SMC-Las Ex ac ed Pho ome y
T. A. Kucha
1
, G. C. Sloan
2,3
, D. R. Mizuno
1
, Ka hleen E. K aeme
1
, M. L. Boye
2
, Ma in A. T. G oenewegen
4
,
O. C. Jones
5
, F. Kempe
6,7,8
, Iain McDonald
9
, Joana M. Oli ei a
10
, Ma a Sewiło
11,12,13
, Sunda S ini asan
14
,
Jacco Th. an Loon
10
, and Albe Zijls a
15
1
Ins i u e o Scien ific Resea ch, Bos on College, 140 Commonweal h A enue, Ches nu Hill, MA 02467, USA
2
Space Telescope Science Ins i u e, 3700 San Ma in D i e, Bal imo e, MD 21218, USA
3
Depa men o Physics and As onomy, Uni e si y o No h Ca olina, Chapel Hill, NC 27599-3255, USA
4
Koninklijke S e enwach an België, Ringlaan 3, B–1180 B ussels, Belgium
5
UK As onomy Technology Cen e, Royal Obse a o y, Black o d Hill, Edinbu gh, EH9 3HJ, UK
6
Ins i u de Ciències de l’Espai (ICE, CSIC), Can Mag ans, s/n, E-08193 Ce danyola del Vallès, Ba celona, Spain
7
ICREA, Pg. Lluís Companys 23, E-08010 Ba celona, Spain
8
Ins i u d’Es udis Espacials de Ca alunya (IEEC), E-08034 Ba celona, Spain
9
Jod ell Bank Cen e o As ophysics, Uni e si y o Manches e , Manches e , M13 9PL, UK
10
Lenna d-Jones Labo a o ies, Keele Uni e si y, ST5 5BG, UK
11
Exoplane s and S ella As ophysics Labo a o y, NASA Godda d Space Fligh Cen e , G eenbel , MD 20771, USA
12
Depa men o As onomy, Uni e si y o Ma yland, College Pa k, MD 20742, USA
13
Cen e o Resea ch and Explo a ion in Space Science and Technology, NASA Godda d Space Fligh Cen e , G eenbel , MD 20771, USA
14
Ins i u o de Radioas onomía y As o ísica, UNAM, An igua Ca e e a a Pá zcua o 8701, Ex-Hda. San José de la Hue a, Mo elia 58089, Mich., México
15
Jod ell Bank Cen e o As ophysics, The Uni e si y o Manches e , Manches e , M13 9PL, UK
Recei ed 2023 No embe 22; e ised 2024 Feb ua y 2; accep ed 2024 Feb ua y 2; published 2024 Ma ch 11
Abs ac
We p esen poin -sou ce pho ome y om he Spi ze Space Telescope's final su ey o he Small Magellanic
Cloud (SMC). We mapped nea ly 30 deg
2
in wo epochs in 2017, wi h he second ex ending o ea ly 2018 a 3.6
and 4.5 μm using he In a ed A ay Came a. This su ey duplica es he oo p in om he SAGE-SMC p og am in
2008. Toge he , hese su eys co e a nea ly 10 y empo al baseline in he SMC. We pe o med ape u e
pho ome y on he mosaicked maps p oduced om he new da a. We did no use any p io ca alogs as inpu s o he
ex ac o in o de o be sensi i e o any mo ing objec s (e.g., o eg ound b own dwa s)and o he ansien
phenomena (e.g., ca aclysmic a iables o FU O i– ype e up ions). We p oduced a poin -sou ce ca alog wi h high-
confidence sou ces o each epoch as well as a combined-epoch ca alog. Fo each epoch and he combined-epoch
da a, we also p oduced a mo e comple e a chi e wi h lowe -confidence sou ces. All o hese da a p oduc s will be
made a ailable o he communi y a he In a ed Science A chi e.
Unified As onomy Thesau us concep s: Small Magellanic Cloud (1468);In a ed pho ome y (792);Celes ial
objec s ca alogs (212)
1. In oduc ion
The Small Magellanic Cloud (SMC)is a nea by, me al-poo
dwa galaxy. I s dis ance (62.44 ±0.94 kpc; G aczyk e al.
2020)and me allici y (Z=0.1–0.2 Z
e
; e.g., Russell &
Dopi a 1992; Choudhu y e al. 2018)make i an ideal a ge
o s udying he e olu ion o bo h he in e s ella medium and
s a s wi h well-cha ac e ized dis ances in a mo e p imi i e
chemical en i onmen han he Milky Way. Consequen ly, he
SMC has been a ge ed by many su eys co e ing a wide
wa eleng h ange, om X- ays o adio (e.g., Za i sky e al.
2002; Cu i & 2MASS Team 2004; Ka o e al. 2007; Udalski
e al. 2008; I a e al. 2010; Cioni e al. 2011; Go don e al.
2011; Habe l e al. 2012; Meixne e al. 2013; Joseph e al.
2019).
The Spi ze Space Telescope (We ne e al. 2004)co e ed
he SMC in whole o in pa in mul iple epochs. The Spi ze
Su ey o he Small Magellanic Cloud (S
3
MC; Bola o e al.
2007)mapped 2.8 deg
2
in he co e o he galaxy (ou lined in
Figu e 1)in all se en pho ome ic fil e s in he In a ed A ay
Came a (IRAC; 3.6, 4.5, 5.8, and 8.0 μm; Fazio e al. 2004)
and he Mul i-band Imaging Pho ome e o Spi ze (24, 70,
and 160 μm; Rieke e al. 2004). A second Spi ze p og am,
Su eying he Agen s o Galaxy E olu ion in he Tidally
S ipped, Low Me allici y Small Magellanic Cloud (SAGE-
SMC; Go don e al. 2011), ollowed. I used he same fil e s bu
co e ed he en i e galaxy and i s en i onmen , including he
ba , wing, and ail ( hese egions a e labeled in Figu e 1).
Spi ze ’s Las Look a he SMC (o SMC-Las )duplica ed he
sky co e age o SAGE-SMC. Bo h su eyed he same 30 deg
2
,
and bo h obse ed in wo epochs spaced 3–4 mon hs apa .
SMC-Las was comple ed du ing Cycle 13, which ook place
du ing he wa m phase o he Spi ze mission. Thus, he only
bands a ailable we e o he 3.6 and 4.5 μm IRAC fil e s.
The S
3
MC p og am obse ed he co e o he SMC ( he
denses egions o he ba )wi h IRAC in 2005 May. The wo
IRAC epochs o SAGE-SMC ollowed in 2008 June and
Sep embe , and he wo SMC-Las epochs we e ob ained in
2017 (Augus –Sep embe )and om 2017 No embe o 2018
Feb ua y. In be ween SAGE-SMC and SMC-Las , he SAGE-
Va p og am mapped an a ea o 2.9 deg
2
in he co e o he
SMC in ou epochs om 2010 Augus o 2011 June du ing he
wa m Spi ze mission (Riebel e al. 2015). Thus, SMC-Las
ex ends he empo al co e age in he cen e o he SMC by
Spi ze o nine epochs a 3.6 and 4.5 μm co e ing a empo al
baseline o o e 12 y . The minimum co e age in he SAGE-
The As onomical Jou nal, 167:149 (16pp), 2024 Ap il h ps://doi.o g/10.3847/1538-3881/ad2601
© 2024. The Au ho (s). Published by he Ame ican As onomical Socie y.
O iginal con en om his wo k may be used unde he e ms
o he C ea i e Commons A ibu ion 4.0 licence. Any u he
dis ibu ion o his wo k mus main ain a ibu ion o he au ho (s)and he i le
o he wo k, jou nal ci a ion and DOI.
1
SMC oo p in is ou epochs wi h a baseline o o e 9 y .
Table 1lis s he epochs om Spi ze su eys o he SMC.
The Wide-fieldIn a edSu eyExpe imen (WISE; W igh
e al. 2010)has p o ided addi ional co e age o he SMC in wo
epochs in 2010 and, s a ing in 2014, wo epochs pe yea as he
Nea -Ea h Objec WISE-Reac i a ed (NEOWISE-R)mission
(Mainze e al. 2014). WISE scans he en i e sky once e e y 6
mon hs. The fi s epoch in 2010 was ob ained in all ou WISE
fil e s. All subsequen epochs ollowed he loss o c yogens and
include jus he 3.4 and 4.6 μmfil e s. The 2023 elease o he
mul iepoch NEOWISE-R ca alog includes a o al o 20 epochs
om 2010 h ough 2022. While WISE p o ides be e empo al
co e age han he Spi ze su eys, i is no as deep, and i is a
lowe esolu ion (∼6″, as discussed in Sec ion 5).
Sec ion 2desc ibes he mosaics om which he poin -sou ce
ca alogs we e gene a ed. Sec ion 3desc ibes he ex ac ion
Figu e 1. The SMC-Las IRAC 3.6 μm(uppe panel)and 4.5 μm(lowe panel)composi e images combining bo h epochs o obse a ion. The majo mo phological
ea u es o he SMC ( he ba , wing, and ail, he la e being a pa o he Magellanic B idge)a e labeled. Red squa es in he 3.6 μm image indica e he wo es fields
( he spa se field in he wing and he c owded field cen e ed on NGC 346, as defined in Sec ion 5). In he 4.5 μm image, he colo ed ou lines show he su ey a eas o
S
3
MC, SAGE-SMC, and SAGE-Va .
2
The As onomical Jou nal, 167:149 (16pp), 2024 Ap il Kucha e al.
p ocess and de ails how he flux and posi ional unce ain ies
we e calcula ed. Sec ion 4desc ibes he con en s o he ca alogs
and a chi es. Sec ion 5compa es he SMC-Las poin sou ces
o SAGE-SMC and WISE in wo es fields. Sec ion 6
summa izes he SMC-Las p ojec .
2. Da a
The pho ome y p esen ed he e we e ex ac ed om mosaic
images p oduced om he SMC-Las su ey (Mizuno e al.
2022). The mosaics consis o 52 pla es, each 1°.06×1°.06, a
bo h 3.6 and 4.5 μm. These da a co e ∼30 deg
2
in he SMC,
including he main ba o he galaxy, he wing ex ending o he
eas , and he ail ex ending u he in ha di ec ion owa d he
La ge Magellanic Cloud.
Al hough he wo IRAC bands ha e sligh ly di e en
esolu ions, he pla es we e mosaicked o ha e he same pixel
size, 0 6, and esolu ion, ∼2″. The obse a ions we e aken
wi h high dynamic ange exposu es a 0.4 and 10.4 s
in eg a ions o co e bo h ain and b igh sou ces and
minimize sa u a ion in he images. Two comple e maps we e
made o e wo epochs sepa a ed by 70 days. The fi s epoch
was ob ained om 2017 Augus 25 o Sep embe 13 and he
second om 2017 No embe 24 o 2018 Feb ua y 12.
Scheduling logis ics equi ed he ex ension o epoch 2 o
78 days o co e he gaps le in he o iginal scheduling. A hi d
map was gene a ed ha combined obse a ions om bo h
epochs, and hese da a a e he basis o Figu e 1.
3. Sou ce Ex ac ion
We used he open sou ce SEx ac o (Sou ce-Ex ac o
e sion 2.14.2; Be in & A nou s 1996) o pe o m ape u e
pho ome y on ou mosaicked images. The so wa e is flexible
in se ing a wide a ie y o pa ame e s needed o pho ome y.
In gene al, he so wa e calcula es a model o he sky
backg ound and hen p oduces a backg ound-sub ac ed image
on which o pe o m he pho ome y. De ec ions a e hen
pho ome ically ex ac ed, “cleaned”o a i ac s (i.e., eex-
amined o con ibu ions o nea es neighbo s), and hen sa ed
o a sou ce lis . The documen a ion wi h he so wa e explains
ha sou ces ha o e lap pa ially, so ha hey show wo clea
peaks, a e ex ac ed as sepa a e sou ces.
3.1. Ape u e Selec ion
The selec ion o ape u e size depends la gely on he
unce ain ies expec ed o he candida e ape u es. The p ima y
sou ce o unce ain y o he ain e sou ces is he backg ound
noise in he ape u e. Fo pixel- o-pixel noise σ
BG
, he
unce ain y in he in-ape u e flux will be p opo ional o
NBG
s, whe e Nis he numbe o pixels in he ape u e, o
equi alen ly, p opo ional o dσ
BG
, whe e dis he ape u e
diame e . The backg ound unce ain y in he o al flux, hen, is
p opo ional o C(d)dσ
BG
, whe e C(d)is he ape u e co ec ion
o diame e d. We es ima e C(d)a small ape u es om he
3.6 and 4.5 μm poin - esponse unc ion (PRF)images supplied
by he In a ed P ocessing and Analysis Cen e (IPAC).
16
C(d)
ises s eeply below abou 4″diame e , and we find ha he
backg ound unce ain y is minimized below a 3″ape u e
diame e o bo h bands.
The second sou ce o ape u e-dependen unce ain y is wha
we e m “small-ape u e”unce ain y, which a ises om he
unce ain ies in he cen oid de e mina ion om he ex ac o ,
pu ing a ying po ions o a sou ce's PRF inside o ou side he
ape u e. This e ec inc eases wi h dec easing ape u e
diame e , and i also inc eases (in absolu e flux unce ain ies)
wi h flux, because i is a ac ional e ec , bu i is pa ially
mi iga ed by he esul ha he cen oid unce ain y gene ally
dec eases wi h inc easing flux.
Ou app oach is o apply he ape u e ha minimizes he o al
unce ain y o each sou ce indi idually, pa icula ly wi h
espec o hese wo compe ing unce ain ies (o he sou ces o
unce ain y a e ei he independen o o depend only weakly on
he ape u e). As he op imum ape u e will a y wi h
backg ound noise and sou ce flux, we se he sou ce ex ac o
o epo he in-ape u e fluxes wi h a ange o ape u e
diame e s: 4″,5″,6″,8″, and 10″. Fo each ape u e, we
calcula e he o al unce ain y and selec he esul s o he
ape u e ha p oduces he minimum unce ain y. See
Sec ion 3.2 o de ails on he calcula ion o unce ain ies.
One excep ion o his scheme is ha o sou ces wi h FWHM
>25, o elonga ion >1.25, a minimum ape u e-size c i e ion
is applied ega dless o he ape u e wi h he minimum
unce ain y, because he ape u e co ec ions (and he esul ing
unce ain ies)a e inaccu a e o such sou ces.
The ca alog fluxes a e he ex ac o - epo ed in-ape u e
fluxes ( o he op imum ape u e)scaled by he ape u e
co ec ions C(d). The ape u e co ec ion o he 10″ape u e
o each band is in e pola ed om he co ec ion alues shown
in he IRAC Use 's Manual ( hei Table 4.8 p o ides c yogenic
and wa m IRAC ape u e co ec ions). The co ec ions o he
emaining ape u es a e de e mined as he median a io o he
10″in-ape u e fluxes o he in-ape u e fluxes a he o he
ape u es o e all he epoch 1 sou ces be ween 1 and 100 mJy
(∼47,000), hen scaled o he 10″ape u e co ec ion. We
es ima e ha he unce ain y in he ape u e co ec ions epo ed
in he IRAC Use 's Manual, abou 1%–2%, should hold o he
ape u e co ec ions applied o ou da a. Table 2shows he
ape u e co ec ions we apply o he wo bands.
Table 1
Spi ze and WISE Obse a ions o he SMC
Spi ze Su eys o he SMC All-Sky
WISE
b
Name Region
a
A ea
(deg
2
)Epoch Epoch
S
3
MC Co e o
he SMC
c
2.8 2005 May 2010 Ap
SAGE-SMC Full SMC 30 2008 Jun 2010 Oc
SAGE-SMC Full SMC 30 2008 Sep 2014 May
SAGE-Va Co e 2.9 2010 Aug 2014 No
SAGE-Va Co e 2.9 2010 Sep 2015 May
SAGE-Va Co e 2.9 2010 Dec 2015 No
SAGE-Va Co e 2.9 2011 Jun 2016 May
SMC-Las Full SMC 30 2017 Aug 2016 No
SMC-Las Full SMC 30 2017 No 2017 Ap
LLLL2017 No
No es.
a
The S
3
MC, SAGE-SMC, and SAGE-Va su ey a eas a e indica ed in he
4.5 μm image in Figu e 1.
b
WISE epochs con inue om 2018 o p esen , wo pe yea .
c
The “ba ”and pa o he “wing”; see Figu e 1.
16
h ps://i sa.ipac.cal ech.edu/da a/SPITZER/docs/i ac/calib a ionfiles/
ps p /
3
The As onomical Jou nal, 167:149 (16pp), 2024 Ap il Kucha e al.
3.2. Unce ain y in Flux
Wi h all such so wa e packages, i is necessa y o unde -
s and exac ly wha he so wa e is measu ing when i ou pu s a
measu emen . Fo example, he ex ac o es ima es he flux
unce ain y gi en he ape u e size(s). Howe e , he unce ain y
calcula ion, pa icula ly o he pho oelec on coun ing s a is-
ics, assumes a fixed in eg a ion ime o he en i e image. Bu
since he images a e mosaics, hey ha e a ying co e ages.
Also, we ha e subs i u ed 0.4 s da a o he e y b igh sou ces
ha sa u a ed he de ec o s in he 10.4 s in eg a ion. Fu he -
mo e, he Spi ze p ocessing pipeline o he aw da a includes
addi ional unce ain ies, pa icula ly o b igh sou ces, such as
unce ain ies associa ed wi h he nonlinea i y co ec ion and
co ec able sa u a ions. Fo hese easons, we ha e chosen o
calcula e unce ain ies sepa a ely om he ex ac o ou pu .
Fo he ape u e pho ome y, we iden i y ou sou ces o
unce ain y om he pixel alues hemsel es: (1)unce ain y in
he backg ound sub ac ed om he da a in he ape u e, (2)
unce ain y in he ape u e in eg a ion om pixel- o-pixel ms
in he ape u e, (3)pho oelec on coun ing unce ain y and
o he unce ain ies o b igh sou ces, and (4) o small
ape u es, unce ain y in he e ec i e ape u e co ec ion due
o unce ain ies in he posi ioning o he cen oid o he sou ce.
Fo ain sou ces (below abou 0.1 mJy), he e is also an
unce ain y associa ed wi h he co ec ion o a small back-
g ound-le el e o .
3.2.1. Backg ound-le el Unce ain y (σ
LEV
)
The sou ce ex ac o p oduces a backg ound ms map. Wi h he
pa ame e s o ou ex ac ions, his map is gene a ed by
de e mining he ms in 14 ×14 pixel bins, wi h some agg essi e
ou lie dele ion. The ms alues o each bin a e smoo hed o e 10
bins and hen in e pola ed o gi e an ms es ima e o each pixel.
While he ms is o mally unique o each pixel, i is basically he
a e age backg ound ms in he su ounding 80″×80″ egion
(10 pixels ×14 bins ×06pixel
−1
).
Fo each sou ce, he backg ound ms σ
BG
is de e mined om
he ms map a he sou ce loca ion. The ex ac o epo s he
backg ound flux sub ac ed o he selec ed ape u e, which is
con e ed o a backg ound b igh ness le el by di iding by he
ape u e a ea. The backg ound le el is de e mined o each
ape u e in a 24 pixel wide “squa e annulus”su ounding he
ape u e, o oughly a 48 ×48 box o pixels. Fo a fla
backg ound, σ
BG
is simply he pixel- o-pixel noise, and he
backg ound-le el unce ain y would be app oxima ely σ
BG
/48.
Bu i σ
BG
con ains significan backg ound a ia ions on a
leng h scale la ge han he sou ce size bu smalle han
14 pixels, hen he backg ound unce ain y would be close o
σ
BG
i sel . We ha e no p ac ical way o di ec ly measu e
backg ound a ia ions a ound a sou ce. Howe e , as an
app oxima ion, we use he backg ound b igh ness alues o
each ape u e: he ms o he le els o e he di e en ape u es
is aken, and he g ea e o his ms (deno ed BG
s
¢)o σ
BG
/48 is
used as he backg ound-le el unce ain y.
Fo each ape u e, hen, he es ima ed backg ound-le el
unce ain y is scaled by he numbe No pixels in he ape u e
o gi e he unce ain y in he in-ape u e flux (in in eg a ed MJy
s
−1
pixel uni s), hen scaled by he solid angle Ao he mosaic
pixels o gi e he in-ape u e flux unce ain y in mJy
(A=8.461595 ×10
−12
s imes he 10
9
MJy- o-mJy con e -
sion). This esul is scaled by he ape u e co ec ion C
k
o
ape u e k(Table 2) o gi e he unce ain y in he o al flux.
This unce ain y is adjus ed by wo addi ional ac o s. Fi s ,
he alues in he backg ound ms images a e smalle han he
expec ed alues o he backg ound ms in he indi idual IRAC
ames di ided by he squa e oo o he co e age. Thus, he
mosaicking p ocess p oduces some amoun o e ec i e
smoo hing. This smoo hing educes he measu ed pixel- o-
pixel ms bu should no a ec he backg ound unce ain ies
o e ex ended egions, so we scaled he backg ound ms alues
upwa d by 20% o app oxima ely ma ch he expec ed
unsmoo hed ms in he mosaics. Second, he IRAC ames
ha e a pixel size o 1 2, while he mosaics ha e 0 6 pixels, so
gene a ing he mosaics e ec i ely ebins he da a by a ac o o
2×2. Fo he unce ain y calcula ions o e ensembles o
pixels, his ebinning inco ec ly inc eases he
n
1
2
- ms benefi
o sums and a e ages by a ac o o 2, so we scale he
calcula ed unce ain ies by 2 o accoun o he ebinning.
The backg ound-le el unce ain y (in mJy)is hen
( )()()()()NAC48 1.2 2.0 .
kLEV BG
ss=
Fo he case whe e he empi ical backg ound ms BG
s
¢is
used, he unce ain y is
()()()NAC
kLEV BG
ss=¢
because BG
s
¢is a di ec measu e (albei an app oxima ion)o
he backg ound-le el unce ain y wi hou he smoo hing and
ebinning adjus men s.
3.2.2. Ape u e Flux Unce ain y Due o Backg ound Noise (σ
SUM
)
The backg ound unce ain y σ
SUM
in he ape u e in eg a ion
is also de e mined om he ex ac o backg ound ms image.
We ake he alue o he backg ound ms σ
BG
a he sou ce
loca ion as cons an o e he ape u e. Wi h his assump ion,
he unce ain y o he sum o image pixel alues o e he
ape u e is
N
B
G
1
2s, whe e Nis he numbe o mosaic pixels in
he ape u e. Wi h he con e sion o mJy and applying he
adjus men s o smoo hing and ebinning, we ha e
()()()()NAC1.2 2.0 .
kSUM BG
1
2
ss=
3.2.3. Pho oelec on Coun ing Unce ain y (σ
PH
)
The pho oelec on coun ing unce ain y (and ela ed b igh -
sou ce unce ain ies)is de e mined om he unce ain y images
gene a ed by Spi ze 's Mopex mosaicking so wa e (Mako oz
& Khan 2005). These images con ain he o al unce ain y o
each mosaic pixel (in MJy s
−1
)as es ima ed om he ini ial
p ocessing o he aw IRAC ames. While hese unce ain y
images con ain backg ound unce ain y alues, he alues do
no accu a ely ep esen he pixel- o-pixel ms alues in ou
images (Mizuno e al. 2022)and also include o e all le el
unce ain y es ima es om he da k sub ac ion, which do no
a ec he unce ain y o he ape u e pho ome y. Fo each
Table 2
IRAC Ape u e Pho ome y Co ec ions
Band Ape u e Size
10″8″6″5″4″3″
3.6 μm 1.070 1.100 1.148 1.205 1.329 1.649
4.5 μm 1.078 1.100 1.141 1.211 1.357 1.685
4
The As onomical Jou nal, 167:149 (16pp), 2024 Ap il Kucha e al.
sou ce, we sub ac he “backg ound”o hese unce ain y
images by aking a 12 ×12 box o pixels su ounding he
sou ce and sub ac ing he median o he pe ime e o he box
as he backg ound. Any esul ing nega i e- alue pixels a e
eplaced wi h ze os. The pho oelec on coun ing unce ain y
σ
PH
is hen
⎜⎟
⎛
⎝
⎞
⎠()()()pAC2.0 ,
i
ikPH 2
1
2
å
s=
whe e he summa ion is o e he mosaic pixels in ape u e k
and p
i
a e he backg ound-sub ac ed pixel alues o he
unce ain y image. No e ha he mosaic smoo hing desc ibed
abo e does a ec he unce ain y image esul s bu no in a
s aigh o wa d way, and i is igno ed in he p esen calcula ion.
3.2.4. Small-ape u e Unce ain y (σ
SA
)
Fo small ape u es, unce ain y in he loca ion o he
cen oid o he sou ce p oduces addi ional unce ain y in he
calcula ed flux due o a ia ion in he ac ion o he flux
con ained wi hin he ape u e depending on he accu acy o he
cen oid. We ha e modeled his small-ape u e e ec wi h he
PRFs supplied by he Spi ze Science Cen e . Fo each band,
he a e age o he 25 images (sampling he ocal plane)is
aken, wi h he backg ound sub ac ion and scaling se o ma ch
he ape u e co ec ions supplied in he IRAC Use ’s Manual
(Table 4.8). A Mon e Ca lo simula ion o he cen oid
unce ain ies was hen pe o med o each ape u e, wi h he
cen e o he ape u e shi ed by andom amoun s in xand y
( ela i e o he PRF cen e )wi h a specified ms o simula e he
cen oid accu acy, and he flux in he ape u e was hen
measu ed. The esul ing sca e in in-ape u e fluxes is a model
o he ac ional unce ain y o he sou ce flux as a unc ion o
he ape u e size and he cen oid ms unce ain y.
Fo each sou ce, he cen oid unce ain y is es ima ed om
an empi ical model (see Sec ion 3.3 below)and hen applied o
he Mon e Ca lo model o es ima e he ac ional flux
unce ain y o each ape u e o ha sou ce. These ac ional
unce ain ies a e hen scaled by he ape u e-co ec ed fluxes (in
mJy) o gi e he unce ain y σ
SA
o each ape u e. This
unce ain y is a maximum o he ain es sou ces in he ca alog,
a ound 5%–10%, and is a pe cen o less abo e abou 0.1 mJy.
3.2.5. Ins umen al Unce ain y (σ
INST
)
The IRAC Use 's Manual (IRAC Ins umen & Ins umen
Suppo Teams 2021)desc ibes wo sou ces o ins umen al
unce ain y. The fi s is he “A ay Loca ion”unce ain y and is
due o he fla -field co ec ion being pe o med wi h he
zodiacal backg ound, and so he e will be a ia ions wi h
di e en sou ce spec a. IPAC supplies a co ec ion image ( o
each band) o a Vega spec um. We use hese o es ima e he
esul ing expec ed sca e in flux alues o ou sou ces by
calcula ing he ms o he co ec ion alues in hese images.
This is an es ima e o he unce ain y o a single co e age.
Because ou da a a e mosaicked om se e al IRAC ames,
his unce ain y is adjus ed by di iding by he squa e oo o he
co e age o ha sou ce.
The second sou ce is he “Pixel Phase”unce ain y, which
a ises om di e en esponses o he image pixels depending
on he p ecise loca ion o he sou ce cen oid on he pixel.
IPAC supplies IDL so wa e o calcula e a co ec ion o he
loca ion o he sou ce wi hin a pixel.
17
This co ec ion is
calcula ed o a g id o loca ion alues on he pixel, and he
expec ed unce ain y is es ima ed om he ms o he co ec ion
alues and again scaled o each sou ce by he squa e oo o
he co e age.
These wo e ec s di e o each band bu oge he gi e an
expec ed unce ain y o abou 2.4% o bo h bands. Howe e ,
hese fixed-pe cen age unce ain ies ha e been adjus ed o
ma ch an empi ical analysis o he unce ain ies. See
Sec ion 3.2.7 below.
3.2.6. Low-flux Backg ound-le el Unce ain y (σ
Δ
)
Sys ema ic a ia ions in he a ios be ween he fluxes
measu ed a he a ious ape u es (4″,5″,6″, and 8″, compa ed
wi h 10″) o sou ces wi h fluxes below ≈0.3 mJy a e
consis en wi h a small backg ound-le el e o , inc easing wi h
dec easing flux. Figu e 2shows he esul ing modeled flux
co ec ions a he h ee smalles ape u es ( he 4″ape u e was
used o mos o he sou ces in his flux ange). The le el e o
is posi i e, p oducing a flux defici , and so he flux co ec ion is
also posi i e. No e ha he backg ound e o seems o be a
Figu e 2. Modeled flux co ec ions due o backg ound-le el e o o he h ee smalles ape u es and o he 3.6 μm band. The esul s o he 4.5 μm band a e simila .
17
h ps://i sa.ipac.cal ech.edu/da a/SPITZER/docs/i ac/calib a ionfiles/
pixelphase/pixel_phase_co ec _gauss.p o
5
The As onomical Jou nal, 167:149 (16pp), 2024 Ap il Kucha e al.

small cons an abo e abou 0.3 mJy, bu we a e omi ing he
co ec ion when i d ops below abou 1%. I we deno e he
co ec ion as Δ
k
( ), o ape u e kand flux , hen he
associa ed unce ain y o he flux co ec ion is es ima ed as
()d
d ,
k
ss=D
D
whe e σis he flux unce ain y o he sou ce om all o he
causes.
3.2.7. To al Unce ain y and Empi ical Analysis
These unce ain ies a e s a is ically independen (wi h he
possible excep ion o σ
Δ
, as we do no know he mechanism
p oducing he unde lying e o , and while σ
LEV
and σ
SUM
bo h
depend on he backg ound noise, any associa ed e o s a e
de i ed om sepa a e ensembles o pixels). The o al
unce ain y o a sou ce is hen
.
2LEV
2
SUM
2
PH
2
SA
2
INST
22
ss s sss s= + +++ +
D
The sepa a e unce ain ies a e also s a is ically independen
ac oss he wo bands. The e m σ
LEV
could ha e some
co ela ion i o a gi en sou ce ha unce ain y is domina ed
by s uc u ed backg ound, which would likely be simila in
bo h bands, bu σ
LEV
is ne e he dominan e m in he o al
unce ain y.
The calcula ed unce ain ies can be checked empi ically wi h
he sou ce ma ches be ween he wo epochs. No ing ha he
backg ound noise le els a he IRAC ame le el o a gi en
band a e app oxima ely he same be ween he wo epochs, and
conside ing only sou ces wi h wo co e ages in he mapping
scans, he flux unce ain ies o a gi en sou ce be ween he wo
epochs should be app oxima ely he same. The flux di e ences
be ween he epoch-ma ched sou ces, hen, should in he
agg ega e eflec he ac ual unce ain ies. Specifically, he ms
o he flux di e ences o an ensemble o sou ces in a na ow
flux ange ( o which he unce ain ies should also gene ally
all in a na ow ange)should be app oxima ely 2 imes he
median unce ain y in he ensemble.
Figu e 3shows he 12
-scaled flux-di e ence ms alues
o he epoch-ma ched sou ces o e a se o flux bins o bo h
bands, he e limi ed o he co e age =2 sou ces (filled blue
squa es). The c i e ion o inclusion in a flux bin is ha he
mean o he epoch 1 and epoch 2 fluxes o a gi en sou ce alls
wi hin he flux bin. Also shown a e he median o mal σ(as
calcula ed abo e)in each bin o he epoch 1 sou ce (open ed
squa es). In he 3.6 μm band in pa icula he e is clea ly a
disc epancy be ween he o mal and empi ical unce ain y
alues, he o mal being abou a ac o o 2 oo small abo e a
ew mJy, bu he e is also a defici in he 4.5 μm band o abou
20%–30% (in bo h cases igno ing he small numbe s a is ics a
he e y high fluxes). Abo e a ew mJy, he only significan
con ibu o o he unce ain y is he ins umen al unce ain y
(Sec ion 3.2.5), so he disc epancy may be a ibu ed o an
inaccu a e accoun ing o ha componen . Wha e e he cause,
we ha e a bi a ily adjus ed he ins umen al unce ain y o
app oxima ely ma ch he empi ical esul s, om 2.4% in bo h
bands o 4.2% and 2.8% in he 3.6 and 4.5 μm bands,
espec i ely.
Figu e 3. Empi ical e o es ima e de i ed om ma ching sou ces ac oss epochs, calcula ed o e flux bins (filled blue squa es). Fo each bin, he ms o he c oss-
epoch flux di e ences is calcula ed and di ided by 2. The open ed squa es a e he median nominal σ o mal unce ain y in he bin, and he filled ed squa es a e he
σ alues a e he adjus men desc ibed in he ex .
6
The As onomical Jou nal, 167:149 (16pp), 2024 Ap il Kucha e al.
C owding o sou ces was no conside ed in he unce ain y
calcula ions. When ano he de ec ed sou ce ell wi hin a gi en
sou ce's ape u e, ha o he sou ce was masked (pixels se o 0
a e backg ound sub ac ion), so ha o a fi s app oxima ion,
he nea by sou ce does no a ec he gi en sou ce's flux o
unce ain y. Howe e , o sou ces wi h a e y close de ec ed
neighbo , less han 3″o so, we see e idence ha he masking
does educe he measu ed flux, al hough his e ec is no
quan ified and is no accoun ed o in he epo ed unce ain y.
Fo he mos “poin like”sou ces (below specified h esholds
o FWHM and elonga ion), he dominan unce ain ies a e he
in-ape u e backg ound noise (σ
SUM
) o fluxes below abou
0.2 mJy. Fo fluxes abo e 0.2 mJy, he ins umen al unce ain y
(σ
INST
)domina es. As hese a e he only wo ape u e-
dependen unce ain ies, o he poin like sou ces, he flux is
he p ima y de e minan o he ape u e. Fo sou ces ha a e
less poin like, he ape u e is usually inc eased om he o mal
unce ain y minimum, la gely because he ape u e co ec ion is
likely o be less accu a e o such sou ces, and a la ge ape u e
should educe he associa ed e o .
3.3. Unce ain y in Posi ion
The ex ac o de e mines he posi ion o he sou ce om he
pixel posi ion o he emission cen oid a e backg ound
sub ac ion, which can in oduce unce ain ies as he ex ac o
subdi ides pixels du ing his p ocess. The accu acy o he
measu ed coo dina es o he sou ces, i.e., he cen oid
unce ain y in he sou ce ex ac ions, is es ima ed wi h he
sou ce ma ches be ween he wo epochs. Fo each band and he
R.A. and decl. sepa a ely, he posi ional di e ences be ween
he ma ched sou ces a e aken, and he ms o he di e ences
a e calcula ed o e flux bins. I we assume ha he posi ional
unce ain ies o gi en sou ces will be equi alen o bo h
epochs (as bo h epochs ha e simila co e age and noise
cha ac e is ics), hen he ms o he di e ences will be
app oxima ely 2 imes he posi ional unce ain y o he
indi idual sou ces ( o sou ces wi h simila unce ain ies).We
find ha he posi ional unce ain ies a e flux-dependen and a e
essen ially equi alen o bo h bands in bo h R.A. and decl.
Figu e 4shows cen oid unce ain ies calcula ed in flux bins.
Figu e 4includes an exponen ial unc ion fi ed o he da a o
he o m
() ae c,
b
CENT
s=+
-
whe e is he log o he sou ce flux. This unc ion is he
cen oid unce ain y used o calcula e he small-ape u e
unce ain y (Sec ion 3.2.4).
In p inciple, hese esul s include he as ome ic unce ain y
in he images as well as he cen oid unce ain y specifically.
Howe e , hese da a a e domina ed by sou ces in egions o
which he e a e hund eds o sou ces used o he as ome ic
co ec ions o each IRAC ame (using ma ches o he Two
Mic on All Sky Su ey, 2MASS). The IRAC as ome y is
co ec ed (wi h a ansla ional adjus men )by ma ching he
measu ed cen oids o he sou ces in he IRAC ame o
2MASS sou ces (see Appendix Band Mizuno e al. 2022 o
de ails). Fo ha p ocedu e, he cen oid accu acy was
app oxima ely 0 5. In he c owded fields, a ypical da a ame
includes abou 400 sou ces and wo co e ages, so he
as ome ic accu acy ( ela i e o he mean 2MASS as ome y
o e he IRAC ame)is abou 0 5/20/1.4 =0 018, gene ally
a small componen o he abo e esul s. This es ima e assumes
pu ely ansla ional unce ain ies in he IRAC as ome y
ela i e o he 2MASS as ome y o any gi en da a ame.
We do, howe e , see a small field- o a ional e o o abou 1′in
he 3.6 μm a ay as ome y in all he da a (see Appendix B).
This e m gi es an e ec i e addi ional sca e o abou 0 03 in
he 3.6 μm coo dina es, which is implici ly included in he
σ
CENT
calcula ion, al hough i is small enough ha di e ences
in posi ional accu acy be ween bands can be igno ed.
Fo he posi ional accu acy epo ed in he ca alog, an
adjus men o he as ome ic accu acy in he spa se fields is
included due o less accu a e ansla ional co ec ions wi h
ewe sou ces. This is app oxima ely 0 5/NC , whe e Nis he
mean numbe o 2MASS sou ces used o he as ome ic
co ec ion and Cis he co e age numbe , so he o al posi ional
Figu e 4. Posi ional unce ain y (single axis)calcula ed om c oss-epoch ma ches o sou ces. Fo each flux bin, he posi ional di e ences be ween epochs in bo h
R.A. and decl. and o bo h bands a e aken. The plo ed alues a e he ms o hese bin ensembles di ided by 2. The epoch 1 3.6 μm sou ce is used o he flux
indexing.
7
The As onomical Jou nal, 167:149 (16pp), 2024 Ap il Kucha e al.
unce ain y o a sou ce is
()NC0. 5 .
POS
2
CENT
22
ss=+
Again, his is ela i e o he local mean 2MASS as ome y
o e an IRAC ame.
This posi ional unce ain y is o a single sou ce de ec ion
along a single axis (R.A. o decl.). Fo sou ces de ec ed in bo h
bands, he epo ed R.A. and decl. a e he unce ain y-weigh ed
a e ages o he coo dina es de e mined o each o he bands,
which p o ides oughly a 2benefi in he posi ional accu acy.
The associa ed single-axis unce ain y o he weigh ed a e ages
is hen scaled upwa d by 2 o gi e a adial unce ain y.
The epo ed posi ional unce ain ies a e andom de ia ions
om he nominal 2MASS as ome y. The absolu e unce ain y
depends on he accu acy o he 2MASS as ome y in he SMC
a he epoch o he SMC-Las obse a ions. The global p ope
mo ion o he SMC is app oxima ely 1.5 mas y
−1
(Zi ick e al.
2018; see also Niede ho e e al. 2021), so o he ∼20 y
be ween he su eys, we expec ∼30 mas o posi ional shi .
Also, Zi ick e al. (2018)obse e p ope mo ions o up o
abou 0.3 mas y
−1
ela i e o he global SMC p ope mo ion in
each o hei ∼35 sepa a e analysis egions, so his can
con ibu e up o ∼6 mas u he shi (al hough Niede ho e
e al. 2021 epo somewha highe isola ed local p ope
mo ions). In addi ion, we ha e some e idence o a sys ema ic
disc epancy o abou 0 03 in he measu ed coo dina es o
sou ces be ween ou wo epochs, possibly due o a small
esidual o se e o in he poin ing co ec ions, which esul s in
a coo dina e shi a di e en oll angles. The ne sys ema ic
e o we expec is he e o e a mos ∼007. The b igh es
objec s in he su ey, howe e , a e mos likely o eg ound
objec s and will ha e much la ge p ope mo ions. Cau ion
should be applied wi h hese objec s when ying o ma ch
ac oss su eys and/o epochs.
4. Da a P oduc s
4.1. Ca alogs and A chi es
Following he p ac ice o he SAGE p ojec (Meixne e al.
2006), SAGE-SMC (Go don e al. 2011), and many o he
Spi ze -based sou ce lis s, we ha e sepa a ed he final lis s o
ex ac ions in a file con aining he mo e eliable sou ces,
e e ed o as he “ca alog,”and a mo e comple e “a chi e,”
which con ains he ca alog sou ces as well as less eliable
ex ac ions. Toge he , hese sou ce lis s balance comple eness
and eliabili y.
Each ca alog and a chi e is published in h ee e sions o a
o al o six files. We applied he sou ce ex ac o sepa a ely o
he mosaics om each obse a ional epoch, and we also
ex ac ed sou ces om he combined-epoch mosaics.
The sou ce ex ac o gene a ed measu ed flux densi ies and
p elimina y es ima ed unce ain ies o all o he selec ed
ape u es. Once he o al unce ain ies we e calcula ed as
desc ibed abo e, he ape u e wi h minimum unce ain y was
ini ially selec ed o he sou ce lis . Fo sou ces de ia ing om
“poin like”(FWHM >25 o elonga ion >1.25), a la ge
ape u e was gene ally selec ed depending on he pa icula
alues o FWHM and elonga ion. C owding has no been
conside ed o he ape u e selec ion. The sou ce lis s include
he selec ed ape u e.
Fo he poin like sou ces, below abou 2.0 mJy, he in-
ape u e backg ound noise unce ain y (σ
SUM
)domina es, and
he smalles ape u e (4″)is usually selec ed. Abo e 2.0 mJy,
he “small-ape u e”unce ain y (σ
SA
)begins o domina e o e
σ
SUM
(as i is a ac ional unce ain y), and he selec ed ape u e
gene ally inc eases wi h inc easing flux. Abo e ∼0.3 mJy, he
ins umen al unce ain y (σ
INST
)domina es he o al e o , bu
his unce ain y does no depend on ape u e.
The c i e ia de e mining whe he sou ces belonged in he
ca alogs o he a chi es di e significan ly om p e ious
Spi ze -based su eys such as SAGE-SMC. Sou ces we e
included in he ca alog only i he signal- o-noise a io (S/N)
was 3 o g ea e in each band. To main ain independence
be ween he obse a ional epochs, sou ces a e no ma ched
be ween he epochs in ei he he ca alog o he a chi e. In a
gi en epoch, sou ces we e ma ched ac oss bands i hey we e
wi hin 1 6 o each o he . Al hough mos sou ces in he a chi e
ha e been ma ched ac oss bands, ha was no a necessa y
c i e ion o be in he a chi e. The e o e, i is possible o an
S/N=3 ex ac ion in IRAC band 1 o be lis ed in he a chi e
wi hou a co esponding band 2 de ec ion, and ice e sa.
We limi ed he size and shape o he sou ces o cull ex ended
sou ces and any emaining a i ac s in he images as well as
alse de ec ions. The ex ac o uses he dis ibu ion o he
sou ce pixels o calcula e he shape o he sou ce om he
second-o de momen s o he emission. These momen s define
he ellip ici y o he sou ce. We p ima ily used he FWHM and
elonga ion ( a io o axes). Fo he ca alog, we limi ed he size
and shape o sou ce o emo e non-poin sou ces. Sou ces in
he ca alogs had o ha e FWHM 3″and elonga ions 2.
These c i e ia we e elaxed o he a chi e, wi h FWHM 5″
and elonga ion 3.5. Thus, sou ce ypes ha can be sligh ly
ex ended such as young s ella objec s a e mo e likely o appea
in he a chi e han he ca alog (see also he discussion by
Sewiło e al. 2013).
We applied flux c i e ia o bo h he ca alog and a chi e o
he h ee epochs. Fain e sou ces (<0.1 mJy)nea b igh e ones
(>1 mJy) ended o be spu ious de ec ions i hey we e wi hin
12″o each o he (e.g., om di ac ion spikes o Ai y ings).
The e o e, we emo ed hese ain sou ces om bo h he
ca alog and a chi e. This s ep culled abou 700 sou ces.
Finally, we mo ed 19 sou ces ha we e b igh e han 3 Jy a
3.6 μm o he a chi e. These objec s a e badly sa u a ed e en in
he sho e in eg a ion imes and hus ha e less eliable flux
densi ies han equi ed o inclusion in he ca alog. The
appendix p o ides a able wi h hese sou ces.
4.2. Modified Julian Da e
The ca alogs and a chi es con ain a column gi ing he mean
Modified Julian Da e (MJD)o he obse a ion, along wi h
maximum and minimum obse a ion imes. These alues a e
based on he da a o a gi en epoch and a e p o ided sepa a ely
o 3.6 and 4.5 μm. The combined-epoch da a also include he
mean MJD, calcula ed in he same manne . The anges in imes
a e wide han hose in a single-epoch ca alog, since he s a s
o he wo epochs a e sepa a ed by 70 days and epoch 2 was
obse ed o e a pe iod o 78 days.
4.3. Comple eness
Figu e 5shows he numbe o sou ces binned by magni ude
o he wo epochs and he combined-epoch ca alog o bo h
IRAC bands. To es ima e comple eness, we ex end he linea
inc ease in he log o he numbe o sou ces as a unc ion o
8
The As onomical Jou nal, 167:149 (16pp), 2024 Ap il Kucha e al.
magni ude and epo he magni ude whe e he cumula i e
numbe o sou ces coun ed eaches 95% and 90% o he
expec ed o al. Table 3gi es he esul s. The imp o ed S/Nin
he mosaics o he combined epochs esul s in significan ly
be e comple eness es ima es bu a he cos o empo al
p ecision.
Table 4compa es SMC-Las o SAGE-SMC, showing he
numbe o sou ces and minimum and maximum flux densi ies
and magni udes o he a ious i e a ions o he ca alogs and
a chi es. He e and in he ollowing sec ions in his pape ,
compa isons o SAGE-SMC a e based on he SAGE-SMC
IRAC “Single F ame +Mosaic Pho ome y”Ca alog and
A chi e, 1.5,
18
which a e based on pho ome y ex ac ed om
he combined epoch 1 and epoch 2 mosaics (Go don e al.
2011). Fo bo h SMC-Las and SAGE-SMC, he a chi e
includes all sou ces in he ca alog. The sou ce coun s e eal a
di e ence in how he ca alogs and a chi es we e sepa a ed in
he wo su eys. In SAGE-SMC, only abou 10% o he
sou ces a e unique o he a chi e, while in SMC-Las , a mo e
conse a i e app oach has led o mo e sou ces unique o he
a chi e han a e in he ca alog. The o al sou ce coun s o he
a chi es in bo h su eys a e compa able.
5. Compa ing SMC-Las o O he Mid-in a ed Su eys
As ou lined in he In oduc ion, mul iple su eys wi h
Spi ze and WISE ha e co e ed he SMC. The WISE mission
has led o he AllWISE ca alog (Cu i e al. 2013), which is
based on he o iginal wo epochs ob ained in 2010, and he
Ca WISE ca alog (Ca WISE2020; Ma occo e al. 2021), which
coadds he fi s 12 WISE, NEOWISE, and NEOWISE-R
epochs ( om 2010 h ough 2018). The AllWISE and Ca WISE
ca alogs equi e S/N>5 o inclusion o a sou ce, bu only in
a single band, as opposed o he SMC-Las ca alogs, which
equi e S/N>3inbo h bands.
The abili y o ex ac a pho ome ic sou ce om he da a
depends on he complexi y o he field. We examined wo es
fields, as shown in Figu e 6(also indica ed in Figu e 1). One
field, e e ed o as he “c owded”o “NGC 346 field,”is
12 12¢´
¢
and cen e ed on NGC 346 (α,δ=14°.7721,
−72°.17589). The o he field samples a spa sely popula ed
egion
2
020¢´
¢
in size and cen e ed a 21°.70, −73°.32. We
e e o i as he “spa se field.”
To compa e he co e age o hese su eys, we pe o med wo
es s in he spa se and c owded fields, using he SAGE-SMC
IRAC “Single F ame +Mosaic Pho ome y”A chi e, 1.5, as he
basis o compa ison. Fi s , we ex ac ed all sou ces om hese
egions o SAGE-SMC, AllWISE, Ca WISE, and all h ee
epochs (1, 2, and combined)o he SMC-Las a chi e. Table 5
gi es he sou ce coun s in each field o he SAGE-SMC a chi e,
Figu e 5. Cumula i e sou ce coun s o he a chi e (blue)and ca alog ( ed) o epoch 1 and he combined-epoch mosaics a bo h 3.6 and 4.5 μm. Epoch 1 beha es
simila ly o epoch 2. Table 3gi es he limi ing magni udes o 95% and 90% comple eness.
Table 3
Limi ing Magni udes and Comple eness o he SMC-Las Su ey
Epoch P oduc 3.6 μm 4.5 μm
95% 90% 95% 90%
1 o 2 Ca alog 15.6 16.2 15.8 16.4
Combined Ca alog 15.7 16.7 16.5 16.9
1 o 2 A chi e 17.5 17.6 17.4 17.5
Combined A chi e 17.8 17.9 17.8 17.9
18
A ailable om he In a ed Science A chi e (IRSA);h ps://i sa.
ipac.cal ech.edu.
9
The As onomical Jou nal, 167:149 (16pp), 2024 Ap il Kucha e al.
in bo h bands, he sepa a e coo dina e co ec ions a e applied
as-is. This h eshold is somewha a bi a y, bu wi h an ms o
abou 0 5 o he cen oid p ecision in he 2MASS ma ches,
he expec ed esidual e o o 50 sou ces is abou 0 07, well
unde he ypical posi ional sca e in he sou ce ex ac ions
om he mosaics (Figu e 4), so combining he e o s a is ics
o bo h bands o his numbe o sou ces (o g ea e )would
ha e li le benefi .
Fo ewe han 50 sou ces in ei he band, he medians o he
R.A. and decl. e o s a e aken o e bo h a ays, and he fixed
ocal plane o se is con e ed o o se s in R.A. and decl. and
hen appo ioned o he sepa a e a ays as an adjus men o he
o e all co ec ion depending on he ac ion o he o al sou ces
on each a ay. This p ocedu e equi es ha he sepa a e R.A.
and decl. e o dis ibu ions o he wo a ays su ficien ly
o e lap so ha he median is close o he a e age, and his is in
ac he case ( he sca e o he indi idual e o s is a ound hal
an a csecond).
Finally, wi h he poin ing efinemen applied as abo e, i was
ound ha he subsequen sou ce ex ac o coo dina es
measu ed om mosaics c ea ed om he co ec ed ames
di e ed sys ema ically om he 2MASS coo dina es by a small
amoun , abou 0 055, also fixed in ocal plane coo dina es,
which is likely due small di e ences in he cen oiding
p ocedu es used and possibly he sligh ly asymme ic PRFs.
As he ex ac o coo dina es a e he ones ul ima ely epo ed,
his o se is also added o he calcula ed poin ing co ec ions.
B.1. Scaling and Ro a ion
The co ec ions applied a e ansla ions in he ocal plane,
because he da a show ha he poin ing e o s a e la gely ime-
a ying e o s in he bo esigh , which o easonable
magni udes o e o will only cause ansla ions in he 5′-size
IRAC a ays. Fo ha eason, i is highly unlikely ha he e
will be any ue ame-by- ame a ia ions in scaling o field
o a ion. Thus, i is easonable o in es iga e he p esence o
fixed e o s in scaling and o a ion in he IRAC ame
as ome y.
Fo his pu pose, we ha e de eloped a χ
2
minimiza ion
p ocedu e o op imize he mapping o he measu ed x,y
cen oids o he sou ces on a gi en ame o he x,ycoo dina es
o he co esponding 2MASS sou ces, using he ame’s
nominal as ome y o map he 2MASS sou ces in o ocal
plane space. Fi e pa ame e s a e op imized: xand yscaling, a
o a ion, and xand y ansla ion.
Op imized pa ame e s a e de e mined o each IRAC ame
o an AOR (using AORs in ai ly c owded egions o be e
s a is ics), hen examined ac oss he AOR o pe sis ence.
While he e is conside able ame- o- ame sca e , pa icula ly
in he o a ion, we find no significan sys ema ic e o s in he
scaling pa ame e s o he 4.5 μm o a ion pa ame e in any o
he AORs examined. Howe e , he op imized 3.6 μm o a ion
shows a median e o o abou 1′, and his e o is p esen a a
nea ly iden ical alue o e all o he AORs. A 1′field o a ion
e o gi es a peak ue angula e o o abou 0 07 in he 5′
IRAC a ays and an e ec i e ms e o o abou 0 03 o an
ensemble o sou ces dis ibu ed ac oss he a ay.
ORCID iDs
T. A. Kucha h ps://o cid.o g/0000-0003-1955-8509
G. C. Sloan h ps://o cid.o g/0000-0003-4520-1044
D. R. Mizuno h ps://o cid.o g/0000-0003-0947-2824
Ka hleen E. K aeme h ps://o cid.o g/0000-0002-
2626-7155
M. L. Boye h ps://o cid.o g/0000-0003-4850-9589
Ma in A. T. G oenewegen h ps://o cid.o g/0000-0003-
2723-6075
O. C. Jones h ps://o cid.o g/0000-0003-4870-5547
F. Kempe h ps://o cid.o g/0000-0003-2743-8240
Iain McDonald h ps://o cid.o g/0000-0003-0356-0655
Joana M. Oli ei a h ps://o cid.o g/0000-0002-0861-7094
Ma a Sewiłoh ps://o cid.o g/0000-0003-2248-6032
Sunda S ini asan h ps://o cid.o g/0000-0002-2996-305X
Jacco Th. an Loon h ps://o cid.o g/0000-0002-1272-3017
Albe Zijls a h ps://o cid.o g/0000-0002-3171-5469
Re e ences
Be in, E., & A nou s, S. 1996, A&AS,117, 393
Bola o, A. D., Simon, J. D., S animi o ić, S., e al. 2007, ApJ,655, 212
Choudhu y, S., Sub amaniam, A., Cole, A. A., & Sohn, Y. J. 2018, MNRAS,
475, 4279
Cioni, M.-R. L., Clemen ini, G., Gi a di, L., e al. 2011, A&A,527, A116
Cu i, R. M. & 2MASS Team 2004, AAS Mee ing, 205, 91.03
Cu i, R. M., Sk u skie, M. F., an Dyk, S., e al. 2012, yCa , II/281yCa
Cu i, R. M., W igh , E. L., Con ow, T., e al. 2013, Explana o y Supplemen o
he AllWISE Da a Release P oduc s, Tech. ep., Cali o nia Ins i u e o
Technology
Fazio, G. G., Ho a, J. L., Allen, L. E., e al. 2004, ApJS,154, 10
Gaia Collabo a ion, Vallena i, A., B own, A. G. A., e al. 2023, A&A,674, A1
Go don, K. D., Meixne , M., Meade, M. R., e al. 2011, AJ,142, 102
G aczyk, D., Pie zyński, G., Thompson, I. B., e al. 2020, ApJ,904, 13
Habe l, F., S u m, R., Balle , J., e al. 2012, A&A,545, A128
IRAC Ins umen and Ins umen Suppo Teams 2021, IRAC Ins umen
Handbook, NASA/IPAC, h ps://i sa.ipac.cal ech.edu/da a/SPITZER/docs/
i ac/i acins umen handbook/home/
I a, Y., Onaka, T., Tanabé, T., e al. 2010, PASJ,62, 273
Joseph, T. D., Filipo ić, M. D., C aw o d, E. J., e al. 2019, MNRAS,
490, 1202
Ka o, D., Nagashima, C., Nagayama, T., e al. 2007, PASJ,59, 615
Mainze , A., Baue , J., Cu i, R. M., e al. 2014, ApJ,792, 30
Mako oz, D., & Khan, I. 2005, in ASP Con . Se . 347, As onomical Da a
Analysis So wa e and Sys ems XIV, ed. P. Shopbell, M. B i on, &
R. Ebe (San F ancisco, CA: ASP),81
Ma occo, F., Eisenha d , P. R. M., Fowle , J. W., e al. 2021, ApJS,253, 8
Meixne , M., Go don, K. D., Indebe ouw, R., e al. 2006, AJ,132, 2268
Meixne , M., Panuzzo, P., Roman-Du al, J., e al. 2013, AJ,146, 62
Mizuno, D. R., K aeme , K. E., Kucha , T. A., & Sloan, G. C. 2022, PASP,
134, 094101
Niede ho e , F., Cioni, M.-R. L., Rubele, S., e al. 2021, MNRAS,502, 2859
Riebel, D., Boye , M. L., S ini asan, S., e al. 2015, ApJ,807, 1
Rieke, G. H., Young, E. T., Engelb ach , C. W., e al. 2004, ApJS,154, 25
Russell, S. C., & Dopi a, M. A. 1992, ApJ,384, 508
SAGE–SMC Team & Go don K. 2011, The SAGE-SMC Da a Desc ip ion:
Deli e y 3 Ap il 2011, h ps://i sa.ipac.cal ech.edu/da a/SPITZER/
SAGE-SMC/docs/sage-smc_deli e y_ap 11.pd
Sewiło, M., Ca lson, L. R., Seale, J. P., e al. 2013, ApJ,778, 15
Sk u skie, M. F., Cu i, R. M., S iening, R., e al. 2006, AJ,131, 1163
Udalski, A., Soszyński, I., Szymański, M. K., e al. 2008, AcA, 58, 329
We ne , M. W., Roellig, T. L., Low, F. J., e al. 2004, ApJS,154, 1
W igh , E. L., Eisenha d , P. R. M., Mainze , A. K., e al. 2010, AJ,140, 1868
Za i sky, D., Ha is, J., Thompson, I. B., G ebel, E. K., & Massey, P. 2002, AJ,
123, 855
Zi ick, P., Kalli ayalil, N., an de Ma el, R. P., e al. 2018, ApJ,864, 55
16
The As onomical Jou nal, 167:149 (16pp), 2024 Ap il Kucha e al.