60 126 61 127 62 128 63 129 64 130 65 131 66 132

catedto map composition ofTitan seas andlakes are envisaged. Additionally, direct measurements of composition and dielectric propertiesbyoneormorein-situprobes(LorenzandMann,2015) couldprovide groundtruthtorobustly infercomposition globally from radar data. Therefore, a combination of orbital data along within situdirectmeasurements would representthe most effi-cientexplorationeffort.

While this effort requires future missions (orbital or landed) to Titan, it is remarkablethat the Cassini mission (and its radar inparticular),althoughnotconceivedforsuchmeasurements,has beenableto obtaindepths andbulk liquidcompositionsoflakes andseaswithonlyahandfulofobservations.

Acknowledgements

M.M. and R.S. would like to acknowledge support from Ital-ianSpaceAgency(ASI)grant2014-041-R.0; G.D.A.was supported by the Italian Ministry ofUniversity andResearch through FIRB-RBFR130ICQ grant. M.M.,A.G.H., andV.P. would like to acknowl-edgesupportfromNASACDAPgrantNNX15AH10G;R.L. acknowl-edgesNASAgrantsNNX13AH14GandNNX13AK97G.J.I.L.is grate-fulfor theministrations of the Cassini missionin supporting his research.We appreciatetheeffortsoftheCassini TOST(Titan Or-biter Science Team) andRADAR Team in planning andexecuting these observations, in particular Yanhua Anderson and Richard West took great care to implement the spacecraft turns and in-strumentsettingstoachievethePungainspectionreportedhere.

Appendix A. Supplementarymaterial

Supplementarymaterialrelatedtothisarticlecanbefound

on-lineathttps://doi.org/10.1016/j.epsl.2018.05.033.

References

Alberti,G.,Festa,L.,Papa,C.,Vingione,G.,2009.Awaveformmodelfornear-Nadir radaraltimetryappliedtotheCassinimissiontoTitan.IEEETrans.Geosci. Re-moteSens. 47(7),2252–2261.https://doi.org/10.1109/TGRS.2009.2012718. Barnes,J.W.,Sotin,C.,Soderblom,J.M.,etal.,2014.Planet.Sci. 3(3).https://doi.org/

10.1186/s13535-014-0003-4.

Blackman,R.B., Tukey,J.W.,1959.Particularpairsofwindows.In: The Measure-mentofPowerSpectra,fromthePointofViewofCommunicationsEngineering. Dover,NewYork,pp. 98–99.

Birch,S.P.D., Hayes,A.G.,Dietrich, W.E.,Howard,A.D.,Bristow,C.S., Malaska,M.J., Lopes, R.M.C.,2017.Geomorphologic mappingofTitan’s polarterrains: con-straining surface processes and landscape evolution. Icarus 282, 214–236.

https://doi.org/10.1016/j.icarus.2016.08.003.

Born,M.,Wolf,E.,1999.PrinciplesofOptics:ElectromagneticTheoryof Propaga-tion,InterferenceandDiffractionofLight,7thed.CambridgeUniversityPress. ISBN 0-521-64222-1,section2.3.3.

Cuomo,K.M.,1992.ABandwidthExtrapolationTechniqueforImprovedRange Res-olutionofCoherentRadarData.ProjectReportCJP-60,Revision1,MITLincoln Laboratory.

Choukroun,M.,Grasset,O.,Tobie,G.,Sotin,C.,2010.Stabilityofmethaneclathrate hydratesunderpressure:influenceonoutgassingprocessesofmethaneon Ti-tan.Icarus 205(2),581–593.https://doi.org/10.1016/j.icarus.2009.08.011. Elachi,C.,etal.,2004.Radar:theCassiniTitanradarmapper.SpaceSci.Rev. 115(1),

71–110.Availablehttp://link.springer.com/article/10.1007%2Fs11214-004-1438 -9[Online].

Grima,C.,Mastrogiuseppe,M.,Hayes,A.G.,Wall,D.S.,Lorenz,R.D.,Hofgartner,J.D., Stiles,B.,Elachi,C.,2017.SurfaceroughnessofTitan’shydrocarbonseas.Earth Planet.Sci. Lett.(ISSN 0012-821X) 474,20–24. https://doi.org/10.1016/j.epsl. 2017.06.007.

Hayes,A.G.,2016.Thelakesand seasofTitan.Ann.Rev.EarthSpaceSci. 44(1), 57–83.

Hayes,A.G.,Aharonson,O.,Callahan,P.,Elachi,C.,Gim,Y.,Kirk,R.,Lewis,K.,Lopes, R.,Lorenz,R.,Lunine,J., Mitchell,K.,Mitri, G.,Stofan,E.,Wall, S.,2008. Hy-drocarbonlakesonTitan:distributionandinteractionwithaporousregolith. Geophys.Res.Lett. 35,9204.

Hayes,A.G.,Lorenz,R.D.,Donelan,M.A.,Manga,M.,Lunine,J.I.,Schneider,T.,Lamb, M.P., Mitchell, J.M.,Fischer, W.W., Graves, S.D.,Tolman, H.L.,Aharonson, O., Encrenaz, P.J., Ventura, B., Casarano, D., Notarnicola, C., 2013. Wind driven capillary-gravitywavesonTitan’slakes:hardtodetectornon-existent?Icarus (ISSN 0019-1035)225(1),403–412.https://doi.org/10.1016/j.icarus.2013.04.004. Hayes,A.G.,Birch,S.P.D.,Dietrich,W.E.,Howard,A.D.,Kirk,R.L.,Poggiali,V., Mas-trogiuseppe, M., Michaelides, R.J., Corlies, P.M., Moore, J.M., Malaska, M.J., Mitchell,K.L.,Lorenz,R.D.,Wood,C.A.,2017.Topographicconstraintsonthe evo-lutionandconnectivityofTitan’slacustrinebasins.Geophys.Res.Lett. Iess,L.,etal., 2012.ThetidesofTitan.Science 337, 457.https://doi.org/10.1126/

science.1219631.

Janssen,M.A.,LeGall,A.,Lopes,R.M.,Lorenz,R.D.,Malaska,M.J.,Hayes,A.G.,Neish, C.D.,Solomonidou,A.,Mitchell,K.L.,Radebaugh,J.,Keihm,S.J.,Choukroun,M., Leyrat,C., Encrenaz,P.J.,Mastrogiuseppe,M.,2016.Titan’ssurfaceat 2.18-cm wavelengthimagedbytheCassiniRADARradiometer:resultsand interpreta-tionsthroughthefirsttenyearsofobservation.Icarus(ISSN 0019-1035)270, 443–459.https://doi.org/10.1016/j.icarus.2015.09.027.

LeGall,A.,etal., 2016.Composition,seasonal change,and bathymetryofLigeia Mare, Titan,derived fromitsmicrowave thermalemission. J. Geophys. Res., Planets 121,233–251.https://doi.org/10.1002/2015JE004920.

Lopes,R.M.C.,etal.,2013.CryovolcanismonTitan:newresultsfromCassiniRADAR andVIMS.J.Geophys.Res.,Planets 118,416–435.https://doi.org/10.1002/jgre. 20062.

Lorenz,R.D.,2014.TheflushingofLigeia:compositionvariationsacrossTitan’sseas inasimplehydrologicalmodel.Geophys.Res.Lett. 41,5764–5770.https://doi. org/10.1002/2014GL061133.

Lorenz,R.D.,Mann,J.,2015.SeakeepingonLigeiaMare:dynamicresponseofa float-ingcapsuletowavesonthe hydrocarbonseasofSaturn’sMoonTitan.Johns HopkinsAPLTech.Dig. 33(2),82–94.

Malaska, M.J., Hodyss, R., Lunine, J.I.,Hayes, A.H., Hofgartner,J.D.,Hollyday, G., Lorenz, R.D., 2017. Laboratory measurements ofnitrogen dissolution in Ti-tanlakefluids.Icarus(ISSN 0019-1035)289,94–105.https://doi.org/10.1016/ j.icarus.2017.01.033.

Mastrogiuseppe, M., et al.,2014a. Thebathymetry ofaTitan sea. Geophys.Res. Lett. 41,1432–1437.https://doi.org/10.1002/2013GL058618.

Mastrogiuseppe, M., et al., 2014b. Titan duneheights retrieval byusingCassini radaraltimeter.Icarus(ISSN 0019-1035)230,191–197.https://doi.org/10.1016/ j.icarus.2013.09.028.

Mastrogiuseppe,M.,etal.,2016.RadarsoundingusingtheCassinialtimeter: wave-formmodelingandMonteCarloapproachfordatainversionofobservationsof Titan’sseas.IEEETrans.Geosci.RemoteSens. 54(10).https://doi.org/10.1109/ TGRS.2016.2563426.

Mastrogiuseppe, M., etal., 2018. Bathymetryand compositionofTitan’sOntario Lacusderived fromMonteCarlo-basedwaveforminversionofCassiniRADAR altimetrydata.Icarus(ISSN 0019-1035)300,203–209.https://doi.org/10.1016/j. icarus.2017.09.009.

Mitchell,K.L., Barmatz,M.B., Jamieson, C.S., Lorenz,R.D., Lunine, J.I.,2015. Lab-oratorymeasurementsofcryogenicliquidalkanemicrowaveabsorptivityand implicationsforthecompositionofLigeiaMare,Titan.Geophys.Res.Lett. 42, 1340–1345.https://doi.org/10.1002/2014GL059475.

Mitri,G.,Bland,M.T.,Showman,A.P.,Radebaugh,J.,Stiles,B.,Lopes,R.M.C.,Lunine, J.I.,Pappalardo,R.T.,2010.MountainsonTitan:modelingandobservations.J. Geophys.Res. 115,E10002.https://doi.org/10.1029/2010JE003592.

Mousis,O.,Lunine,J.I.,Hayes,A.G.,Hofgartner,J.D.,2016.Thefateofethanein Ti-tan’shydrocarbonlakesandseas.Icarus.

Poggiali, V., et al., 2016. Liquid-filled canyons on Titan.Geophys. Res. Lett. 43, 7887–7894.https://doi.org/10.1002/2016GL069679.

Radebaugh,J.,et al.,2007.Mountainson TitanobservedbyCassiniRadar.Icarus (ISSN 0019-1035)192(1),77–91.https://doi.org/10.1016/j.icarus.2007.06.020. Stofan, E.R.,2007. The lakesofTitan. Nature 445,61–64.Available http://www.

nature.com/nature/journal/v445/n7123/abs/nature05438.html[Online]. Tan,S.P.,Jeffrey,S.K.,Donald,E.J.,Mastrogiuseppe,M.,Hertanto,A.,Marion,G.M.,

2015. Titan’sliquids:exoticbehavioranditsimplicationsonglobalfluid cir-culation.Icarus (ISSN 0019-1035) 250,64–75.https://doi.org/10.1016/j.icarus. 2014.11.029.

Tan,S.P.,Kargel,J.S.,2018.Multiphase-equilibriaanalysis:applicationinmodeling theatmosphericandlacustrinechemicalsystemsofSaturn’smoonTitan.Fluid Phase Equilib.(ISSN 0378-3812) 458,153–169. https://doi.org/10.1016/j.fluid. 2017.11.020.

Ulaby, F.T., Moore, R.K., Fung, A.K., 1982. Microwave Remote Sensing, vol. 2. Addison–Wesley,Reading,MA(Chapters11and12).

West,R.D.,et al.,2009.CassiniRADARsequenceplanningandinstrument perfor-mance.IEEETrans.Geosci.RemoteSens. 47(6),1777–1795.https://doi.org/10. 1109/TGRS.2008.2007217.

Wood,C.A.,etal.,2010.ImpactcratersonTitan.Icarus(ISSN 0019-1035)206(1), 334–344.https://doi.org/10.1016/j.icarus.2009.08.021.

JID:EPSL AID:15086 /SCO [m5G; v1.236; Prn:23/05/2018; 14:31] P.8 (1-7) 1 67 2 68 3 69 4 70 5 71 6 72 7 73 8 74 9 75 10 76 11 77 12 78 13 79 14 80 15 81 16 82 17 83 18 84 19 85 20 86 21 87 22 88 23 89 24 90 25 91 26 92 27 93 28 94 29 95 30 96 31 97 32 98 33 99 34 100 35 101 36 102 37 103 38 104 39 105 40 106 41 107 42 108 43 109 44 110 45 111 46 112 47 113 48 114 49 115 50 116 51 117 52 118 53 119 54 120 55 121 56 122 57 123 58 124 59 125 60 126 61 127 62 128 63 129 64 130 65 131 66 132 XMLVIEW:extended

Appendix A. Supplementarymaterial

ThefollowingistheSupplementarymaterialrelatedtothisarticle.

begin ecomponent begin ecomponent begin ecomponent begin ecomponent begin ecomponent begin ecomponent begin ecomponent begin ecomponent begin ecomponent begin ecomponent begin ecomponent begin ecomponent begin ecomponent begin ecomponent begin ecomponent begin ecomponent begin ecomponent

Label:

MMC

caption:

T108CassiniRADARobservation:GeometryandSystemparameters.

link:

APPLICATION : mmc1

JID:EPSL AID:15086 /SCO [m5G; v1.236; Prn:23/05/2018; 14:31] P.9 (1-7) 1 67 2 68 3 69 4Q20 70 5 71 6 72 7 73 8 74 9 75 10 76 11 77 12 78 13 79 14 80 15 81 16 82 17 83 18 84 19 85 20 86 21 87 22 88 23 89 24 90 25 91 26 92 27 93 28 94 29 95 30 96 31 97 32 98 33 99 34 100 35 101 36 102 37 103 38 104 39 105 40 106 41 107 42 108 43 109 44 110 45 111 46 112 47 113 48 114 49 115 50 116 51 117 52 118 53 119 54 120 55 121 56 122 57 123 58 124 59 125 60 126 61 127 62 128 63 129 64 130 65 131 66 132

Sponsor names

Donotcorrectthispage.Pleasemarkcorrectionstosponsornamesandgrantnumbersinthemaintext.

ItalianSpaceAgency,country=Italy,grants=

2014-041-R.0

NASA,country=UnitedStates,grants=

NNX15AH10G

JID:EPSL AID:15086 /SCO [m5G; v1.236; Prn:23/05/2018; 14:31] P.10 (1-7) 1 67 2 68 3 69 4 70 5 71 6 72 7 73 8 74 9 75 10 76 11 77 12 78 13 79 14 80 15 81 16 82 17 83 18 84 19 85 20 86 21 87 22 88 23 89 24 90 25 91 26 92 27 93 28 94 29 95 30 96 31 97 32 98 33 99 34 100 35 101 36 102 37 103 38 104 39 105 40 106 41 107 42 108 43 109 44 110 45 111 46 112 47 113 48 114 49 115 50 116 51 117 52 118 53 119 54 120 55 121 56 122 57 123 58 124 59 125 60 126 61 127 62 128 63 129 64 130 65 131 66 132

Highlights

•

FirstbathymetricmeasurementsofPungamare,Titan’sthird-largestsea.

•

Maximumdepthmeasuredalongthetrackisabout110m.

•

Compositionisconsistentwith80%ofmethaneand20%ofnitrogen.

•

Littletonoethaneispresentonthemare.