HighResolutionALMADataof theFomalhautDebrisDiskConfirmsApsidalWidth Variation
sacani
315 views
10 slides
Sep 10, 2025
Slide 1 of 10
1
2
3
4
5
6
7
8
9
10
About This Presentation
Wepresent long-baselineobservationsof theFomalhautouterdebrisdiskat223GHz(1.3mm)fromAtacama LargeMillimeter/submillimeterArray (ALMA) Cycle 5, whichwe use alongwith archival short-baseline observationstoproducea0.57resolutionmosaicofthediskatasensitivityof7μJybm−1.Weuseradialprofiles tomeasurethe...
Slide Content
HighResolutionALMADataoftheFomalhautDebrisDiskConfirmsApsidalWidth
Variation
JayS.Chittidi
1,2
aa ,MeredithA.MacGregor
1
aa ,JoshuaBennettLovell
3
aa ,GaspardDuchene
4,5
aa ,MarkWyatt
6
aa ,OljaPanic
7
aa ,
PaulKalas
4,8,9
aa ,MargaretPan
3
,A.MeredithHughes
10
aa ,DavidJ.Wilner
3
aa ,GrantM.Kennedy
11
aa ,LucaMatrà
12
aa ,
MichaelP.Fitzgerald
13
aa ,andKateY.L.Su
14
aa
1
DepartmentofPhysics&Astronomy,JohnHopkinsUniversity,3400NorthCharlesStreet,Baltimore,MD21218,USA
2
DepartmentofAstrophysicalandPlanetarySciences,UniversityofColorado,2000ColoradoAvenue,Boulder,CO80309,USA
3
CenterforAstrophysics,Harvard&Smithsonian,60GardenStreet,Cambridge,MA02138-1516,USA
4
AstronomyDepartment,UniversityofCalifornia,Berkeley,CA94720-3411,USA
5
UniversitéGrenobleAlpes/CNRS, InstitutdePlanétologieetd’AstrophysiquedeGrenoble,38000Grenoble,France
6
InstituteofAstronomy,UniversityofCambridge,MadingleyRoad,CambridgeCB30HA,UK
7
SchoolofPhysicsandAstronomy,UniversityofLeeds,LeedsLS29JT,UK
8
InstituteofAstrophysics,FORTH,GR-71110Heraklion,Greece
9
CarlSaganCenter,SETIInstitute,189BernardoAvenue,MountainView,CA94043,USA
10
DepartmentofAstronomy,VanVleckObservatory,WesleyanUniversity,Middletown,CT06459,USA
11
DepartmentofPhysicsandCentreforExoplanetsandHabitability,UniversityofWarwick,GibbetHillRoad,CoventryCV47AL,UK
12
SchoolofPhysics,TrinityCollegeDublin,TheUniversityofDublin,CollegeGreen,Dublin2,Ireland
13
DepartmentofPhysicsandAstronomy,UCLA,LosAngeles,CA90095,USA
14
SpaceScienceInstitute,4750WalnutStreet,Suite205,Boulder,CO80301,USA
Received2025March19;revised2025July10;accepted2025July11;published2025September4
Abstract
Wepresentlong-baselineobservationsoftheFomalhautouterdebrisdiskat223GHz(1.3mm)fromAtacama
LargeMillimeter/submillimeter Array(ALMA) Cycle5,whichweusealongwitharchivalshort-baseline
observationstoproducea0.
57resolutionmosaicofthediskatasensitivityof7μJybm
−1
.Weuseradialprofiles
tomeasurethediskattheansaeandfindthatthesoutheast(SE)sideofthediskis4auwiderthanthenorthwest
(NW) sideasobservedbyALMA.WealsofindthatthepeakbrightnessoftheNWansais21%±1%brighter
thantheSEansa.WeperformMarkovChainMonteCarlofitsoftheALMAvisibilitiesusingtwoanalytical,
eccentricdiskmodels.Ourresultssuggestthatthemodelincludingadispersionparameterfortheproper
eccentricity(e
p
),whichaccountsforadditionalscatterintheeccentricityofindividualorbits,ispreferredoverthe
modelwithoutone.Suchamodelimpliesthatself-gravitation,particlecollisions,andclose-packingcouldplaya
roleinshapingtheoverallstructureoftheFomalhautdisk,asisseenineccentricplanetaryrings.Crucially,
neithermodelcanreproducethebrightnessorwidthasymmetryneartheNWansa.Noemissionfromthe
intermediatebeltisdetected,allowingustoplacea3σupperlimitof396μJyat1.3mm.Wealsodiscovera
spectrallineinarchivalCycle3datacenteredatν
obs
≈230.25GHzatthelocationofthe“GreatDustCloud,”
whoseredshiftfromtheexpectedCOlineforFomalhautconfirmsthesourceisabackgroundgalaxy.
UnifiedAstronomyThesaurusconcepts:Debrisdisks(363); Planetformation(1241)
1.Introduction
Fomalhautisanearby(∼7.7pc,F.vanLeeuwen2007) A3V
starwithanestimatedageof440Myr(E.E.Mamajek2012).
ObservationsfromtheInfraredAstronomicalSatellitedecades
agorevealedastronginfraredexcessindicativeofcircum-
stellardust,makingthesystemanexcitingsitetostudy
planetarydynamicsanddiskevolution(D.E.Backman&
F.Paresce1993). Low-resolutionimagingobservationswith
theSubmillimetreCommon-UserBolometerArrayatthe
JamesClerkMaxwellTelescoperevealedasymmetricemis-
sionfromthedebrisdiskatthesouthernansa(W.S.Holland
etal.1998). M.C.Wyattetal.(1999) notedthatthiscouldbe
evidenceofaneccentricdisksculptedbyaneccentric,hidden
planet,resultinginincreasedthermalemissionnearthedisk’s
pericenter(presumablythesouthernansa). Spitzerobserva-
tionspresentedinK.R.Stapelfeldtetal.(2004) showed
furtherevidenceofthis“pericenterglow.”HubbleSpace
Telescopeimagingresolvedtheoutercoldbelt(∼140au,
P.Kalasetal.2005) andlatertheappearanceofwhatwas
initiallythoughttobeaplanetbutisnowthoughttobean
expandingdustcloudjustinteriortothedisk(P.Kalasetal.
2008; S.J.Kenyonetal.2014; S.M.Lawleretal.2015;
A.Gáspár&G.Rieke2020). TheopticalimagingwithHubble
revealedthatthediskwaseccentric,andbothHSTandfar-
infraredimagingwithHerschelobservedpericenterglow
(B.Ackeetal.2012). Notably,longer-wavelengthimaging
usingfacilitiessuchastheAtacamaLargeMillimeter/
submillimeterArray(ALMA) hasshownapronounced
brightnessenhancementatthenorthernansa,analogously
called“apocenterglow”(M.A.MacGregoretal.2017).
M.Panetal.(2016) arguethatapocenterglowisduetoa
surfacedensityenhancementfromparticlestravelingslowest
(andhencespendingmoretime) atapocenter,whilerecent
theoreticalworkdonebyE.M.Lynch&J.B.Lovell(2022)
arguethatsuchfeaturesalsodependontheeccentricityprofile
ofthedisk.JWSTMIRIimagingpresentedinA.Gáspáretal.
(2023) furtherrevealedthepresenceofaninnerdiskofwarm
TheAstrophysicalJournalLetters,990:L40(10pp), 2025September10 https://doi.org/10.3847/2041-8213/adfadb
©2025.TheAuthor(s). PublishedbytheAmericanAstronomicalSociety.
aaaaaaa
Originalcontentfromthisworkmaybeusedundertheterms
oftheCreativeCommonsAttribution4.0licence. Anyfurther
distributionofthisworkmustmaintainattributiontotheauthor(s) andthetitle
ofthework,journalcitationandDOI.
1
Variation
JayS.Chittidi
1,2
aa ,MeredithA.MacGregor
1
aa ,JoshuaBennettLovell
3
aa ,GaspardDuchene
4,5
aa ,MarkWyatt
6
aa ,OljaPanic
7
aa ,
PaulKalas
4,8,9
aa ,MargaretPan
3
,A.MeredithHughes
10
aa ,DavidJ.Wilner
3
aa ,GrantM.Kennedy
11
aa ,LucaMatrà
12
aa ,
MichaelP.Fitzgerald
13
aa ,andKateY.L.Su
14
aa
1
DepartmentofPhysics&Astronomy,JohnHopkinsUniversity,3400NorthCharlesStreet,Baltimore,MD21218,USA
2
DepartmentofAstrophysicalandPlanetarySciences,UniversityofColorado,2000ColoradoAvenue,Boulder,CO80309,USA
3
CenterforAstrophysics,Harvard&Smithsonian,60GardenStreet,Cambridge,MA02138-1516,USA
4
AstronomyDepartment,UniversityofCalifornia,Berkeley,CA94720-3411,USA
5
UniversitéGrenobleAlpes/CNRS, InstitutdePlanétologieetd’AstrophysiquedeGrenoble,38000Grenoble,France
6
InstituteofAstronomy,UniversityofCambridge,MadingleyRoad,CambridgeCB30HA,UK
7
SchoolofPhysicsandAstronomy,UniversityofLeeds,LeedsLS29JT,UK
8
InstituteofAstrophysics,FORTH,GR-71110Heraklion,Greece
9
CarlSaganCenter,SETIInstitute,189BernardoAvenue,MountainView,CA94043,USA
10
DepartmentofAstronomy,VanVleckObservatory,WesleyanUniversity,Middletown,CT06459,USA
11
DepartmentofPhysicsandCentreforExoplanetsandHabitability,UniversityofWarwick,GibbetHillRoad,CoventryCV47AL,UK
12
SchoolofPhysics,TrinityCollegeDublin,TheUniversityofDublin,CollegeGreen,Dublin2,Ireland
13
DepartmentofPhysicsandAstronomy,UCLA,LosAngeles,CA90095,USA
14
SpaceScienceInstitute,4750WalnutStreet,Suite205,Boulder,CO80301,USA
Received2025March19;revised2025July10;accepted2025July11;published2025September4
Abstract
Wepresentlong-baselineobservationsoftheFomalhautouterdebrisdiskat223GHz(1.3mm)fromAtacama
LargeMillimeter/submillimeter Array(ALMA) Cycle5,whichweusealongwitharchivalshort-baseline
observationstoproducea0.
57resolutionmosaicofthediskatasensitivityof7μJybm
−1
.Weuseradialprofiles
tomeasurethediskattheansaeandfindthatthesoutheast(SE)sideofthediskis4auwiderthanthenorthwest
(NW) sideasobservedbyALMA.WealsofindthatthepeakbrightnessoftheNWansais21%±1%brighter
thantheSEansa.WeperformMarkovChainMonteCarlofitsoftheALMAvisibilitiesusingtwoanalytical,
eccentricdiskmodels.Ourresultssuggestthatthemodelincludingadispersionparameterfortheproper
eccentricity(e
p
),whichaccountsforadditionalscatterintheeccentricityofindividualorbits,ispreferredoverthe
modelwithoutone.Suchamodelimpliesthatself-gravitation,particlecollisions,andclose-packingcouldplaya
roleinshapingtheoverallstructureoftheFomalhautdisk,asisseenineccentricplanetaryrings.Crucially,
neithermodelcanreproducethebrightnessorwidthasymmetryneartheNWansa.Noemissionfromthe
intermediatebeltisdetected,allowingustoplacea3σupperlimitof396μJyat1.3mm.Wealsodiscovera
spectrallineinarchivalCycle3datacenteredatν
obs
≈230.25GHzatthelocationofthe“GreatDustCloud,”
whoseredshiftfromtheexpectedCOlineforFomalhautconfirmsthesourceisabackgroundgalaxy.
UnifiedAstronomyThesaurusconcepts:Debrisdisks(363); Planetformation(1241)
1.Introduction
Fomalhautisanearby(∼7.7pc,F.vanLeeuwen2007) A3V
starwithanestimatedageof440Myr(E.E.Mamajek2012).
ObservationsfromtheInfraredAstronomicalSatellitedecades
agorevealedastronginfraredexcessindicativeofcircum-
stellardust,makingthesystemanexcitingsitetostudy
planetarydynamicsanddiskevolution(D.E.Backman&
F.Paresce1993). Low-resolutionimagingobservationswith
theSubmillimetreCommon-UserBolometerArrayatthe
JamesClerkMaxwellTelescoperevealedasymmetricemis-
sionfromthedebrisdiskatthesouthernansa(W.S.Holland
etal.1998). M.C.Wyattetal.(1999) notedthatthiscouldbe
evidenceofaneccentricdisksculptedbyaneccentric,hidden
planet,resultinginincreasedthermalemissionnearthedisk’s
pericenter(presumablythesouthernansa). Spitzerobserva-
tionspresentedinK.R.Stapelfeldtetal.(2004) showed
furtherevidenceofthis“pericenterglow.”HubbleSpace
Telescopeimagingresolvedtheoutercoldbelt(∼140au,
P.Kalasetal.2005) andlatertheappearanceofwhatwas
initiallythoughttobeaplanetbutisnowthoughttobean
expandingdustcloudjustinteriortothedisk(P.Kalasetal.
2008; S.J.Kenyonetal.2014; S.M.Lawleretal.2015;
A.Gáspár&G.Rieke2020). TheopticalimagingwithHubble
revealedthatthediskwaseccentric,andbothHSTandfar-
infraredimagingwithHerschelobservedpericenterglow
(B.Ackeetal.2012). Notably,longer-wavelengthimaging
usingfacilitiessuchastheAtacamaLargeMillimeter/
submillimeterArray(ALMA) hasshownapronounced
brightnessenhancementatthenorthernansa,analogously
called“apocenterglow”(M.A.MacGregoretal.2017).
M.Panetal.(2016) arguethatapocenterglowisduetoa
surfacedensityenhancementfromparticlestravelingslowest
(andhencespendingmoretime) atapocenter,whilerecent
theoreticalworkdonebyE.M.Lynch&J.B.Lovell(2022)
arguethatsuchfeaturesalsodependontheeccentricityprofile
ofthedisk.JWSTMIRIimagingpresentedinA.Gáspáretal.
(2023) furtherrevealedthepresenceofaninnerdiskofwarm
TheAstrophysicalJournalLetters,990:L40(10pp), 2025September10 https://doi.org/10.3847/2041-8213/adfadb
©2025.TheAuthor(s). PublishedbytheAmericanAstronomicalSociety.
aaaaaaa
Originalcontentfromthisworkmaybeusedundertheterms
oftheCreativeCommonsAttribution4.0licence. Anyfurther
distributionofthisworkmustmaintainattributiontotheauthor(s) andthetitle
ofthework,journalcitationandDOI.
1
grains,apreviouslyundiscoveredintermediatebelt(IB)at
∼90aujustinteriortothemainbelt,andanextendedhaloof
dustfromtheouterbeltalsopresentinHSTimaging(P.Kalas
etal.2013). Recently,M.Sommeretal.(2025) presentedan
analysisthatdemonstratedthattheemissionfromtheinnerand
possiblyIBscouldbethenaturalconsequencesofPoynting–
Robertsondragactingonsmallerdustgrainsfromtheouter
belt.Later,JWSTNIRCAMimagingpresentedinM.Ygouf
etal.(2024) placedconstraintsonplanetmasses�M
J
beyond
about8au.Theyfoundonepossiblenewcandidatesource
(“S7”) thatfutureobservationswillneedtosearchfortoverify
ifitisassociatedwithFomalhaut.Thepresenceofmultiple
eccentricdebrisbeltsandtheirmultiwavelengthbrightness
distributionsprovidesanexcitingbutcomplexsystemtostudy
theprocessesthatdrivetheevolutionofplanetarysystems.
ALMAhasrevolutionizedourunderstandingofplane-
tesimaldisksmakingmanydozensofdetectionsofKuiper
BeltanalogsincludingtheFomalhautouterbeltitself
(A.G.Sepulvedaetal.2019; S.Marino2019, 2022).
A.C.Boleyetal.(2012) observedthediskinALMACycle
0andfoundthatthedisk’sedgesweresharperthan
expectationsforthesystem,andappearedconsistentwith
sculptingfromaninnerandpossiblyouterplanetarymass
companion.M.A.MacGregoretal.(2017) usedALMAto
produceamosaicmapofthediskinBand6at223GHz,
providinghighresolution(naturalweightbeamsizeof×1.561.15
)andsensitivity(rmsof14μJybeam
−1
)inthe
submillimeterregimeforthedisk,andshowedthatamodel
whichtreatsthecomplexeccentricityasthevectorsumwitha
properandforcedcomponentwithindependentphasepara-
meterswasagoodmatchtotheobservations.G.M.Kennedy
(2020) foundthatthosesameobservationsshowedevidence
thatthenorthwest(NW) ansaofthediskwasnarrowerthanthe
southeast(SE)ansabyabout4au(measuredfromthefull
widthathalf-maximum;FWHM), anddemonstratedthata
modifiedversionofthecomplexeccentricitymodelthat
includedadispersionintheforcedeccentricitycouldallowfor
anarrowerNWansa(proximatetothediskapocenter).
Here,wepresentnewALMAobservationsthatrevealthe
Fomalhautdebrisdiskatunprecedentedresolutionatmilli-
meterwavelengthsallowingustomoreaccuratelyconstrain
thegeometryoftheouterbelt.Thisletterisorganizedas
follows.InSection2,wediscusstheALMAdatasets,their
processing,andourmethodtocorrectforthepropermotion
betweenthethreeepochs.InSection3,wepresentcontinuum
imagesofthealigneddatasetsanddiscusstheCLEAN
parametersthatwereusedtoproducethem.Wethenpresent
MarkovChainMonteCarlo(MCMC) fitresultsfortwo
eccentricdiskmodelsinSection4.Wemeasureanddiscuss
thewidthsofthediskattheansaeusingradialprofilesofthe
diskimagesthathavebeenregriddedtoR–θspacein
Section5.1.Specifically,wefindthatwhilewecanstill
modelthebulkparametersofFomalhaut’seccentricring,our
modelsdonotprovideagoodphysicalinterpretationforthe
disk’sasymmetriesidentifiedinthehigher-resolutiondata
(i.e.,thewidthandbrightnessdifferencesattheansae). We
alsopresentthedetectionofaspectrallinecoincidentwiththe
“GreatDustCloud”(GDC) sourcediscussedinA.Gáspár
etal.(2023) andG.M.Kennedyetal.(2023), which
corroboratesitsnatureasabackgroundgalaxyratherthanan
objectassociatedwiththeFomalhautsystem.Finally,we
discussourinterpretationinthecontextofalternative
descriptionsofthissystem,suchasthosewithdifferencesin
theirradialprofileparameterizationsandorbitaleccentricity
distributionsoforbitaleccentricities,includingthatof
J.B.Lovelletal.(2025) analyzingthesesameobservations.
2.Data
ThreeepochsofALMAdatainBand6wereobtained
fromthearchive:ahigh-resolutionpointingonthecentral
starobservedinCycle2(J.A.Whiteetal.2017,
ID#2013.1.00486.S), aseven-pointingmosaicfromCycle3
(M.A.MacGregoretal.2017, ID#2015.1.00966.S), andtwo
high-resolutionpointingsatthediskapsesfromCycle5
(ID#2017.1.01043.S). TheCycle5observationsaresummar-
izedinTable1.TheobservationsfromCycle2werecalibrated
usingALMApipelineversion4.3.1,theCycle3datawere
processedwith4.5.3,andtheCycle5datawithversion5.1.1-
5,allofwhicharethepipelineversionsrecommendedby
ALMAsupportforeachrespectivedataset.Theflux
calibrationanddataweightswereinspectedacrossthethree
datasetsandwerefoundtobeconsistentacrossthethree
epochs,giventheirrespectivebaselinecoverage,andsono
furtherchangesweremadetoeitherproduct.Weusedtwo
versionsofCASAforthiswork:version6.5.1-23wasusedfor
alldatamanipulationtaskssuchasaveragingandvisibility
modelsubtraction,whileversion6.6.0-20wasusedexclu-
sivelyforimagingduetorelevantupdatestotclean.
ThedatawerethenconcatenatedintoasingleCASA
measurementset.Toreducethedatavolume,themeasurement
setwasaverageddownto8channelsforthethreecontinuum
spectralwindowsand128channelsforthespectralwindow
centeredonthe230.538GHzCOline.Wethentimeaveraged
thedatato30sintervals.
Self-calibrationwasattemptedtocorrectforphase
offsetsintroducedfromthestar’spropermotion(328.95,
−164.67masyr
−1
,comparabletothenetbeamsizeacrossthe
threeepochs,F.vanLeeuwen2007), butwasunsuccessfuldue
tolowsignal-to-noiseratio.Instead,weusedtheCASA
implementationofuvmodelfit(I.Martí-Vidaletal.2014)
tofitforthepositionofthestarforeachuniqueobservation
andpointing(atotalof52fits).Wethenusedthelast
observationfromtheCycle3dataasthereferencepositionof
thestar,α=22
h
57
m
39.450801andδ=−29°37′22.69400 ,
andusedthefixplanetsandfixvisfunctionsto
manuallycorrectthephasecenterandUVWbaselinepositions
tothereferencestellarposition.Tovalidatethepropermotion
correction,weimagedeachofthe52separatepointings/
observationsandanalyzedthestellarpositionintheimagesfor
eachsetofpointings.Wefoundthatthestellarcentroid
positionwasprecisetotwopixels(0.1
,orabout1/5ofthe
synthesizedbeamoftheimageinSection3).Thescatteris
dominatedbythestarcoincidingwiththefirstnullinthe
Table1
ALMAObservationsfromProject2017.1.01043.S
Date Antennas Baselines PWV Obs.Time
(m) (mm) (minutes)
2018Sep8 45 15.1–783.5 1.3 73.1
45 15.1–783.5 1.2 72.7
2018Sep17 45 15.1–1245.6 0.4 78.4
2018Sep23 48 15.1–1397.8 0.5 73.1
2018Sep26 47 15.1–1397.8 0.8 73.7
2
TheAstrophysicalJournalLetters,990:L40(10pp), 2025September10 Chittidietal.
∼90aujustinteriortothemainbelt,andanextendedhaloof
dustfromtheouterbeltalsopresentinHSTimaging(P.Kalas
etal.2013). Recently,M.Sommeretal.(2025) presentedan
analysisthatdemonstratedthattheemissionfromtheinnerand
possiblyIBscouldbethenaturalconsequencesofPoynting–
Robertsondragactingonsmallerdustgrainsfromtheouter
belt.Later,JWSTNIRCAMimagingpresentedinM.Ygouf
etal.(2024) placedconstraintsonplanetmasses�M
J
beyond
about8au.Theyfoundonepossiblenewcandidatesource
(“S7”) thatfutureobservationswillneedtosearchfortoverify
ifitisassociatedwithFomalhaut.Thepresenceofmultiple
eccentricdebrisbeltsandtheirmultiwavelengthbrightness
distributionsprovidesanexcitingbutcomplexsystemtostudy
theprocessesthatdrivetheevolutionofplanetarysystems.
ALMAhasrevolutionizedourunderstandingofplane-
tesimaldisksmakingmanydozensofdetectionsofKuiper
BeltanalogsincludingtheFomalhautouterbeltitself
(A.G.Sepulvedaetal.2019; S.Marino2019, 2022).
A.C.Boleyetal.(2012) observedthediskinALMACycle
0andfoundthatthedisk’sedgesweresharperthan
expectationsforthesystem,andappearedconsistentwith
sculptingfromaninnerandpossiblyouterplanetarymass
companion.M.A.MacGregoretal.(2017) usedALMAto
produceamosaicmapofthediskinBand6at223GHz,
providinghighresolution(naturalweightbeamsizeof×1.561.15
)andsensitivity(rmsof14μJybeam
−1
)inthe
submillimeterregimeforthedisk,andshowedthatamodel
whichtreatsthecomplexeccentricityasthevectorsumwitha
properandforcedcomponentwithindependentphasepara-
meterswasagoodmatchtotheobservations.G.M.Kennedy
(2020) foundthatthosesameobservationsshowedevidence
thatthenorthwest(NW) ansaofthediskwasnarrowerthanthe
southeast(SE)ansabyabout4au(measuredfromthefull
widthathalf-maximum;FWHM), anddemonstratedthata
modifiedversionofthecomplexeccentricitymodelthat
includedadispersionintheforcedeccentricitycouldallowfor
anarrowerNWansa(proximatetothediskapocenter).
Here,wepresentnewALMAobservationsthatrevealthe
Fomalhautdebrisdiskatunprecedentedresolutionatmilli-
meterwavelengthsallowingustomoreaccuratelyconstrain
thegeometryoftheouterbelt.Thisletterisorganizedas
follows.InSection2,wediscusstheALMAdatasets,their
processing,andourmethodtocorrectforthepropermotion
betweenthethreeepochs.InSection3,wepresentcontinuum
imagesofthealigneddatasetsanddiscusstheCLEAN
parametersthatwereusedtoproducethem.Wethenpresent
MarkovChainMonteCarlo(MCMC) fitresultsfortwo
eccentricdiskmodelsinSection4.Wemeasureanddiscuss
thewidthsofthediskattheansaeusingradialprofilesofthe
diskimagesthathavebeenregriddedtoR–θspacein
Section5.1.Specifically,wefindthatwhilewecanstill
modelthebulkparametersofFomalhaut’seccentricring,our
modelsdonotprovideagoodphysicalinterpretationforthe
disk’sasymmetriesidentifiedinthehigher-resolutiondata
(i.e.,thewidthandbrightnessdifferencesattheansae). We
alsopresentthedetectionofaspectrallinecoincidentwiththe
“GreatDustCloud”(GDC) sourcediscussedinA.Gáspár
etal.(2023) andG.M.Kennedyetal.(2023), which
corroboratesitsnatureasabackgroundgalaxyratherthanan
objectassociatedwiththeFomalhautsystem.Finally,we
discussourinterpretationinthecontextofalternative
descriptionsofthissystem,suchasthosewithdifferencesin
theirradialprofileparameterizationsandorbitaleccentricity
distributionsoforbitaleccentricities,includingthatof
J.B.Lovelletal.(2025) analyzingthesesameobservations.
2.Data
ThreeepochsofALMAdatainBand6wereobtained
fromthearchive:ahigh-resolutionpointingonthecentral
starobservedinCycle2(J.A.Whiteetal.2017,
ID#2013.1.00486.S), aseven-pointingmosaicfromCycle3
(M.A.MacGregoretal.2017, ID#2015.1.00966.S), andtwo
high-resolutionpointingsatthediskapsesfromCycle5
(ID#2017.1.01043.S). TheCycle5observationsaresummar-
izedinTable1.TheobservationsfromCycle2werecalibrated
usingALMApipelineversion4.3.1,theCycle3datawere
processedwith4.5.3,andtheCycle5datawithversion5.1.1-
5,allofwhicharethepipelineversionsrecommendedby
ALMAsupportforeachrespectivedataset.Theflux
calibrationanddataweightswereinspectedacrossthethree
datasetsandwerefoundtobeconsistentacrossthethree
epochs,giventheirrespectivebaselinecoverage,andsono
furtherchangesweremadetoeitherproduct.Weusedtwo
versionsofCASAforthiswork:version6.5.1-23wasusedfor
alldatamanipulationtaskssuchasaveragingandvisibility
modelsubtraction,whileversion6.6.0-20wasusedexclu-
sivelyforimagingduetorelevantupdatestotclean.
ThedatawerethenconcatenatedintoasingleCASA
measurementset.Toreducethedatavolume,themeasurement
setwasaverageddownto8channelsforthethreecontinuum
spectralwindowsand128channelsforthespectralwindow
centeredonthe230.538GHzCOline.Wethentimeaveraged
thedatato30sintervals.
Self-calibrationwasattemptedtocorrectforphase
offsetsintroducedfromthestar’spropermotion(328.95,
−164.67masyr
−1
,comparabletothenetbeamsizeacrossthe
threeepochs,F.vanLeeuwen2007), butwasunsuccessfuldue
tolowsignal-to-noiseratio.Instead,weusedtheCASA
implementationofuvmodelfit(I.Martí-Vidaletal.2014)
tofitforthepositionofthestarforeachuniqueobservation
andpointing(atotalof52fits).Wethenusedthelast
observationfromtheCycle3dataasthereferencepositionof
thestar,α=22
h
57
m
39.450801andδ=−29°37′22.69400 ,
andusedthefixplanetsandfixvisfunctionsto
manuallycorrectthephasecenterandUVWbaselinepositions
tothereferencestellarposition.Tovalidatethepropermotion
correction,weimagedeachofthe52separatepointings/
observationsandanalyzedthestellarpositionintheimagesfor
eachsetofpointings.Wefoundthatthestellarcentroid
positionwasprecisetotwopixels(0.1
,orabout1/5ofthe
synthesizedbeamoftheimageinSection3).Thescatteris
dominatedbythestarcoincidingwiththefirstnullinthe
Table1
ALMAObservationsfromProject2017.1.01043.S
Date Antennas Baselines PWV Obs.Time
(m) (mm) (minutes)
2018Sep8 45 15.1–783.5 1.3 73.1
45 15.1–783.5 1.2 72.7
2018Sep17 45 15.1–1245.6 0.4 78.4
2018Sep23 48 15.1–1397.8 0.5 73.1
2018Sep26 47 15.1–1397.8 0.8 73.7
2
TheAstrophysicalJournalLetters,990:L40(10pp), 2025September10 Chittidietal.
primarybeamfortheNWansapointingfromtheseven-
pointingCycle3data.Theresultingmeasurementsetisproper
motioncorrectedforanyemissioncomovingwiththestarand
debrisdisk,butwithsmearedemissionforbackground
sources.
3.Imaging
Thedatawereimagedtogetherusingthemultifrequency
synthesistaskintcleanimplementedinCASAv.6.6.0-20in
threeways:firstwehighlightjusttheNWandSEansae
pointingsfromonlytheCycle5data,thenweevaluatethose
sametwopointingsbutwiththeshorter-baselineCycle3data
tohighlighttheimprovementtothetotalflux,andfinallywe
mosaicalloftheavailabledatainBand6toproduceasingle
image.Weoptedtousethemultiscaledeconvolution
methodwiththeparameterscales=[0,10,20], which
allowsforGaussiansourcesintheCLEANmodelmapequalto
thenumberofpixelsspecifiedbyscales(0isequivalentto
typicalCLEANpointsource). Wefoundthatthisresultedina
CLEANmodelthatlookedmorelikeacontinuous,resolved
diskratherthanacollectionofpointsourcesasproducedwith
thehogbomalgorithm.Sincethemosaicswereconstructed
withunevensensitivityacrossthepointings,andthedisk
extendsacrossanyindividualpointing,alloftheimages
presentedhereareprimary-beamcorrected.Theimagepixel
scalewassetsothatthesynthesizedbeammajoraxiscovered
approximately10pixelsgiventheselecteddataforeach
image,andthenthechoiceofscaleswasdetermined
experimentallysothatthelargestscalewassmallerthanthe
apparentwidthofthedisk.Inaddition,weusedanelliptical
CLEANmaskthatencompassedtheouterdiskandstarwith
parameters:α=22
h
57
m
39.443347,δ=−29°37′21.12931 ,
a=25
″
,b=12. 5,andθ=338°. Fortheimageofjustthe
Cycle5data,weimagedthedatatothe5%gainlevelto
includethestrongly-detectedstar,butusedaprimary-beam
maskatthe30%gainlevelinlieuoftheellipticalmask
discussedabovetoavoidcreatinganomalouslybrightCLEAN
sourcesattheedgeoftheprimarybeam.ThecombinedCycle
3and5image,aswellasthefullmosaicwereimagedtothe
20%gainlevel.AllimageswereCLEANed to3×therms
noiselevel.Wepresentnaturally-weightedimagesinthisletter
inFigure1.Briggs-weightedimagesarepresentedand
analyzedinJ.B.Lovelletal.(2025).
InFigure1(a),weshowtheimageoftheCycle5data,while
inFigure1(b)weshowthecombinedCycle3andCycle5
data.Thermsnoisenearthephasecentersisabout7μJy,and
thenaturally-weightedbeamsizesare×0.500.39
and×0.570.44
,respectively.InFigure1(c),wepresentthe
mosaicimageofalloftheavailabledata.Thesynthesized
beamsizeis×1.090.81
andthermsnoisenearthepointing
centersisalso7μJy.Thedifferentresolutionsbetweenthe
Cycle2andCycle5dataandthelackofhigh-resolution
observationsforthefour“intermediate”pointingsadjacentto
thediskminoraxisresultinimageartifactsthatincreasethe
rmsnoisenearthestar.Inaddition,theinconsistentresolution
effectivelyresultsinarestoringbeamsizethatislargerthan
thenaturally-weightedbeamsizeforthedataatthediskansae
(Figure1(b)),andsointheanalysisthatfollowsweconsider
boththefullandpartialmosaicsdependingonourusecase.
ThediskmodelinginSection4bypassesthenonlinearity
inherenttoCLEANandtheaboveartifactsaltogetherby
directlycomparingthedataandmodelvisibilities.
3.1.ImageAnalysis
Inordertoanalyzeazimuthalvariationsinthewidthofthe
disk,wedeprojectthediskimagefromFigure1(b)(whichhasa
smallerbeamsizethanthefullmosaic) usingthebest-fit
inclinationandpositionanglefromtheMCMCresultspresented
laterinTable2,i=66°.
5andPA=335°. 84.Wethenresample
theimageontoagridofthecircumstellarradiusandazimuthal
anglebyaveragingthefluxofpixelscorrespondingtothesame
dR–dθ bin,presentedinFigure2(a).Theangularcoordinatesare
measuredrelativetothedisk’spositionangle(i.e.,thesky
plane), so0°and360°correspondto335°.
84onthesky.Inthis
view,thetruepericenterandapocenterappeartooccurat≈45°,
consistentwiththeresultspresentedlaterfortheω
f
parameter
fromtheMCMCfits(thoughwenotethattheapparentapsides
fromtheseplotsaloneareobscuredbyboththediskscaleheight
andchangeinresolutionduetothebeampositionangle). The
projectionleadstoincreasedartifactsatthediskminoraxes,
thoughthisregionismaskedbyourchoiceoftheprimarybeam
limitinthetwo-pointingmosaic.
Wethenusethedeprojectedmaptocreate1Dradial
profilesusinga10°wedgecenteredatthediskansae(0°and
180°alongthex-axis) inFigure2(e).The3σuncertaintyin
theALMAdataprofilesareplottedbyusingthe
7μJybeam
−1
measuredintheimageandasafunctionof
numberofbinnedpixelsrelativetothenumberofpixels
inthebeamarea:σ
profile
=7μJy/
/ /
N N
pixbeampi
xbin .We
comparethepeak-to-peakbrightnessratiobetweenthe
radialprofilesandmeasurea21%±1%brightnessenhance-
mentattheNWansa,alargerdifferencethanreportedin
M.A.MacGregoretal.(2017). Thislikelyresultsfromthe
lowresolutionofthe2017observations,whichdonotfully
resolvethedisk’swidth.ConsideringjusttheSEansaof
thedisk,thepeakbrightnessis0.21±0.02and0.25±
0.03mJyarcsec
−2
inthe2017dataandthenewdata
presentedhere(Figure1(a)),respectively,consistentwithin
themutualuncertainties.Thedifferenceinpeakbrightnessis
morenoticeableattheNWansawherethepeaksare
0.23±0.02and0.30±0.03mJyarcsec
−2
,respectively.
However,theNWansaiswheretheresolutiondifference
betweenthetwodatasetsismostsignificantgiventhenarrow
diskwidth.Wecomputetheintegratedfluxtobe0.23±
0.02mJyand0.27±0.03mJy,respectively,toaccountfor
thisresolutiondifferenceandnotethatthesevaluesalso
overlapwithinthemutualuncertainties.
Wecomputethewidthofthediskateachansabymeasuring
theFWHMoftheradialprofiles.TheFWHMis16aufortheSE
ansaprofile,and12auforNWansaprofile,resultinginawidth
differenceof4au.Thewidthsaredenotedasthehighlighted
regionsinFigure2(e).G.M.Kennedy(2020) alsomeasureda
4auFWHMwidthdifference,althoughtherespectiveansaealso
eachappeared4auwideratlowerresolution.
Theradialprofilesanalyzedherealsohaveacomplex
shape.Weidentifythepresenceof“shoulders”ofemission
locatedatabout15–20auinteriorandexteriortotheemission
peaks(seethecoloredarrowsinFigure2(e)).Theinneredge
oftheapocenteransaradialprofileisalsosharperthanthe
outeredgeorthepericenteransaprofile.Weassesshowthese
extendedfeaturesinfluencetheprofileshapebymeasuring
thefullwidthat20%maximum.Atthislowerfluxthreshold,
theSEansawidthis29auandtheNWansawidthis21au,or
twicethewidthdifferenceasbefore.Thissuggeststhatthese
lowsurfacebrightnessfeaturesaresignificant.Inthe
3
TheAstrophysicalJournalLetters,990:L40(10pp), 2025September10 Chittidietal.
pointingCycle3data.Theresultingmeasurementsetisproper
motioncorrectedforanyemissioncomovingwiththestarand
debrisdisk,butwithsmearedemissionforbackground
sources.
3.Imaging
Thedatawereimagedtogetherusingthemultifrequency
synthesistaskintcleanimplementedinCASAv.6.6.0-20in
threeways:firstwehighlightjusttheNWandSEansae
pointingsfromonlytheCycle5data,thenweevaluatethose
sametwopointingsbutwiththeshorter-baselineCycle3data
tohighlighttheimprovementtothetotalflux,andfinallywe
mosaicalloftheavailabledatainBand6toproduceasingle
image.Weoptedtousethemultiscaledeconvolution
methodwiththeparameterscales=[0,10,20], which
allowsforGaussiansourcesintheCLEANmodelmapequalto
thenumberofpixelsspecifiedbyscales(0isequivalentto
typicalCLEANpointsource). Wefoundthatthisresultedina
CLEANmodelthatlookedmorelikeacontinuous,resolved
diskratherthanacollectionofpointsourcesasproducedwith
thehogbomalgorithm.Sincethemosaicswereconstructed
withunevensensitivityacrossthepointings,andthedisk
extendsacrossanyindividualpointing,alloftheimages
presentedhereareprimary-beamcorrected.Theimagepixel
scalewassetsothatthesynthesizedbeammajoraxiscovered
approximately10pixelsgiventheselecteddataforeach
image,andthenthechoiceofscaleswasdetermined
experimentallysothatthelargestscalewassmallerthanthe
apparentwidthofthedisk.Inaddition,weusedanelliptical
CLEANmaskthatencompassedtheouterdiskandstarwith
parameters:α=22
h
57
m
39.443347,δ=−29°37′21.12931 ,
a=25
″
,b=12. 5,andθ=338°. Fortheimageofjustthe
Cycle5data,weimagedthedatatothe5%gainlevelto
includethestrongly-detectedstar,butusedaprimary-beam
maskatthe30%gainlevelinlieuoftheellipticalmask
discussedabovetoavoidcreatinganomalouslybrightCLEAN
sourcesattheedgeoftheprimarybeam.ThecombinedCycle
3and5image,aswellasthefullmosaicwereimagedtothe
20%gainlevel.AllimageswereCLEANed to3×therms
noiselevel.Wepresentnaturally-weightedimagesinthisletter
inFigure1.Briggs-weightedimagesarepresentedand
analyzedinJ.B.Lovelletal.(2025).
InFigure1(a),weshowtheimageoftheCycle5data,while
inFigure1(b)weshowthecombinedCycle3andCycle5
data.Thermsnoisenearthephasecentersisabout7μJy,and
thenaturally-weightedbeamsizesare×0.500.39
and×0.570.44
,respectively.InFigure1(c),wepresentthe
mosaicimageofalloftheavailabledata.Thesynthesized
beamsizeis×1.090.81
andthermsnoisenearthepointing
centersisalso7μJy.Thedifferentresolutionsbetweenthe
Cycle2andCycle5dataandthelackofhigh-resolution
observationsforthefour“intermediate”pointingsadjacentto
thediskminoraxisresultinimageartifactsthatincreasethe
rmsnoisenearthestar.Inaddition,theinconsistentresolution
effectivelyresultsinarestoringbeamsizethatislargerthan
thenaturally-weightedbeamsizeforthedataatthediskansae
(Figure1(b)),andsointheanalysisthatfollowsweconsider
boththefullandpartialmosaicsdependingonourusecase.
ThediskmodelinginSection4bypassesthenonlinearity
inherenttoCLEANandtheaboveartifactsaltogetherby
directlycomparingthedataandmodelvisibilities.
3.1.ImageAnalysis
Inordertoanalyzeazimuthalvariationsinthewidthofthe
disk,wedeprojectthediskimagefromFigure1(b)(whichhasa
smallerbeamsizethanthefullmosaic) usingthebest-fit
inclinationandpositionanglefromtheMCMCresultspresented
laterinTable2,i=66°.
5andPA=335°. 84.Wethenresample
theimageontoagridofthecircumstellarradiusandazimuthal
anglebyaveragingthefluxofpixelscorrespondingtothesame
dR–dθ bin,presentedinFigure2(a).Theangularcoordinatesare
measuredrelativetothedisk’spositionangle(i.e.,thesky
plane), so0°and360°correspondto335°.
84onthesky.Inthis
view,thetruepericenterandapocenterappeartooccurat≈45°,
consistentwiththeresultspresentedlaterfortheω
f
parameter
fromtheMCMCfits(thoughwenotethattheapparentapsides
fromtheseplotsaloneareobscuredbyboththediskscaleheight
andchangeinresolutionduetothebeampositionangle). The
projectionleadstoincreasedartifactsatthediskminoraxes,
thoughthisregionismaskedbyourchoiceoftheprimarybeam
limitinthetwo-pointingmosaic.
Wethenusethedeprojectedmaptocreate1Dradial
profilesusinga10°wedgecenteredatthediskansae(0°and
180°alongthex-axis) inFigure2(e).The3σuncertaintyin
theALMAdataprofilesareplottedbyusingthe
7μJybeam
−1
measuredintheimageandasafunctionof
numberofbinnedpixelsrelativetothenumberofpixels
inthebeamarea:σ
profile
=7μJy/
/ /
N N
pixbeampi
xbin .We
comparethepeak-to-peakbrightnessratiobetweenthe
radialprofilesandmeasurea21%±1%brightnessenhance-
mentattheNWansa,alargerdifferencethanreportedin
M.A.MacGregoretal.(2017). Thislikelyresultsfromthe
lowresolutionofthe2017observations,whichdonotfully
resolvethedisk’swidth.ConsideringjusttheSEansaof
thedisk,thepeakbrightnessis0.21±0.02and0.25±
0.03mJyarcsec
−2
inthe2017dataandthenewdata
presentedhere(Figure1(a)),respectively,consistentwithin
themutualuncertainties.Thedifferenceinpeakbrightnessis
morenoticeableattheNWansawherethepeaksare
0.23±0.02and0.30±0.03mJyarcsec
−2
,respectively.
However,theNWansaiswheretheresolutiondifference
betweenthetwodatasetsismostsignificantgiventhenarrow
diskwidth.Wecomputetheintegratedfluxtobe0.23±
0.02mJyand0.27±0.03mJy,respectively,toaccountfor
thisresolutiondifferenceandnotethatthesevaluesalso
overlapwithinthemutualuncertainties.
Wecomputethewidthofthediskateachansabymeasuring
theFWHMoftheradialprofiles.TheFWHMis16aufortheSE
ansaprofile,and12auforNWansaprofile,resultinginawidth
differenceof4au.Thewidthsaredenotedasthehighlighted
regionsinFigure2(e).G.M.Kennedy(2020) alsomeasureda
4auFWHMwidthdifference,althoughtherespectiveansaealso
eachappeared4auwideratlowerresolution.
Theradialprofilesanalyzedherealsohaveacomplex
shape.Weidentifythepresenceof“shoulders”ofemission
locatedatabout15–20auinteriorandexteriortotheemission
peaks(seethecoloredarrowsinFigure2(e)).Theinneredge
oftheapocenteransaradialprofileisalsosharperthanthe
outeredgeorthepericenteransaprofile.Weassesshowthese
extendedfeaturesinfluencetheprofileshapebymeasuring
thefullwidthat20%maximum.Atthislowerfluxthreshold,
theSEansawidthis29auandtheNWansawidthis21au,or
twicethewidthdifferenceasbefore.Thissuggeststhatthese
lowsurfacebrightnessfeaturesaresignificant.Inthe
3
TheAstrophysicalJournalLetters,990:L40(10pp), 2025September10 Chittidietal.
subsequentmodelingsection,weconsidertwoeccentricdisk
models.Ofkeyconcerniswhethereitherofthesemodelsare
abletoreproducethe21%brightnessenhancementat
apocenterandtheapproximately4audifferenceinwidth
identifiedinthedata.
4.EccentricDiskModels
Theunderlyingorbitalparametersthatdescribethesurface
brightnessdistributionaremodeledusingaparticle-based
approachthatwasinvestigatedinM.A.MacGregoretal.
(2017) andG.M.Kennedy(2020) (andmorerecentlyforthe
HD53143diskinM.A.MacGregoretal.2022). Weconsider
twodiskmodelswithcomplexforcedandpropereccentricities,e
f
ande
p
.Thefirstmodelmostcloselymirrorsthemodelinitially
studiedinM.A.MacGregoretal.(2017), whilethesecondmodel
includesanadditionalfreeparameterforaGaussiandispersionin
thepropereccentricity,whichG.M.Kennedy(2020) demon-
stratedcouldallowforanarrowerapocenterthanpericenter.20
00
10
00
0
00
10
00
20
00
a)
0
20
40
60
80
100
Jy beam
1
b)
0
20
40
60
80
100
Jy beam
1
10
00
0
00
10
00
20
00
20
00
10
00
0
00
10
00
20
00
Apocenter
NW Ansa
Pericenter
SE Ansa
c)
0
50
100
150
200
250
Jy beam
1
20
00
10
00
0
00
10
00
20
00
d)
50 AU
4
5
6
7
8
p
j
MJy sr
1
j
RA Oset (
00
)
Dec Oset (
00
)
Figure1.Topleft:AmosaicoftheCycle5(C5)long-baselinedata.Thedatawereimagedtothe5%gainleveltohighlightthedetectionofthestar,whichsitsatthe
edgeoftheprimarybeambetweentheNWandSEansaepointings.Thenaturally-weightedbeamsizeis×0.500.39
(indicatedbytheellipseinthelowerleft
cornerofeachpanel) andthermsnoisenearthepointingcentersis7μJybm
−1
.Theaxesarethestellocentricoffsetinarcsecondsandtheimageisnorth-aligned.
Topright:AmosaicoftheNWandSEansaepointingsusingthecoalignedshorter-baselineCycle3(C3)dataaswellasthelongerbaselineC5data.Thenaturally-
weightedbeamsizeisslightlylarger,0×.570.44
,andthermsisalso7μJybm
−1
.Bottomleft:AmosaicofallsevenpointingsofALMABand6observations
fromC2,C3,andC5.Thenaturally-weightedbeamsizeis×1.090.81
andthermsnoisenearthepointingcentersattheansaeis7μJybm
−1
,butdecreasesto
12μJybm
−1
towardtheIB(denotedbythedashedwhitecurves,seeSection5.2).Thelackoflong-baselineobservationsattheintermediatepointingsleadstoa
beamsizethatislargerthanthenaturally-weightedbeamcorrespondingtojusttheNWandSEansaedata,asinthetopright.Bottomright:JWST25.5μmimage
fromA.Gáspáretal.(2023) withoverlaidcontoursfromthebottomrightALMAmosaiccorrespondingtothe10σ,20σ,and30σfluxlevels.Thisimagehasbeen
scaledtohighlighttheintermediateandinnerdisks,andthepositionsofthecontourshavebeencorrectedforthepropermotionbetweenthetwoimagedepochs.The
fainterdashedcurvesdenotetheIBboundariesasinthebottomleft.
4
TheAstrophysicalJournalLetters,990:L40(10pp), 2025September10 Chittidietal.
models.Ofkeyconcerniswhethereitherofthesemodelsare
abletoreproducethe21%brightnessenhancementat
apocenterandtheapproximately4audifferenceinwidth
identifiedinthedata.
4.EccentricDiskModels
Theunderlyingorbitalparametersthatdescribethesurface
brightnessdistributionaremodeledusingaparticle-based
approachthatwasinvestigatedinM.A.MacGregoretal.
(2017) andG.M.Kennedy(2020) (andmorerecentlyforthe
HD53143diskinM.A.MacGregoretal.2022). Weconsider
twodiskmodelswithcomplexforcedandpropereccentricities,e
f
ande
p
.Thefirstmodelmostcloselymirrorsthemodelinitially
studiedinM.A.MacGregoretal.(2017), whilethesecondmodel
includesanadditionalfreeparameterforaGaussiandispersionin
thepropereccentricity,whichG.M.Kennedy(2020) demon-
stratedcouldallowforanarrowerapocenterthanpericenter.20
00
10
00
0
00
10
00
20
00
a)
0
20
40
60
80
100
Jy beam
1
b)
0
20
40
60
80
100
Jy beam
1
10
00
0
00
10
00
20
00
20
00
10
00
0
00
10
00
20
00
Apocenter
NW Ansa
Pericenter
SE Ansa
c)
0
50
100
150
200
250
Jy beam
1
20
00
10
00
0
00
10
00
20
00
d)
50 AU
4
5
6
7
8
p
j
MJy sr
1
j
RA Oset (
00
)
Dec Oset (
00
)
Figure1.Topleft:AmosaicoftheCycle5(C5)long-baselinedata.Thedatawereimagedtothe5%gainleveltohighlightthedetectionofthestar,whichsitsatthe
edgeoftheprimarybeambetweentheNWandSEansaepointings.Thenaturally-weightedbeamsizeis×0.500.39
(indicatedbytheellipseinthelowerleft
cornerofeachpanel) andthermsnoisenearthepointingcentersis7μJybm
−1
.Theaxesarethestellocentricoffsetinarcsecondsandtheimageisnorth-aligned.
Topright:AmosaicoftheNWandSEansaepointingsusingthecoalignedshorter-baselineCycle3(C3)dataaswellasthelongerbaselineC5data.Thenaturally-
weightedbeamsizeisslightlylarger,0×.570.44
,andthermsisalso7μJybm
−1
.Bottomleft:AmosaicofallsevenpointingsofALMABand6observations
fromC2,C3,andC5.Thenaturally-weightedbeamsizeis×1.090.81
andthermsnoisenearthepointingcentersattheansaeis7μJybm
−1
,butdecreasesto
12μJybm
−1
towardtheIB(denotedbythedashedwhitecurves,seeSection5.2).Thelackoflong-baselineobservationsattheintermediatepointingsleadstoa
beamsizethatislargerthanthenaturally-weightedbeamcorrespondingtojusttheNWandSEansaedata,asinthetopright.Bottomright:JWST25.5μmimage
fromA.Gáspáretal.(2023) withoverlaidcontoursfromthebottomrightALMAmosaiccorrespondingtothe10σ,20σ,and30σfluxlevels.Thisimagehasbeen
scaledtohighlighttheintermediateandinnerdisks,andthepositionsofthecontourshavebeencorrectedforthepropermotionbetweenthetwoimagedepochs.The
fainterdashedcurvesdenotetheIBboundariesasinthebottomleft.
4
TheAstrophysicalJournalLetters,990:L40(10pp), 2025September10 Chittidietal.
Forbothmodels,themeanlongitude,l,andargumentof
periastron,ω
p
,oforbitingparticlesaredrawnfromauniform
distributionbetween0and2π.Then,theparticlespopulatethe
complexeccentricityplanedefinedbythreefreeparameters:
theforcedeccentricity,e
f
,theforcedargumentofperiastron,
ω
f
,andthepropereccentricity,e
p
.Forthesecondmodel,the
propereccentricityisdrawnfromanormaldistributionwhose
meanise
p
andstandarddeviationise
p .Toavoidnonphysical
parameters,wetaketheabsolutevalueofe
p
soe
p
>0,and
redrawparticlesfromthedistributionwhene
p
>1.Wethen
solveKepler’sequationforthetrueanomaly,f,withthe
keplercode(D.Foreman-Mackeyetal.2021).
Next,thesemimajoraxisofeachparticleisdrawnfroma
uniformdistributiondefinedbetweena−Δa/2 anda+Δa/
2,whereaandΔaarefreeparameters.Then,theradial
positionofeachparticleissolvedbytheequation( )
()
()=
+
r
a e
e f
1
1 cos
, 1
2
wherefisthetrueanomaly.
Theparticlesaregivenaheight,z,aboutthediskmidplane
andaredrawnfromanexponentialdistributiondefinedbya
singlefreeparameterscaleheight,h,suchthatz=h/r.We
alsoaccountforthedisk’sgeometry,fittingforaninclination,
i,andpositionangle,PA,definednorthtoeast.R.A.anddecl.
offsets(positiveinthenorthandeastdirections,respectively)
accountforanyglobalpointingoffset.
Tocreateanimage,webintheparticlesontoa2Dspatial
grid(histogram) whosevaluesarethenscaledbyr
−0.5
inorder
tosimulateatemperatureprofile.Thetotaldiskfluxis
normalizedsuchthatF
belt
=∫I
ν
dΩandapointsource
representingthestarisaddedwithfluxF
star
.Tosimplify
comparison,weensurethatbothmodelshave12free
parametersbyfixingthestellarfluxtoF
star
=0.735mJyfor
thesecondmodelwith>0
e
p
.Thislowervalueforthestellar
fluxwasbasedonanearlyfittothedata,butthisis
independentoftheparametersforthediskandshouldnot
impactthebest-fitparameters.However,whenwecomputethe
Bayesianinformationcriterion(BIC) usingthemedian
parametersfromtheemceeposteriors,weusethestellarflux
fromthefirstmodelforanevencomparison.
G.M.Kennedy(2020) andJ.B.Lovell&E.M.Lynch
(2023) demonstratedthatlinedensitymodelssuchastheone
consideredhereneedasufficientnumberofparticlesinorder
toreducetheshot-noiseassociatedwiththerandomly-
generatedmodelandtoeffectivelysamplethesurfacedensity
distributionunderthebeam.G.M.Kennedy(2020) used10
7
particlestomodeltheC3datawithadispersionparameter,
whichweestimateresultedina0.2%model-inducederror
basedontheshot-noiseanalysispresentedinFigureA1of
J.B.Lovell&E.M.Lynch(2023). Weusethatsameanalysis
todeterminethatwewouldneedO
(10
8
)particlesinorderto
achieveasimilarlevelofshot-noiseattheC5resolutionof
0.
5.Weusegalario(M.Tazzarietal.2018) tosamplethe
modelimagesintovisibilitiestocomparewiththeALMA
data.ForeachuniqueobservationID,spectralwindow,and
field,themodelimageisoffsettothecorrectrelativepointing
beforecomparingwiththevisibilities,andthusthenetχ
2
for
eachmodelimageisthesumoftheχ
2
valuesforeachunique
offset.TheparameterspaceisexploredwiththeMCMC
packageemcee(D.Foreman-Mackeyetal.2013) using80
walkersand18,027stepsforthefirstmodeland22,199steps
forthesecondmodel.Weassessthemodelsashaving
convergedwhenthechainswererunforatleast50timesthe
longestautocorrelationtimeforanyoftheparameters.
Analyticmodelslikethesehavelimitationsandoftenfailto
completelymodelcomplexsystems.Forexample,themodels
weemploydonotaccountforeitherdensityenhancementsor
diskbroadeningasafunctionoftrueanomaly.TheFomalhaut
debrisdiskislikelydynamicallyinfluencedbyatleastone
planet(e.g.,A.C.Boleyetal.2012), soacompletemodelof
thesystemwouldrequireuseofN-bodysimulations.However,
asnotedabove,thisanalyticmodeliscomputationally
intensive.AttemptingtofoldanN-bodysimulationintothis
MCMCframeworkwouldtakeaprohibitiveamountof
computingresources.Asaresult,analyticmodelsare
extremelyusefulforunderstandingdiskgeometryandare
widelyusedthroughouttheliterature.SomeN-bodysimula-
tionsareincludedinthecompanionpapertothisletter
(J.B.Lovelletal.2025), andwedefermorecomplicated
modelingtofuturework.
4.1.ModelingResults
Wepresentthemedianposteriorparametersforbothmodels
inTable2,andinFigure3weshowthefull-resolution(i.e.,
notCLEAN’ed orconvolvedwiththesynthesizedbeam) model
imagesatthediskansae.Withthese,wealsopresentthe
residualsthathavebeenrepackagedintoaCASAmeasurement
setandimagedwiththesametcleanparametersasthefull-
mosaicpresentedinFigure1(c).
Thenominaldisksemimajoraxis,inclination,position
angle,andscaleheightforbothmodelsareidenticalandarein
goodagreementwiththeanalogous“uniformsimple”and
“uniformfull”modelfitspresentedinTable1of
G.M.Kennedy(2020). Theforcedandpropereccentricities
forthetwomodelspresentedherearealsoingoodagreement
withintheirparameteruncertainties,buttheydodifferslightly
fromtheresultsinG.M.Kennedy(2020). Here,wefindthat
forthemodelwithoutdispersion,theforcedeccentricityis
about0.02higherandthepropereccentricityisabout0.03
lowerthantheG.M.Kennedy(2020) results.Forthemodel
withdispersion,theresultsaremoreinagreement,butthee
p
parameterisabout0.03lowerinthiswork.Thiscouldbedue
Table2
EccentricDiskModelPosteriors
Parameter ModelWithoute
p
ModelWithe
p
a[au]+
139.48 0.39
0.39 +
139.48 0.42
0.43
Δa[au]+
10.88 2.95
2.33 +
6.66 2.62
2.06
F
disk
[mJy]+
21.95 0.53
0.54 +
22.99 0.68
0.72
F
star
[mJy]+
0.77 0.01
0.01 ⋯
i[°]+
66.50 0.10
0.11 +
66.50 0.11
0.11
PA[°]+
335.84 0.12
0.11 +
335.84 0.11
0.11
e
f+
0.15 0.01
0.01 +
0.14 0.01
0.01
e
p+
0.03 0.01
0.01 +
0.02 0.01
0.01 e
p
⋯+
0.06 0.01
0.01
ω
f
[°]+
45.94 1.51
1.38 +
41.89 4.18
2.94
R.A.
off
[″]+
0.08 0.01
0.01 +
0.08 0.01
0.01
Decl.
off
[″]+
0.05 0.01
0.01 +
0.05 0.01
0.01
h+
0.01 0.01
0.01 +
0.02 0.01
0.01
BIC 27815115 27811668
5
TheAstrophysicalJournalLetters,990:L40(10pp), 2025September10 Chittidietal.
periastron,ω
p
,oforbitingparticlesaredrawnfromauniform
distributionbetween0and2π.Then,theparticlespopulatethe
complexeccentricityplanedefinedbythreefreeparameters:
theforcedeccentricity,e
f
,theforcedargumentofperiastron,
ω
f
,andthepropereccentricity,e
p
.Forthesecondmodel,the
propereccentricityisdrawnfromanormaldistributionwhose
meanise
p
andstandarddeviationise
p .Toavoidnonphysical
parameters,wetaketheabsolutevalueofe
p
soe
p
>0,and
redrawparticlesfromthedistributionwhene
p
>1.Wethen
solveKepler’sequationforthetrueanomaly,f,withthe
keplercode(D.Foreman-Mackeyetal.2021).
Next,thesemimajoraxisofeachparticleisdrawnfroma
uniformdistributiondefinedbetweena−Δa/2 anda+Δa/
2,whereaandΔaarefreeparameters.Then,theradial
positionofeachparticleissolvedbytheequation( )
()
()=
+
r
a e
e f
1
1 cos
, 1
2
wherefisthetrueanomaly.
Theparticlesaregivenaheight,z,aboutthediskmidplane
andaredrawnfromanexponentialdistributiondefinedbya
singlefreeparameterscaleheight,h,suchthatz=h/r.We
alsoaccountforthedisk’sgeometry,fittingforaninclination,
i,andpositionangle,PA,definednorthtoeast.R.A.anddecl.
offsets(positiveinthenorthandeastdirections,respectively)
accountforanyglobalpointingoffset.
Tocreateanimage,webintheparticlesontoa2Dspatial
grid(histogram) whosevaluesarethenscaledbyr
−0.5
inorder
tosimulateatemperatureprofile.Thetotaldiskfluxis
normalizedsuchthatF
belt
=∫I
ν
dΩandapointsource
representingthestarisaddedwithfluxF
star
.Tosimplify
comparison,weensurethatbothmodelshave12free
parametersbyfixingthestellarfluxtoF
star
=0.735mJyfor
thesecondmodelwith>0
e
p
.Thislowervalueforthestellar
fluxwasbasedonanearlyfittothedata,butthisis
independentoftheparametersforthediskandshouldnot
impactthebest-fitparameters.However,whenwecomputethe
Bayesianinformationcriterion(BIC) usingthemedian
parametersfromtheemceeposteriors,weusethestellarflux
fromthefirstmodelforanevencomparison.
G.M.Kennedy(2020) andJ.B.Lovell&E.M.Lynch
(2023) demonstratedthatlinedensitymodelssuchastheone
consideredhereneedasufficientnumberofparticlesinorder
toreducetheshot-noiseassociatedwiththerandomly-
generatedmodelandtoeffectivelysamplethesurfacedensity
distributionunderthebeam.G.M.Kennedy(2020) used10
7
particlestomodeltheC3datawithadispersionparameter,
whichweestimateresultedina0.2%model-inducederror
basedontheshot-noiseanalysispresentedinFigureA1of
J.B.Lovell&E.M.Lynch(2023). Weusethatsameanalysis
todeterminethatwewouldneedO
(10
8
)particlesinorderto
achieveasimilarlevelofshot-noiseattheC5resolutionof
0.
5.Weusegalario(M.Tazzarietal.2018) tosamplethe
modelimagesintovisibilitiestocomparewiththeALMA
data.ForeachuniqueobservationID,spectralwindow,and
field,themodelimageisoffsettothecorrectrelativepointing
beforecomparingwiththevisibilities,andthusthenetχ
2
for
eachmodelimageisthesumoftheχ
2
valuesforeachunique
offset.TheparameterspaceisexploredwiththeMCMC
packageemcee(D.Foreman-Mackeyetal.2013) using80
walkersand18,027stepsforthefirstmodeland22,199steps
forthesecondmodel.Weassessthemodelsashaving
convergedwhenthechainswererunforatleast50timesthe
longestautocorrelationtimeforanyoftheparameters.
Analyticmodelslikethesehavelimitationsandoftenfailto
completelymodelcomplexsystems.Forexample,themodels
weemploydonotaccountforeitherdensityenhancementsor
diskbroadeningasafunctionoftrueanomaly.TheFomalhaut
debrisdiskislikelydynamicallyinfluencedbyatleastone
planet(e.g.,A.C.Boleyetal.2012), soacompletemodelof
thesystemwouldrequireuseofN-bodysimulations.However,
asnotedabove,thisanalyticmodeliscomputationally
intensive.AttemptingtofoldanN-bodysimulationintothis
MCMCframeworkwouldtakeaprohibitiveamountof
computingresources.Asaresult,analyticmodelsare
extremelyusefulforunderstandingdiskgeometryandare
widelyusedthroughouttheliterature.SomeN-bodysimula-
tionsareincludedinthecompanionpapertothisletter
(J.B.Lovelletal.2025), andwedefermorecomplicated
modelingtofuturework.
4.1.ModelingResults
Wepresentthemedianposteriorparametersforbothmodels
inTable2,andinFigure3weshowthefull-resolution(i.e.,
notCLEAN’ed orconvolvedwiththesynthesizedbeam) model
imagesatthediskansae.Withthese,wealsopresentthe
residualsthathavebeenrepackagedintoaCASAmeasurement
setandimagedwiththesametcleanparametersasthefull-
mosaicpresentedinFigure1(c).
Thenominaldisksemimajoraxis,inclination,position
angle,andscaleheightforbothmodelsareidenticalandarein
goodagreementwiththeanalogous“uniformsimple”and
“uniformfull”modelfitspresentedinTable1of
G.M.Kennedy(2020). Theforcedandpropereccentricities
forthetwomodelspresentedherearealsoingoodagreement
withintheirparameteruncertainties,buttheydodifferslightly
fromtheresultsinG.M.Kennedy(2020). Here,wefindthat
forthemodelwithoutdispersion,theforcedeccentricityis
about0.02higherandthepropereccentricityisabout0.03
lowerthantheG.M.Kennedy(2020) results.Forthemodel
withdispersion,theresultsaremoreinagreement,butthee
p
parameterisabout0.03lowerinthiswork.Thiscouldbedue
Table2
EccentricDiskModelPosteriors
Parameter ModelWithoute
p
ModelWithe
p
a[au]+
139.48 0.39
0.39 +
139.48 0.42
0.43
Δa[au]+
10.88 2.95
2.33 +
6.66 2.62
2.06
F
disk
[mJy]+
21.95 0.53
0.54 +
22.99 0.68
0.72
F
star
[mJy]+
0.77 0.01
0.01 ⋯
i[°]+
66.50 0.10
0.11 +
66.50 0.11
0.11
PA[°]+
335.84 0.12
0.11 +
335.84 0.11
0.11
e
f+
0.15 0.01
0.01 +
0.14 0.01
0.01
e
p+
0.03 0.01
0.01 +
0.02 0.01
0.01 e
p
⋯+
0.06 0.01
0.01
ω
f
[°]+
45.94 1.51
1.38 +
41.89 4.18
2.94
R.A.
off
[″]+
0.08 0.01
0.01 +
0.08 0.01
0.01
Decl.
off
[″]+
0.05 0.01
0.01 +
0.05 0.01
0.01
h+
0.01 0.01
0.01 +
0.02 0.01
0.01
BIC 27815115 27811668
5
TheAstrophysicalJournalLetters,990:L40(10pp), 2025September10 Chittidietal.
toacombinationofthechoiceoftheuniformdistributionof
semimajoraxesandthe“shoulders.”AsseeninFigure2(e),
thepeakemissionsforbothmodelsneartheapocentersideof
thediskoccurabout4aufurtheroutthanthedata,towardthe
exterior“shoulder.”Theeccentricityparametersandchoiceof
distributioncouldbothaffectwherethesepeaksoccur,and100
120
140
160
180a) Data
NW Ansa
SE Ansa SE Ansa
0
20
40
60
80
100
Jy beam
1
100
120
140
160
180b) ep
= 0
Pericenter
Apocenter
0
20
40
60
80
100
Jy beam
1
100
120
140
160
180c) ep
= 0:06
Pericenter
Apocenter
0
20
40
60
80
100
Jy beam
1
0 45 90 135 180 225 270 315 360
Azimuthal Angle [
]
100
120
140
160
180d) JWST
0
10
20
30
40
MJy sr
1
100 120 140 160 180
R[AU]
0:0
0:5
1:0
F
[mJy beam
1
]
16 AU 12 AU
e) Data
ep
= 0
ep
= 0:06
JWST
0
250
500
F
[MJy sr
1
]
R
[AU]
Figure2.Firstrow:Adeprojectedmapofthecircumstellarradiusvs.azimuthalangleoftheFomalhautouterdiskproducedusingthemosaicinFigure1(b).The
angularcoordinatesarewithrespecttothediskpositionangleof335°.
84.Secondrow:Aradiusvs.azimuthmapforthebest-fitmodelwithoutadispersionin
eccentricity,afterimagingwithCASA. Thirdrow:Sameasthesecond,butforthebest-fitmodelwithadispersionineccentricity.Fourthrow:Aradiusvs.azimuth
mapofthe25.5μmJWSTimageofthedisk.UnliketheALMAimages,theNWansaissignificantlydimmerthantheSEansa.Fifthrow:Radialprofilesproduced
using10°radialcutscenteredat0°(SEansanearapocenter,red)and180°(NWansanearapocenter,yellow) fromtheabovemaps.FortheALMAdata,the
21%±1%brightnessasymmetrybetweentheansaeisstronglyapparent.ThelabeledshadedregionsarethemeasuredFWHMofthedata.“Shoulders”ofemission
at20auinteriorandexteriortothepeakfluxintheSEansaprofileandexteriortoNWansaprofilearedenotedbythecoloredarrows.Theradiusofpeakemission
differbetweenthedataandthemodels,whichisdiscussedinthetext.
6
TheAstrophysicalJournalLetters,990:L40(10pp), 2025September10 Chittidietal.
semimajoraxesandthe“shoulders.”AsseeninFigure2(e),
thepeakemissionsforbothmodelsneartheapocentersideof
thediskoccurabout4aufurtheroutthanthedata,towardthe
exterior“shoulder.”Theeccentricityparametersandchoiceof
distributioncouldbothaffectwherethesepeaksoccur,and100
120
140
160
180a) Data
NW Ansa
SE Ansa SE Ansa
0
20
40
60
80
100
Jy beam
1
100
120
140
160
180b) ep
= 0
Pericenter
Apocenter
0
20
40
60
80
100
Jy beam
1
100
120
140
160
180c) ep
= 0:06
Pericenter
Apocenter
0
20
40
60
80
100
Jy beam
1
0 45 90 135 180 225 270 315 360
Azimuthal Angle [
]
100
120
140
160
180d) JWST
0
10
20
30
40
MJy sr
1
100 120 140 160 180
R[AU]
0:0
0:5
1:0
F
[mJy beam
1
]
16 AU 12 AU
e) Data
ep
= 0
ep
= 0:06
JWST
0
250
500
F
[MJy sr
1
]
R
[AU]
Figure2.Firstrow:Adeprojectedmapofthecircumstellarradiusvs.azimuthalangleoftheFomalhautouterdiskproducedusingthemosaicinFigure1(b).The
angularcoordinatesarewithrespecttothediskpositionangleof335°.
84.Secondrow:Aradiusvs.azimuthmapforthebest-fitmodelwithoutadispersionin
eccentricity,afterimagingwithCASA. Thirdrow:Sameasthesecond,butforthebest-fitmodelwithadispersionineccentricity.Fourthrow:Aradiusvs.azimuth
mapofthe25.5μmJWSTimageofthedisk.UnliketheALMAimages,theNWansaissignificantlydimmerthantheSEansa.Fifthrow:Radialprofilesproduced
using10°radialcutscenteredat0°(SEansanearapocenter,red)and180°(NWansanearapocenter,yellow) fromtheabovemaps.FortheALMAdata,the
21%±1%brightnessasymmetrybetweentheansaeisstronglyapparent.ThelabeledshadedregionsarethemeasuredFWHMofthedata.“Shoulders”ofemission
at20auinteriorandexteriortothepeakfluxintheSEansaprofileandexteriortoNWansaprofilearedenotedbythecoloredarrows.Theradiusofpeakemission
differbetweenthedataandthemodels,whichisdiscussedinthetext.
6
TheAstrophysicalJournalLetters,990:L40(10pp), 2025September10 Chittidietal.
mayexplainwhyourresultsappeartodifferfromtheliterature
andthedata.Finally,themodelwithdispersionhasaslightly
lowerargumentofperiastronthatiscomparabletothe
41°±1°constraintinG.M.Kennedy(2020). Thedifference
invaluesbetweenourtwomodelsdoesnotappeartobe
statisticallysignificant.
Themostnotabledifferencesbetweenthetwomodelsinthis
workarethedisk’swidthandthetotaldiskflux.Theinclusion
ofthee
p
parameterwasexpectedtoreducethedisk’swidth
suchthattheexpression()()+a a
e
2 2
p
shouldcorrespond
tothedisk’swidthwithoutthedispersionparameter.Indeed,we
findthattheexpressionwouldyieldaneffectivediskwidthof
10.70au,inexcellentagreementwiththe10.88auresultforthe
modelwithoutthedispersionparameter.G.M.Kennedy(2020)
reportedabest-fitdispersionparameterof0.09,higherthanwe
findhere,butthisislikelyrelatedtothedifferencesnoted
above.Themodelwithdispersionhasatotaldiskfluxabout
1mJybrighterthantheothermodel,thoughthisdifferenceis
onlyatthe2σsignificancelevel.Thismaybeinfluencedby
the“shoulders”observedintheradialprofiles,sincethemodel
withdispersionhasawiderprofilethatoverlapswiththese
featureswhilethemodelwithoutthedispersiondoesnot.
M.A.MacGregoretal.(2017) reportedatotaldiskfluxof
24.7±0.1mJy,higherthaneithermodelhere,eitherbecause
the“shoulders”areunresolvedinthoseobservationsorthe
inclusionofthelong-baselinevisibilitiesweighdownthetotal
fluxinthemodelfitsinthiswork.Notably,neithermodelisable
toreproducethe21%fluxdifferencebetweentheansae,instead
producingroughlyequalpeaks.
WecomputetheBICforourtwomodels:BIC () ()= +knln , 2 vis
2
wherekisthenumberofparameters(12forbothmodels) and
n
vis
isthetotalnumberofvisibilities(realandimaginary,
>8.6×10
6
).WeincludetheBICvaluesinTable2.Themodel
withthelowestBICisthepreferredmodel,withdifferencesof
>10indicatingthatthelowerBICisvery-stronglypreferred
(R.E.Kass&A.E.Raftery1995). Wefindthatthemodel
withe
p
isvery-stronglypreferredoverthemodelwithout,in
agreementwithsimilarfindingsinG.M.Kennedy(2020).
5.Discussion
5.1.VariationoftheDisk’sWidth
Wegenerateadditionalradialmapsforthetwobest-fit
models(theseareimagedwithCLEANusingthesame
parametersasthetwo-pointingmosaic) andforthe25.5μm
JWSTMIRIimageusingthesametechniquepresentedin
Section3.1.Wecompareradialprofilecutswiththedatain
Figure2(e).Rathercritically,neithermodelinvestigatedinthis
workseemstobeabletoreproducethe21%brightness
asymmetryorthe4auwidthdifference.Thepeakfluxesinthe
modelprofilesarenearlythesameateachansaeandthe
measuredFWHMappearstobe1auwideratapocenteransa—
thecompleteoppositeofthetrendinthedata.Thesemeasured
widthsarepresentedinTable3.
Themodelradialprofilesalsoelucidatetheresidualspresent
inFigure3—thesebest-fitmodelshavetheirpeakemission
occurringatslightlylowerradiinearpericenterandnoticeably
higherradiinearapocentercomparedtothedata.Thismaybe
duetoourchoiceofauniformdistributionforthediskparticle
semimajoraxes.Forexample,adifferentprescription,suchas
aGaussian,couldalterwherethepeakbrightnessoccurs.
Someofthekeyparametersinthemodelfitsarecorrelated
(e.g.,Δaande
p
inthe=0
e
p model) orhavenon-Gaussian
posteriordistributions,andsothemedianparametervalues
selectedforthesemodelsmaynottrulybethe“best-fit”ones.20
00
15
00
10
00
Model
ep
=0
25 AU
ep
=0:06
5
00
10
00
15
00
ep
=0 ep
=0:06
0
0.02
0.04
0.06
0.08
0.1
Jy px
1
0
00
5
00
10
00
20
00
15
00
10
00
Residual
0
00
5
00
10
00
10
00
5
00
0
00
5
00
10
00
15
00
10
00
5
00
0
00
25 AU
-50
-25
0
25
50
Jy beam
1
RA Oset (
00
)
Dec Oset (
00
)
Figure3.Toprow:Apocenterandpericenterplotsofthebest-fiteccentricdiskmodelswithout(firstcolumn) andwith(secondcolumn) afreeparameterdescribinga
Gaussiandispersioninthepropereccentricityfortheparticles(seeTable2).Theseimagesarefull-resolution,i.e.,notconvolvedwiththesynthesizedbeam,inorder
tohighlighttheunderlyingdifferencesinintensity.Bottomrow:Residualbetweenthedataandmodelwithcontourshighlighting−10σ, −5σ,5σ,and10σnoise
level.ThesewereimagedwithtcleanwiththesameparametersasinFigure1(c).TheellipsesindicatethesizeoftheALMAbeam.
7
TheAstrophysicalJournalLetters,990:L40(10pp), 2025September10 Chittidietal.
andthedata.Finally,themodelwithdispersionhasaslightly
lowerargumentofperiastronthatiscomparabletothe
41°±1°constraintinG.M.Kennedy(2020). Thedifference
invaluesbetweenourtwomodelsdoesnotappeartobe
statisticallysignificant.
Themostnotabledifferencesbetweenthetwomodelsinthis
workarethedisk’swidthandthetotaldiskflux.Theinclusion
ofthee
p
parameterwasexpectedtoreducethedisk’swidth
suchthattheexpression()()+a a
e
2 2
p
shouldcorrespond
tothedisk’swidthwithoutthedispersionparameter.Indeed,we
findthattheexpressionwouldyieldaneffectivediskwidthof
10.70au,inexcellentagreementwiththe10.88auresultforthe
modelwithoutthedispersionparameter.G.M.Kennedy(2020)
reportedabest-fitdispersionparameterof0.09,higherthanwe
findhere,butthisislikelyrelatedtothedifferencesnoted
above.Themodelwithdispersionhasatotaldiskfluxabout
1mJybrighterthantheothermodel,thoughthisdifferenceis
onlyatthe2σsignificancelevel.Thismaybeinfluencedby
the“shoulders”observedintheradialprofiles,sincethemodel
withdispersionhasawiderprofilethatoverlapswiththese
featureswhilethemodelwithoutthedispersiondoesnot.
M.A.MacGregoretal.(2017) reportedatotaldiskfluxof
24.7±0.1mJy,higherthaneithermodelhere,eitherbecause
the“shoulders”areunresolvedinthoseobservationsorthe
inclusionofthelong-baselinevisibilitiesweighdownthetotal
fluxinthemodelfitsinthiswork.Notably,neithermodelisable
toreproducethe21%fluxdifferencebetweentheansae,instead
producingroughlyequalpeaks.
WecomputetheBICforourtwomodels:BIC () ()= +knln , 2 vis
2
wherekisthenumberofparameters(12forbothmodels) and
n
vis
isthetotalnumberofvisibilities(realandimaginary,
>8.6×10
6
).WeincludetheBICvaluesinTable2.Themodel
withthelowestBICisthepreferredmodel,withdifferencesof
>10indicatingthatthelowerBICisvery-stronglypreferred
(R.E.Kass&A.E.Raftery1995). Wefindthatthemodel
withe
p
isvery-stronglypreferredoverthemodelwithout,in
agreementwithsimilarfindingsinG.M.Kennedy(2020).
5.Discussion
5.1.VariationoftheDisk’sWidth
Wegenerateadditionalradialmapsforthetwobest-fit
models(theseareimagedwithCLEANusingthesame
parametersasthetwo-pointingmosaic) andforthe25.5μm
JWSTMIRIimageusingthesametechniquepresentedin
Section3.1.Wecompareradialprofilecutswiththedatain
Figure2(e).Rathercritically,neithermodelinvestigatedinthis
workseemstobeabletoreproducethe21%brightness
asymmetryorthe4auwidthdifference.Thepeakfluxesinthe
modelprofilesarenearlythesameateachansaeandthe
measuredFWHMappearstobe1auwideratapocenteransa—
thecompleteoppositeofthetrendinthedata.Thesemeasured
widthsarepresentedinTable3.
Themodelradialprofilesalsoelucidatetheresidualspresent
inFigure3—thesebest-fitmodelshavetheirpeakemission
occurringatslightlylowerradiinearpericenterandnoticeably
higherradiinearapocentercomparedtothedata.Thismaybe
duetoourchoiceofauniformdistributionforthediskparticle
semimajoraxes.Forexample,adifferentprescription,suchas
aGaussian,couldalterwherethepeakbrightnessoccurs.
Someofthekeyparametersinthemodelfitsarecorrelated
(e.g.,Δaande
p
inthe=0
e
p model) orhavenon-Gaussian
posteriordistributions,andsothemedianparametervalues
selectedforthesemodelsmaynottrulybethe“best-fit”ones.20
00
15
00
10
00
Model
ep
=0
25 AU
ep
=0:06
5
00
10
00
15
00
ep
=0 ep
=0:06
0
0.02
0.04
0.06
0.08
0.1
Jy px
1
0
00
5
00
10
00
20
00
15
00
10
00
Residual
0
00
5
00
10
00
10
00
5
00
0
00
5
00
10
00
15
00
10
00
5
00
0
00
25 AU
-50
-25
0
25
50
Jy beam
1
RA Oset (
00
)
Dec Oset (
00
)
Figure3.Toprow:Apocenterandpericenterplotsofthebest-fiteccentricdiskmodelswithout(firstcolumn) andwith(secondcolumn) afreeparameterdescribinga
Gaussiandispersioninthepropereccentricityfortheparticles(seeTable2).Theseimagesarefull-resolution,i.e.,notconvolvedwiththesynthesizedbeam,inorder
tohighlighttheunderlyingdifferencesinintensity.Bottomrow:Residualbetweenthedataandmodelwithcontourshighlighting−10σ, −5σ,5σ,and10σnoise
level.ThesewereimagedwithtcleanwiththesameparametersasinFigure1(c).TheellipsesindicatethesizeoftheALMAbeam.
7
TheAstrophysicalJournalLetters,990:L40(10pp), 2025September10 Chittidietal.
Onesubtledetailinthemodelprofilesthatincludea
dispersionparameteristhatthefluxfallsoffmoregently,
resultinginwiderlowsurfacebrightnessfeatures.Thiscould
atleastpartiallyexplainthe“shoulders”seeninthedata,and
mayalsobethereasonthismodelisstatisticallypreferredover
theotheronedespitethefactthatitcannotreproducethe
brightnessasymmetryorwidthdifference.These“shoulders”
appearslightlymoredistinctfromthecentralpeakofemission
comparedtothemoresmoothprofilefromthemodel,which
mayindicatethattheseareseparatefeatures.Edgesharpness
canbeusedtoconstrainthepropertiesofsculptingplanets
(e.g.,T.D.Pearceetal.2024). Thefactthatthepreferred
modelhassmootheredgescouldalsoimplythatother
dynamicalprocessesmightbeinvolvedincreatingthe
eccentricityintheFomalhautdisk,suchasself-gravitation,
particlecollisions,andclose-packingseeninplanetaryringsin
thesolarsystem(S.F.Dermott&C.D.Murray1980). More
detailedN-bodysimulationsareneededtofullyexplorethis.
Theexterior“shoulders”arealsoreminiscentofthehalos
detectedinotherdebrisdisks,suchasHD32297,HD61005,
andq
1
Eri(M.A.MacGregoretal.2018; J.B.Lovelletal.
2021), andtargetsfromtheARKSsurvey(2025,in
preparation,privatecommunication). Thedifferencesinthe
shapeoftheradialprofilesatthetwoansae,includingthe
exteriorshoulders,appearconsistentwithanN-bodysimula-
tionofaneccentricplanetsculptinganexteriordiskpresented
inAppendixB1ofT.D.Pearceetal.(2024), withanarrower
apocenterandanexteriorhalo/shoulder,thoughtheSEansa
profileofFomalhautappearsmoreGaussianthaninthatwork.
Higher-resolutionobservationsinthefuturemaybeableto
resolvejusthowdistinctthesefeaturesare.
FluxfromtheIBintheJWSTMIRIimageandtheextended
exteriorhalocomplicatethewidthmeasurement,andsowe
insteadcitealowerlimitbymeasuringthewidthbetween
wherethefluxradiallyoutwardsfallstothemeanfluxlevel
about25auinteriortothepeakemission.Inaddition,theMIRI
beamsizeis∼1″or∼7.7au,whichismuchlargerthanthe
ALMAbeam.Weplaceupperlimitsof51and31auontheSE
andNWansaeintheJWSTimage,orawidthdifferenceof
about20aubetweentheansae.However,ifweinstead
comparedwherethefluxlevelattheouteredgeoftheSEansa
iscomparabletothefluxbeyondthedimmerNWansa,theSE
widthcouldbeasmuchas20auwiderthanestimatedhere.
Theseestimatesarebettercomparedtothefullwidthat20%
maximummeasurementsof29and21auattheSEandNW
ansaefromtheALMAdata(Section3.1).
Finally,wenotethatrelativelyhighinclinationofthedisk
degradesourabilitytoaccuratelymeasurethedeprojecteddisk
widthawayfromthemajoraxis(andiscompoundedbythe
increaseinnoiseandartifactsinthesesameregionsas
discussedinSection3).
5.2.FluxConstraintsontheIntermediateBelt
WedonotdetecttheIBrevealedbyJWST/MIRI
(A.Gáspáretal.2023) intheALMAmosaicat1.3mm.In
ordertoconstrainthetotalfluxfortheIBat1.3mm,we
estimateboundariesfortheIBusingparametersfrom
A.Gáspáretal.(2023) andourforcedandpropereccentricity
model.A.Gáspáretal.(2023) notethattheorbitalboundaries
oftheIBarenotwell-definedduetothebrightinnerbeltand
thefainterNEansa.Theyusedellipsefittingtoestimatean
innerboundaryofa=83au,e=0.31,andanouterboundary
ofa=104au,e=0.265.TheyalsofoundthattheIBand
innerbeltshadslightlyvaryinginclinationsandposition
angles,butforthesakeofsimplicityweignorethose
differenceshere.Instead,weassumethattheIBhasasimilar
forcedargumentofpericenterastheouterbelt,ω
f
=45°,and
traceouttheorbitsfortheboundariescorrespondingtoall
meanlongitudes,l=[0,2π].Theseboundariesareplottedas
whitedashedlinesinFigures1(c)and(d).
Theapproximatermsnoiseandartifactsinthefull-mosaic
image(Figure1(c))increaseawayfromthepointingcenterat
thediskansae,asdiscussedinSection3.Tomeasurethiseffect,
wecreateanannulusregioninCASAdefinedbytheabove
boundaries,andusetheimstattasktocomputetherms,about
12μJybm
−1
.Wethereforeputa3σupperlimitonthepeakflux
oftheIBat1.33mmof36μJybm
−1
.Theareadenotedbythese
approximateboundariescorrespondston
beams
=121.Wethen
usetheexpression( )/= ×F n nr
ms
lim beams b
eams
toplacea3σ
upperlimitof396μJyontheIBfluxat1.33mm.Using
thesesameboundaries,wecompute56mJyoffluxinthe
JWST25.5μmimage,althoughwenotethatthisestimateis
likelycontaminatedwithfluxfromtheinnerbelt.
M.Sommeretal.(2025) suggesttheIBcouldbeexplained
byP–Rdragalongwithunmodeledfeatures,likeresonant
trappingofsmallgrains,withoutnecessarilyinvokinga
second,collisionally-activedustbelt.FutureSEDmodeling
couldusetheupperlimitstheseobservationsprovidetoclarify
thenatureoftheIB.
5.3.GreatDustCloudDetection
ThespectralwindowcorrespondingtotheC5datacentered
ontheν
21
=230.538GHzCO2–1lineoftheresidual
measurementset(i.e.,thebest-fite
p
continuummodelwas
subtractedoff)wasimagedasacubewithtclean, anda
Keplerianspatio-spectralmaskwasusedinordertodetermine
ifCOemissionfromtheexocometarydebriscouldbedetected
asinL.Matràetal.(2017). Noemissionwasdetected,butthis
resultisconsistentwiththepreviousworkgiventhehigher
resolutionhere.
However,uponreexaminationofthediskmodel-subtracted
C3COdata,wediscoveredmultipleconsecutivechannelsof
emissioncoincidentwiththepositionoftheGDCobject
discussedinA.Gáspáretal.(2023) andG.M.Kennedyetal.
(2023). Theemissionoccursbetween230.1and230.4GHz,
correspondingtoalinewidthof≈300kms
−1
.Fomalhaut’s
systemicradialvelocityisv
r
=6.5kms
−1
(G.A.Gontcharov
2006), whichwouldimplythattheemittinggasismovingwith
velocities>400kms
−1
iftheywerefromtheexpected
230GHzCO2–1line.Giventhedynamicsofdebrisdisks
andtherelativesize,velocity,andlocationofthisfeature,it
seemsimplausiblethatitisconnectedtoafeatureinthe
Fomalhautdebrisdiskandismuchmorelikelytobea
Table3
DiskAnsaeFWHMMeasurements
DataSet SEAnsa NWAnsa
(au) (au)
ALMABand6 16 12e
p
=0 12 13e
p
=0.06 11 12
8
TheAstrophysicalJournalLetters,990:L40(10pp), 2025September10 Chittidietal.
dispersionparameteristhatthefluxfallsoffmoregently,
resultinginwiderlowsurfacebrightnessfeatures.Thiscould
atleastpartiallyexplainthe“shoulders”seeninthedata,and
mayalsobethereasonthismodelisstatisticallypreferredover
theotheronedespitethefactthatitcannotreproducethe
brightnessasymmetryorwidthdifference.These“shoulders”
appearslightlymoredistinctfromthecentralpeakofemission
comparedtothemoresmoothprofilefromthemodel,which
mayindicatethattheseareseparatefeatures.Edgesharpness
canbeusedtoconstrainthepropertiesofsculptingplanets
(e.g.,T.D.Pearceetal.2024). Thefactthatthepreferred
modelhassmootheredgescouldalsoimplythatother
dynamicalprocessesmightbeinvolvedincreatingthe
eccentricityintheFomalhautdisk,suchasself-gravitation,
particlecollisions,andclose-packingseeninplanetaryringsin
thesolarsystem(S.F.Dermott&C.D.Murray1980). More
detailedN-bodysimulationsareneededtofullyexplorethis.
Theexterior“shoulders”arealsoreminiscentofthehalos
detectedinotherdebrisdisks,suchasHD32297,HD61005,
andq
1
Eri(M.A.MacGregoretal.2018; J.B.Lovelletal.
2021), andtargetsfromtheARKSsurvey(2025,in
preparation,privatecommunication). Thedifferencesinthe
shapeoftheradialprofilesatthetwoansae,includingthe
exteriorshoulders,appearconsistentwithanN-bodysimula-
tionofaneccentricplanetsculptinganexteriordiskpresented
inAppendixB1ofT.D.Pearceetal.(2024), withanarrower
apocenterandanexteriorhalo/shoulder,thoughtheSEansa
profileofFomalhautappearsmoreGaussianthaninthatwork.
Higher-resolutionobservationsinthefuturemaybeableto
resolvejusthowdistinctthesefeaturesare.
FluxfromtheIBintheJWSTMIRIimageandtheextended
exteriorhalocomplicatethewidthmeasurement,andsowe
insteadcitealowerlimitbymeasuringthewidthbetween
wherethefluxradiallyoutwardsfallstothemeanfluxlevel
about25auinteriortothepeakemission.Inaddition,theMIRI
beamsizeis∼1″or∼7.7au,whichismuchlargerthanthe
ALMAbeam.Weplaceupperlimitsof51and31auontheSE
andNWansaeintheJWSTimage,orawidthdifferenceof
about20aubetweentheansae.However,ifweinstead
comparedwherethefluxlevelattheouteredgeoftheSEansa
iscomparabletothefluxbeyondthedimmerNWansa,theSE
widthcouldbeasmuchas20auwiderthanestimatedhere.
Theseestimatesarebettercomparedtothefullwidthat20%
maximummeasurementsof29and21auattheSEandNW
ansaefromtheALMAdata(Section3.1).
Finally,wenotethatrelativelyhighinclinationofthedisk
degradesourabilitytoaccuratelymeasurethedeprojecteddisk
widthawayfromthemajoraxis(andiscompoundedbythe
increaseinnoiseandartifactsinthesesameregionsas
discussedinSection3).
5.2.FluxConstraintsontheIntermediateBelt
WedonotdetecttheIBrevealedbyJWST/MIRI
(A.Gáspáretal.2023) intheALMAmosaicat1.3mm.In
ordertoconstrainthetotalfluxfortheIBat1.3mm,we
estimateboundariesfortheIBusingparametersfrom
A.Gáspáretal.(2023) andourforcedandpropereccentricity
model.A.Gáspáretal.(2023) notethattheorbitalboundaries
oftheIBarenotwell-definedduetothebrightinnerbeltand
thefainterNEansa.Theyusedellipsefittingtoestimatean
innerboundaryofa=83au,e=0.31,andanouterboundary
ofa=104au,e=0.265.TheyalsofoundthattheIBand
innerbeltshadslightlyvaryinginclinationsandposition
angles,butforthesakeofsimplicityweignorethose
differenceshere.Instead,weassumethattheIBhasasimilar
forcedargumentofpericenterastheouterbelt,ω
f
=45°,and
traceouttheorbitsfortheboundariescorrespondingtoall
meanlongitudes,l=[0,2π].Theseboundariesareplottedas
whitedashedlinesinFigures1(c)and(d).
Theapproximatermsnoiseandartifactsinthefull-mosaic
image(Figure1(c))increaseawayfromthepointingcenterat
thediskansae,asdiscussedinSection3.Tomeasurethiseffect,
wecreateanannulusregioninCASAdefinedbytheabove
boundaries,andusetheimstattasktocomputetherms,about
12μJybm
−1
.Wethereforeputa3σupperlimitonthepeakflux
oftheIBat1.33mmof36μJybm
−1
.Theareadenotedbythese
approximateboundariescorrespondston
beams
=121.Wethen
usetheexpression( )/= ×F n nr
ms
lim beams b
eams
toplacea3σ
upperlimitof396μJyontheIBfluxat1.33mm.Using
thesesameboundaries,wecompute56mJyoffluxinthe
JWST25.5μmimage,althoughwenotethatthisestimateis
likelycontaminatedwithfluxfromtheinnerbelt.
M.Sommeretal.(2025) suggesttheIBcouldbeexplained
byP–Rdragalongwithunmodeledfeatures,likeresonant
trappingofsmallgrains,withoutnecessarilyinvokinga
second,collisionally-activedustbelt.FutureSEDmodeling
couldusetheupperlimitstheseobservationsprovidetoclarify
thenatureoftheIB.
5.3.GreatDustCloudDetection
ThespectralwindowcorrespondingtotheC5datacentered
ontheν
21
=230.538GHzCO2–1lineoftheresidual
measurementset(i.e.,thebest-fite
p
continuummodelwas
subtractedoff)wasimagedasacubewithtclean, anda
Keplerianspatio-spectralmaskwasusedinordertodetermine
ifCOemissionfromtheexocometarydebriscouldbedetected
asinL.Matràetal.(2017). Noemissionwasdetected,butthis
resultisconsistentwiththepreviousworkgiventhehigher
resolutionhere.
However,uponreexaminationofthediskmodel-subtracted
C3COdata,wediscoveredmultipleconsecutivechannelsof
emissioncoincidentwiththepositionoftheGDCobject
discussedinA.Gáspáretal.(2023) andG.M.Kennedyetal.
(2023). Theemissionoccursbetween230.1and230.4GHz,
correspondingtoalinewidthof≈300kms
−1
.Fomalhaut’s
systemicradialvelocityisv
r
=6.5kms
−1
(G.A.Gontcharov
2006), whichwouldimplythattheemittinggasismovingwith
velocities>400kms
−1
iftheywerefromtheexpected
230GHzCO2–1line.Giventhedynamicsofdebrisdisks
andtherelativesize,velocity,andlocationofthisfeature,it
seemsimplausiblethatitisconnectedtoafeatureinthe
Fomalhautdebrisdiskandismuchmorelikelytobea
Table3
DiskAnsaeFWHMMeasurements
DataSet SEAnsa NWAnsa
(au) (au)
ALMABand6 16 12e
p
=0 12 13e
p
=0.06 11 12
8
TheAstrophysicalJournalLetters,990:L40(10pp), 2025September10 Chittidietal.
yet-unidentifiedspectrallineemissionfromabackground
galaxy.InFigure4,weproduceaMoment-0imageofthe
spectrallinefromaprimary-beam-correctedcube,andusetwo
spatialfilterstoextract1Dsourceandnoisespectra.Noother
sourcesappearinthesechannelsabovethenoiselevel.The
sourcewasnotdetectedintheC5spectralwindowcontaining
the
12
CO(2–1) transition,likelybecausethislocation
correspondstobelowthe20%gainleveloftheprimarybeam
andthusmaybelostinthenoise.
M.Ygoufetal.(2024) fittheavailablephotometryofthe
galaxyusingtemplatespectraofgalaxiesatz=0.80,0.21,
and0.56andconcludedthattheGDCwasanultraluminous
infraredgalaxy.Ontheassumptionthatthislineemanates
fromCO(plausiblythebrightestlineandsomostlikelytobe
detected), weconsideronepossibilitythatthespectrallinewe
identifiedisthe
12
CO(3–2) transitionatν
32
=345.796GHz.
Giventhattheobservedcentralfrequencyofthelineisat
ν
obs
≈230.25GHz,weestimateacorrespondingredshiftof
z=0.502.Itisunclearhowuncertaintheredshiftestimates
fromthefitspresentedinM.Ygoufetal.(2024) are,butthis
couldsupportthez=0.56,NGC6240-analogtheyproposed.
6.Conclusions
Weusedlong-baselineobservationsfromALMACycle5
andshort-baselineobservationsfromCycle3toproducea
0.
57(4.4au)imageoftheFomalhautouterdebrisdiskata
sensitivityof7μJybm
−1
.
1.WegeneratedradialprofilesofthenewALMAimage
andthe25.5μmJWST/MIRI imagefromA.Gáspár
etal.(2023) byregriddingthemtoanR–θgrid.Wethen
measuredtheFWHMoftheradialprofiles,andfound
thattheSEsideofthedisknearpericenteris4auwider
thanattheNWsidenearapocenterwithALMA,and
about20auwiderwithJWST.Wealsoobserveda
21%±1%brighterNWsidecomparedtotheSEsidein
theALMAimagefromthepeakbrightnessoftheradial
profiles.
2.WeperformedMCMCfitsoftwo,particle-baseddisk
modelswithproperandforcedcomponenteccentricities
totheALMAvisibilities.OurBICanalysissuggeststhat
themodelthatincludesae
p
parameterispreferredover
themodelwithoutone,supportingthefindingsin
G.M.Kennedy(2020).
3.Neithermodelisabletoreplicatethe4auwidth
difference,andinsteadhavea1auwiderapocenter.
FuturemodelingshouldtestwhetheraGaussiansemi-
majoraxisdistributionfortheparticlesasopposedtothe
uniformoneemployedherecanalleviatethistension.
4.Neithermodelcouldreproducethe21%±1%brightness
asymmetrynearapocentermeasuredfromtheradial
profileofthedata,suggestingthereisayetmissing
componenttothephysicsunderlyingthedistributionof
materialinthedisk.J.B.Lovelletal.(2025) presenta
modelwithaneccentricitygradientthatcansimulta-
neouslyaccountforthewidthandbrightnessdifference.
5.WedonotdetectanyemissionfromtheIBdiscoveredin
JWST/MIRI imaging,butareabletoplacea3σupper
limitof396μJyforthetotalfluxat1.33mm.
6.WediscoveredaspectrallineinthearchivalCycle3data
centeredatν
obs
≈230.25GHzatthelocationofthe
GDC,redshiftedfromtheexpectedCO(2–1) lineby
morethan400kms
−1
.Thishighvelocitysupportsthe
conclusionthattheobjectisabackgroundgalaxy.We
suggestthatthespectrallinecouldbefromCO(3–2)
emissionatarestfrequencyof345GHz,implyinga
redshiftofz≈0.502.
Fomalhaut’sproximityallowsobservationstoresolveits
structureathigherresolutionthanothersystems,which
continuestomakeitanidealtargettoexploretheearly
evolutionofplanetarysystems.Thesenewdatareveal
variationintheazimuthalstructureoftheouterdiskthat
isnotwell-fittedbycurrenteccentricmodels.Future0
00
5
00
10
00
0
00
5
00
10
00
RA Oset (
00
)
Dec Oset (
00
)
25 AU
0
100
200
300
mJybeam
1
kms
1
1000 500 0 500 1000
v21vr[kms
1
]
500
250
0
250
500
750
F
[mJy]
GDC Spectrum
Noise Spectrum
Figure4.Left:AMoment-0imageoftheGDCobtainedbyintegratingbetween230.1and230.4GHz(173–563kms
−1
channelsrelativetotheexpectedCOline)of
theCycle3residualmeasurementset(i.e.,thebest-fitdiskcontinuummodelhasbeensubtractedoff).Thecyancircledenotestheareaoftheregionusedtoextract
thespectrumofthesource,andtheorange-dashedcircleisdenotesaregionofequivalentsizewithcharacteristicnoise.Right:Theresultingextractedspectraofthe
backgroundgalaxyandnoiseregion.Thesourceemissionisredshiftedby>400kms
−1
fromtheexpectedDoppler-shiftedtransitionof
12
COJ=2–1corresponding
toFomalhaut(v
r
=6.5kms
−1
,G.A.Gontcharov2006), suggestingthesourceisunaffiliatedwiththesystem.Thewidthoftheemissionis≈300kms
−1
,consistent
withvelocitydispersionfromrotatingdiskgalaxies.
9
TheAstrophysicalJournalLetters,990:L40(10pp), 2025September10 Chittidietal.
galaxy.InFigure4,weproduceaMoment-0imageofthe
spectrallinefromaprimary-beam-correctedcube,andusetwo
spatialfilterstoextract1Dsourceandnoisespectra.Noother
sourcesappearinthesechannelsabovethenoiselevel.The
sourcewasnotdetectedintheC5spectralwindowcontaining
the
12
CO(2–1) transition,likelybecausethislocation
correspondstobelowthe20%gainleveloftheprimarybeam
andthusmaybelostinthenoise.
M.Ygoufetal.(2024) fittheavailablephotometryofthe
galaxyusingtemplatespectraofgalaxiesatz=0.80,0.21,
and0.56andconcludedthattheGDCwasanultraluminous
infraredgalaxy.Ontheassumptionthatthislineemanates
fromCO(plausiblythebrightestlineandsomostlikelytobe
detected), weconsideronepossibilitythatthespectrallinewe
identifiedisthe
12
CO(3–2) transitionatν
32
=345.796GHz.
Giventhattheobservedcentralfrequencyofthelineisat
ν
obs
≈230.25GHz,weestimateacorrespondingredshiftof
z=0.502.Itisunclearhowuncertaintheredshiftestimates
fromthefitspresentedinM.Ygoufetal.(2024) are,butthis
couldsupportthez=0.56,NGC6240-analogtheyproposed.
6.Conclusions
Weusedlong-baselineobservationsfromALMACycle5
andshort-baselineobservationsfromCycle3toproducea
0.
57(4.4au)imageoftheFomalhautouterdebrisdiskata
sensitivityof7μJybm
−1
.
1.WegeneratedradialprofilesofthenewALMAimage
andthe25.5μmJWST/MIRI imagefromA.Gáspár
etal.(2023) byregriddingthemtoanR–θgrid.Wethen
measuredtheFWHMoftheradialprofiles,andfound
thattheSEsideofthedisknearpericenteris4auwider
thanattheNWsidenearapocenterwithALMA,and
about20auwiderwithJWST.Wealsoobserveda
21%±1%brighterNWsidecomparedtotheSEsidein
theALMAimagefromthepeakbrightnessoftheradial
profiles.
2.WeperformedMCMCfitsoftwo,particle-baseddisk
modelswithproperandforcedcomponenteccentricities
totheALMAvisibilities.OurBICanalysissuggeststhat
themodelthatincludesae
p
parameterispreferredover
themodelwithoutone,supportingthefindingsin
G.M.Kennedy(2020).
3.Neithermodelisabletoreplicatethe4auwidth
difference,andinsteadhavea1auwiderapocenter.
FuturemodelingshouldtestwhetheraGaussiansemi-
majoraxisdistributionfortheparticlesasopposedtothe
uniformoneemployedherecanalleviatethistension.
4.Neithermodelcouldreproducethe21%±1%brightness
asymmetrynearapocentermeasuredfromtheradial
profileofthedata,suggestingthereisayetmissing
componenttothephysicsunderlyingthedistributionof
materialinthedisk.J.B.Lovelletal.(2025) presenta
modelwithaneccentricitygradientthatcansimulta-
neouslyaccountforthewidthandbrightnessdifference.
5.WedonotdetectanyemissionfromtheIBdiscoveredin
JWST/MIRI imaging,butareabletoplacea3σupper
limitof396μJyforthetotalfluxat1.33mm.
6.WediscoveredaspectrallineinthearchivalCycle3data
centeredatν
obs
≈230.25GHzatthelocationofthe
GDC,redshiftedfromtheexpectedCO(2–1) lineby
morethan400kms
−1
.Thishighvelocitysupportsthe
conclusionthattheobjectisabackgroundgalaxy.We
suggestthatthespectrallinecouldbefromCO(3–2)
emissionatarestfrequencyof345GHz,implyinga
redshiftofz≈0.502.
Fomalhaut’sproximityallowsobservationstoresolveits
structureathigherresolutionthanothersystems,which
continuestomakeitanidealtargettoexploretheearly
evolutionofplanetarysystems.Thesenewdatareveal
variationintheazimuthalstructureoftheouterdiskthat
isnotwell-fittedbycurrenteccentricmodels.Future0
00
5
00
10
00
0
00
5
00
10
00
RA Oset (
00
)
Dec Oset (
00
)
25 AU
0
100
200
300
mJybeam
1
kms
1
1000 500 0 500 1000
v21vr[kms
1
]
500
250
0
250
500
750
F
[mJy]
GDC Spectrum
Noise Spectrum
Figure4.Left:AMoment-0imageoftheGDCobtainedbyintegratingbetween230.1and230.4GHz(173–563kms
−1
channelsrelativetotheexpectedCOline)of
theCycle3residualmeasurementset(i.e.,thebest-fitdiskcontinuummodelhasbeensubtractedoff).Thecyancircledenotestheareaoftheregionusedtoextract
thespectrumofthesource,andtheorange-dashedcircleisdenotesaregionofequivalentsizewithcharacteristicnoise.Right:Theresultingextractedspectraofthe
backgroundgalaxyandnoiseregion.Thesourceemissionisredshiftedby>400kms
−1
fromtheexpectedDoppler-shiftedtransitionof
12
COJ=2–1corresponding
toFomalhaut(v
r
=6.5kms
−1
,G.A.Gontcharov2006), suggestingthesourceisunaffiliatedwiththesystem.Thewidthoftheemissionis≈300kms
−1
,consistent
withvelocitydispersionfromrotatingdiskgalaxies.
9
TheAstrophysicalJournalLetters,990:L40(10pp), 2025September10 Chittidietal.
observationscouldprovidefurtherinsightsintothedisk’s
radialsubstructureonspatialscales,wherewecouldmeasure
itseccentricmorphologyandtestwhetheritisconsistentwith
thearchitecturecharacteristicofsculptingbyaninternal
planet.
Acknowledgments
WeacknowledgethecontributionsofWayneHollandtothe
initialALMAproposalthatresultedinthisletter.Dr.Holland
unfortunatelypassedawaybeforethisworkreachedits
conclusion.
ThislettermakesuseofthefollowingALMAdata:ADS/JAO.
ALMA#2013.1.00486.S, #2015.1.00966.S,and#2017.1.01043.
S.ALMAisapartnershipofESO(representingitsmember
states), NSF(USA) andNINS(Japan), togetherwithNRC
(Canada), MOSTandASIAA(Taiwan), andKASI(Republicof
Korea), incooperationwiththeRepublicofChile.TheJoint
ALMAObservatoryisoperatedbyESO,AUI/NRAO and
NAOJ.TheNationalRadioAstronomyObservatoryisafacilityof
theNationalScienceFoundationoperatedundercooperative
agreementbyAssociatedUniversities,Inc.
ThisworkutilizedtheAlpinehighperformancecomputing
resourceattheUniversityofColoradoBoulder.Alpineis
jointlyfundedbytheUniversityofColoradoBoulder,the
UniversityofColoradoAnschutz,ColoradoStateUniversity,
andtheNationalScienceFoundation(award2201538).
ThisworkutilizedtheBlancacondocomputingresourceat
theUniversityofColoradoBoulder.Blancaisjointlyfunded
bycomputingusersandtheUniversityofColoradoBoulder.
M.A.M.acknowledgessupportforpartofthisresearchfrom
theNationalAeronauticsandSpaceAdministration(NASA)
underawardnumber19-ICAR19_2-0041.
J.B.L.acknowledgestheSmithsonianInstituteforfunding
viaaSubmillimeterArray(SMA) Fellowship.
J.S.C.gratefullyacknowledgesKirkLongforoptimizing
theparticle-basedmodelspresentedinthiswork.
L.M.acknowledgesfundingbytheEuropeanUnionthrough
theE-BEANSERCproject(grantNo.100117693). Viewsand
opinionsexpressedare,however,thoseoftheauthor(s) only
anddonotnecessarilyreflectthoseoftheEuropeanUnionor
theEuropeanResearchCouncilExecutiveAgency.Neitherthe
EuropeanUnionnorthegrantingauthoritycanbeheld
responsibleforthem.
O.P.acknowledgessupportfromtheScienceandTechnol-
ogyFacilitiesCounciloftheUnitedKingdom,grantNo.ST/
T000287/1.
A.M.H.gratefullyacknowledgessupportfromtheNational
ScienceFoundationundergrantNo.ASTR-2307920.
Software:numpy(C.R.Harrisetal.2020), matplotlib
(J.D.Hunter2007), CASA(J.P.McMullinetal.2007),
uvmodelfit (I.Martí-Vidaletal.2014), astropy,
regions, spectral-cube, (AstropyCollaborationetal.
2022), emcee(D.Foreman-Mackeyetal.2013), galario
(M.Tazzarietal.2018), kepler(D.Foreman-Mackeyetal.
2021), scipy(P.Virtanenetal.2020), numba(S.K.Lam
etal.2015).
ORCIDiDs
JayS.Chittidi
aahttps://orcid.org/0000-0002-4985-028X
MeredithA.MacGregor
aahttps://orcid.org/0000-0001-
7891-8143
JoshuaBennettLovell
aahttps://orcid.org/0000-0002-4248-5443
GaspardDuchene
aahttps://orcid.org/0000-0002-5092-6464
MarkWyatt
aahttps://orcid.org/0000-0001-9064-5598
OljaPanic
aahttps://orcid.org/0000-0002-6648-2968
PaulKalas
aahttps://orcid.org/0000-0002-6221-5360
A.MeredithHughes
aahttps://orcid.org/0000-0002-
4803-6200
DavidJ.Wilner
aahttps://orcid.org/0000-0003-1526-7587
GrantM.Kennedy
aahttps://orcid.org/0000-0001-6831-7547
LucaMatrà
aahttps://orcid.org/0000-0003-4705-3188
MichaelP.Fitzgerald
aahttps://orcid.org/0000-0002-
0176-8973
KateY.L.Su
aahttps://orcid.org/0000-0002-3532-5580
References
Acke,B.,Min,M.,Dominik,C.,etal.2012,A&A, 540,A125
AstropyCollaboration,Price-Whelan,A.M.,Lim,P.L.,etal.2022,ApJ,
935,167
Backman,D.E.,&Paresce,F.1993,inProtostarsandPlanetsIII,ed.
E.H.Levy&J.I.Lunine(Tucson,AZ:Univ.ArizonaPress), 1253
Boley,A.C.,Payne,M.J.,Corder,S.,etal.2012,ApJL, 750,L21
Dermott,S.F.,&Murray,C.D.1980,Icar,43,338
Foreman-Mackey,D.,Hogg,D.W.,Lang,D.,&Goodman,J.2013,PASP,
125,306
Foreman-Mackey,D.,Luger,R.,Agol,E.,etal.2021,JOSS, 6,3285
Gáspár,A.,&Rieke,G.2020,PNAS, 117,9712
Gáspár,A.,Wolff,S.G.,Rieke,G.H.,etal.2023,NatAs, 7,790
Gontcharov,G.A.2006,AstL, 32,759
Harris,C.R.,Millman,K.J.,vanderWalt,S.J.,etal.2020,Natur, 585,357
Holland,W.S.,Greaves,J.S.,Zuckerman,B.,etal.1998,Natur, 392,788
Hunter,J.D.2007,CSE,9,90
Kalas,P.,Graham,J.R.,Chiang,E.,etal.2008,Sci,322,1345
Kalas,P.,Graham,J.R.,&Clampin,M.2005,Natur, 435,1067
Kalas,P.,Graham,J.R.,Fitzgerald,M.P.,&Clampin,M.2013,ApJ,775,56
Kass,R.E.,&Raftery,A.E.1995,JASA, 90,773
Kennedy,G.M.2020,RSOS, 7,200063
Kennedy,G.M.,Lovell,J.B.,Kalas,P.,&Fitzgerald,M.P.2023,MNRAS,
524,2698
Kenyon,S.J.,Currie,T.,&Bromley,B.C.2014,ApJ,786,70
Lam,S.K.,Pitrou,A.,&Seibert,S.2015,inProc.SecondWorkshoponthe
LLVMCompilerInfrastructureinHPC,LLVM’15(NewYork:ACM), 1
Lawler,S.M.,Greenstreet,S.,&Gladman,B.2015,ApJL, 802,L20
Lovell,J.B.,Lynch,E.,Chittidi,J.S.,etal.2025,ApJ,990,145
Lovell,J.B.,&Lynch,E.M.2023,MNRAS, 525,L36
Lovell,J.B.,Marino,S.,Wyatt,M.C.,etal.2021,MNRAS, 506,1978
Lynch,E.M.,&Lovell,J.B.2022,MNRAS, 510,2538
MacGregor,M.A.,Hurt,S.A.,Stark,C.C.,etal.2022,ApJL, 933,L1
MacGregor,M.A.,Matrà,L.,Kalas,P.,etal.2017,ApJ,842,8
MacGregor,M.A.,Weinberger,A.J.,Hughes,A.M.,etal.2018,ApJ,
869,75
Mamajek,E.E.2012,ApJL, 754,L20
Marino,S.2019,ApJ,881,84
Marino,S.2022,arXiv:2202.03053
Martí-Vidal,I.,Vlemmings,W.H.T.,Muller,S.,&Casey,S.2014,A&A,
563,A136
Matrà,L.,MacGregor,M.A.,Kalas,P.,etal.2017,ApJ,842,9
McMullin,J.P.,Waters,B.,Schiebel,D.,Young,W.,&Golap,K.2007,in
ASPConf.Ser.376,AstronomicalDataAnalysisSoftwareandSystems
XVI,ed.R.A.Shaw,F.Hill,&D.J.Bell(SanFrancisco,CA:ASP), 127
Pan,M.,Nesvold,E.R.,&Kuchner,M.J.2016,ApJ,832,81
Pearce,T.D.,Krivov,A.V.,Sefilian,A.A.,etal.2024,MNRAS, 527,3876
Sepulveda,A.G.,Matrà,L.,Kennedy,G.M.,etal.2019,ApJ,881,84
Sommer,M.,Wyatt,M.,&Han,Y.2025,MNRAS, 539,439
Stapelfeldt,K.R.,Holmes,E.K.,Chen,C.,etal.2004,ApJS, 154,458
Tazzari,M.,Beaujean,F.,&Testi,L.2018,MNRAS, 476,4527
vanLeeuwen,F.2007,A&A, 474,653
Virtanen,P.,Gommers,R.,Oliphant,T.E.,etal.2020,NatMe, 17,261
White,J.A.,Boley,A.C.,Dent,W.R.F.,Ford,E.B.,&Corder,S.2017,
MNRAS, 466,4201
Wyatt,M.C.,Dermott,S.F.,Telesco,C.M.,etal.1999,ApJ,527,918
Ygouf,M.,Beichman,C.A.,Llop-Sayson,J.,etal.2024,AJ,167,26
10
TheAstrophysicalJournalLetters,990:L40(10pp), 2025September10 Chittidietal.
radialsubstructureonspatialscales,wherewecouldmeasure
itseccentricmorphologyandtestwhetheritisconsistentwith
thearchitecturecharacteristicofsculptingbyaninternal
planet.
Acknowledgments
WeacknowledgethecontributionsofWayneHollandtothe
initialALMAproposalthatresultedinthisletter.Dr.Holland
unfortunatelypassedawaybeforethisworkreachedits
conclusion.
ThislettermakesuseofthefollowingALMAdata:ADS/JAO.
ALMA#2013.1.00486.S, #2015.1.00966.S,and#2017.1.01043.
S.ALMAisapartnershipofESO(representingitsmember
states), NSF(USA) andNINS(Japan), togetherwithNRC
(Canada), MOSTandASIAA(Taiwan), andKASI(Republicof
Korea), incooperationwiththeRepublicofChile.TheJoint
ALMAObservatoryisoperatedbyESO,AUI/NRAO and
NAOJ.TheNationalRadioAstronomyObservatoryisafacilityof
theNationalScienceFoundationoperatedundercooperative
agreementbyAssociatedUniversities,Inc.
ThisworkutilizedtheAlpinehighperformancecomputing
resourceattheUniversityofColoradoBoulder.Alpineis
jointlyfundedbytheUniversityofColoradoBoulder,the
UniversityofColoradoAnschutz,ColoradoStateUniversity,
andtheNationalScienceFoundation(award2201538).
ThisworkutilizedtheBlancacondocomputingresourceat
theUniversityofColoradoBoulder.Blancaisjointlyfunded
bycomputingusersandtheUniversityofColoradoBoulder.
M.A.M.acknowledgessupportforpartofthisresearchfrom
theNationalAeronauticsandSpaceAdministration(NASA)
underawardnumber19-ICAR19_2-0041.
J.B.L.acknowledgestheSmithsonianInstituteforfunding
viaaSubmillimeterArray(SMA) Fellowship.
J.S.C.gratefullyacknowledgesKirkLongforoptimizing
theparticle-basedmodelspresentedinthiswork.
L.M.acknowledgesfundingbytheEuropeanUnionthrough
theE-BEANSERCproject(grantNo.100117693). Viewsand
opinionsexpressedare,however,thoseoftheauthor(s) only
anddonotnecessarilyreflectthoseoftheEuropeanUnionor
theEuropeanResearchCouncilExecutiveAgency.Neitherthe
EuropeanUnionnorthegrantingauthoritycanbeheld
responsibleforthem.
O.P.acknowledgessupportfromtheScienceandTechnol-
ogyFacilitiesCounciloftheUnitedKingdom,grantNo.ST/
T000287/1.
A.M.H.gratefullyacknowledgessupportfromtheNational
ScienceFoundationundergrantNo.ASTR-2307920.
Software:numpy(C.R.Harrisetal.2020), matplotlib
(J.D.Hunter2007), CASA(J.P.McMullinetal.2007),
uvmodelfit (I.Martí-Vidaletal.2014), astropy,
regions, spectral-cube, (AstropyCollaborationetal.
2022), emcee(D.Foreman-Mackeyetal.2013), galario
(M.Tazzarietal.2018), kepler(D.Foreman-Mackeyetal.
2021), scipy(P.Virtanenetal.2020), numba(S.K.Lam
etal.2015).
ORCIDiDs
JayS.Chittidi
aahttps://orcid.org/0000-0002-4985-028X
MeredithA.MacGregor
aahttps://orcid.org/0000-0001-
7891-8143
JoshuaBennettLovell
aahttps://orcid.org/0000-0002-4248-5443
GaspardDuchene
aahttps://orcid.org/0000-0002-5092-6464
MarkWyatt
aahttps://orcid.org/0000-0001-9064-5598
OljaPanic
aahttps://orcid.org/0000-0002-6648-2968
PaulKalas
aahttps://orcid.org/0000-0002-6221-5360
A.MeredithHughes
aahttps://orcid.org/0000-0002-
4803-6200
DavidJ.Wilner
aahttps://orcid.org/0000-0003-1526-7587
GrantM.Kennedy
aahttps://orcid.org/0000-0001-6831-7547
LucaMatrà
aahttps://orcid.org/0000-0003-4705-3188
MichaelP.Fitzgerald
aahttps://orcid.org/0000-0002-
0176-8973
KateY.L.Su
aahttps://orcid.org/0000-0002-3532-5580
References
Acke,B.,Min,M.,Dominik,C.,etal.2012,A&A, 540,A125
AstropyCollaboration,Price-Whelan,A.M.,Lim,P.L.,etal.2022,ApJ,
935,167
Backman,D.E.,&Paresce,F.1993,inProtostarsandPlanetsIII,ed.
E.H.Levy&J.I.Lunine(Tucson,AZ:Univ.ArizonaPress), 1253
Boley,A.C.,Payne,M.J.,Corder,S.,etal.2012,ApJL, 750,L21
Dermott,S.F.,&Murray,C.D.1980,Icar,43,338
Foreman-Mackey,D.,Hogg,D.W.,Lang,D.,&Goodman,J.2013,PASP,
125,306
Foreman-Mackey,D.,Luger,R.,Agol,E.,etal.2021,JOSS, 6,3285
Gáspár,A.,&Rieke,G.2020,PNAS, 117,9712
Gáspár,A.,Wolff,S.G.,Rieke,G.H.,etal.2023,NatAs, 7,790
Gontcharov,G.A.2006,AstL, 32,759
Harris,C.R.,Millman,K.J.,vanderWalt,S.J.,etal.2020,Natur, 585,357
Holland,W.S.,Greaves,J.S.,Zuckerman,B.,etal.1998,Natur, 392,788
Hunter,J.D.2007,CSE,9,90
Kalas,P.,Graham,J.R.,Chiang,E.,etal.2008,Sci,322,1345
Kalas,P.,Graham,J.R.,&Clampin,M.2005,Natur, 435,1067
Kalas,P.,Graham,J.R.,Fitzgerald,M.P.,&Clampin,M.2013,ApJ,775,56
Kass,R.E.,&Raftery,A.E.1995,JASA, 90,773
Kennedy,G.M.2020,RSOS, 7,200063
Kennedy,G.M.,Lovell,J.B.,Kalas,P.,&Fitzgerald,M.P.2023,MNRAS,
524,2698
Kenyon,S.J.,Currie,T.,&Bromley,B.C.2014,ApJ,786,70
Lam,S.K.,Pitrou,A.,&Seibert,S.2015,inProc.SecondWorkshoponthe
LLVMCompilerInfrastructureinHPC,LLVM’15(NewYork:ACM), 1
Lawler,S.M.,Greenstreet,S.,&Gladman,B.2015,ApJL, 802,L20
Lovell,J.B.,Lynch,E.,Chittidi,J.S.,etal.2025,ApJ,990,145
Lovell,J.B.,&Lynch,E.M.2023,MNRAS, 525,L36
Lovell,J.B.,Marino,S.,Wyatt,M.C.,etal.2021,MNRAS, 506,1978
Lynch,E.M.,&Lovell,J.B.2022,MNRAS, 510,2538
MacGregor,M.A.,Hurt,S.A.,Stark,C.C.,etal.2022,ApJL, 933,L1
MacGregor,M.A.,Matrà,L.,Kalas,P.,etal.2017,ApJ,842,8
MacGregor,M.A.,Weinberger,A.J.,Hughes,A.M.,etal.2018,ApJ,
869,75
Mamajek,E.E.2012,ApJL, 754,L20
Marino,S.2019,ApJ,881,84
Marino,S.2022,arXiv:2202.03053
Martí-Vidal,I.,Vlemmings,W.H.T.,Muller,S.,&Casey,S.2014,A&A,
563,A136
Matrà,L.,MacGregor,M.A.,Kalas,P.,etal.2017,ApJ,842,9
McMullin,J.P.,Waters,B.,Schiebel,D.,Young,W.,&Golap,K.2007,in
ASPConf.Ser.376,AstronomicalDataAnalysisSoftwareandSystems
XVI,ed.R.A.Shaw,F.Hill,&D.J.Bell(SanFrancisco,CA:ASP), 127
Pan,M.,Nesvold,E.R.,&Kuchner,M.J.2016,ApJ,832,81
Pearce,T.D.,Krivov,A.V.,Sefilian,A.A.,etal.2024,MNRAS, 527,3876
Sepulveda,A.G.,Matrà,L.,Kennedy,G.M.,etal.2019,ApJ,881,84
Sommer,M.,Wyatt,M.,&Han,Y.2025,MNRAS, 539,439
Stapelfeldt,K.R.,Holmes,E.K.,Chen,C.,etal.2004,ApJS, 154,458
Tazzari,M.,Beaujean,F.,&Testi,L.2018,MNRAS, 476,4527
vanLeeuwen,F.2007,A&A, 474,653
Virtanen,P.,Gommers,R.,Oliphant,T.E.,etal.2020,NatMe, 17,261
White,J.A.,Boley,A.C.,Dent,W.R.F.,Ford,E.B.,&Corder,S.2017,
MNRAS, 466,4201
Wyatt,M.C.,Dermott,S.F.,Telesco,C.M.,etal.1999,ApJ,527,918
Ygouf,M.,Beichman,C.A.,Llop-Sayson,J.,etal.2024,AJ,167,26
10
TheAstrophysicalJournalLetters,990:L40(10pp), 2025September10 Chittidietal.
Tags
Categories
TechnologyDownload
Download Slideshow Get the original presentation fileQuick Actions
Statistics
- Views 315
- Slides 10
- Age 82 days
Related Slideshows
11
8-top-ai-courses-for-customer-support-representatives-in-2025.pptx
JeroenErne2
44 views
10
7-essential-ai-courses-for-call-center-supervisors-in-2025.pptx
JeroenErne2
45 views
13
25-essential-ai-courses-for-user-support-specialists-in-2025.pptx
JeroenErne2
36 views
11
8-essential-ai-courses-for-insurance-customer-service-representatives-in-2025.pptx
JeroenErne2
33 views
21
Know for Certain
DaveSinNM
19 views
17
PPT OPD LES 3ertt4t4tqqqe23e3e3rq2qq232.pptx
novasedanayoga46
23 views