astro-ph/9511025 6 Nov 1995 AURA Inc under co op erative Observatories which is op erated by the Asso ciation of Universities for Research in Astronomy TRACING OPTICAL e mail Alfred P Sloan Visiting Visiting Astronomy Department University of California ragl astron b erkeley edu jrg astron berkeley edu Astronomer Lick Observatory University of California Astronomer Received THE AND Research Fellow Rosa COLOR Kitt A Gonz alez DYNAMICS INFRARED Peak August agreement with GRADIENTS National and OF SURFACE accepted Observatory James R Graham the National DISK IN Berkeley CA GALAXIES M National Octob er PHOTOMETRY Science Foundation Optical WITH Astronomy
ABSTRACT
We present optical and IR surface photometry of M NGC at
0 0
g r i J and K We also present a K image of M NGC for
S
comparison We use these data to investigate if the presence and strength
of spiral structure are wavelength dep endent and hence if they are features
in the old stellar disk or in the gas and dust only or asso ciated with star
formation or some combination The arms of M are still very prominent
in the infrared although broader and smo other than at optical wavelengths
Fourier decomp osition of the disk light reveals that the radial distribution of
p ower dep ends on wavelength which in turn implies that the spiral structure
traced in the visual i e young p opulation I and dust is di erent from the one
detected at m i e old stellar disk While M has signi cantly more p ower
in m than in any other mo de at all radii M shows noticeable p ower at
m m and m We observe radial mo dulation of the p ower and a
dep endency of p ower with wavelength that are consistent with mo dal theory of
spiral structure
A central motivation for our research is the fundamental idea of density
wave theory that the passage of a spiral density wave triggers star formation
as it sho cks the molecular gas in the disk If this is the case color gradients
due to stellar p opulation age gradients should o ccur across the spiral arms We
have found a gradient consistent with this scenario in a reddening free Q like
parameter at kp c galacto centric distance across one of the arms of M this
photometric parameter Q is very sensitive to red sup ergiants We rule out
r J g i
S
that the change in Q across the arm is mainly due to dust We also see a
r J g i
S
di erence in Q going from the interarm regions to the arms that indicates
r J g i
S
that arms cannot b e due exclusively to crowding of stellar orbits
We present the rst measurement of the angular sp eed of the spiral
p
pattern and of the lo cation of the corotation radius derived from the drift
velocity of the young stars away from their birth site The measured Q
r J g i
S
implies a star formation rate for M within the range of M yr a disk
stellar mass surface density of M p c and a maximum contribution of
0
p ercent from red sup ergiants to the K light in a small region and much
0
smaller on average We measure a K arm interarm contrast of to o high for
M to b e a truly isolated galaxy
Subject headings galaxies individua l M galaxies individua l M
galaxies infrared galaxies kinematics and dynamics galaxies photometry galaxies spirals
Introduction
The morphology of a spiral galaxy is of great imp ortance b ecause it is often
the only available clue to the galaxy s underlying structure and dynamics
However the morphology of spirals is highly dep endent on the wavelength at
which they are observed the IR morphology is known for only a few spirals
and in several of these cases the IR and optical morphologies di er signi cantly
i e in the presence or absence of bars the degree of o cculence of the arms
even the apparent Hubble type Hackwell Schweizer Scoville et al
Thronson et al Blo ck Wainscoat Blo ck et al
Past studies of spirals have used optical surface photometry to prob e the
dynamics of stellar disks e g Schweizer Talbot Jensen Dufour
Elmegreen Iye et al Elmegreen Elmegreen
Elmegreen Elmegreen Seiden Elmegreen Elmegreen Montenegro
However even at I the observed morphology of disk galaxies is strongly
in uenced by extinction Combined optical and near infrared data extending
to a wavelength of m o er a b etter opp ortunity for measuring extinction
and reddening Correcting for extinction is crucial when using morphology as
a prob e of disk dynamics since reddening and extinction can lead to incorrect
estimates of the contribution of young stars and the true arm interarm density
contrast Also given that the e ect of dust extinction is much less severe in the
IR relative to the optical A A the IR surface brightness distribution
K B
is a more direct measure of the stellar disk mass surface density distribution
Rix Rieke
We present the rst results of an ongoing optical IR survey of spiral
galaxies of di erent types and a range of inclinations The primary goal of our
study is to discover the nature and triggering mechanisms of spiral structure
Mo dal theory Bertin et al a b for example predicts that only lower order
mo des m and m will survive in an old stellar disk Conversely swing
ampli cation Toomre Toomre Kalna js predicts that gas should
form a stronger and more tightly wound spiral than the old stellar p opulation
The Sc I galaxy M o ccupies a sp ecial place in our survey Its spiral pattern
app ears to b e exceptionally strong Schweizer it has an unusual three arm
spiral with a very prominent m mo de i e one arm is much stronger than the
other two In the optical its arms show a hint of the azimuthally asymmetric
pro les that would b e pro duced if star formation is induced by the passage of a
spiral density wave Schweizer Assuming a distance of Mp c Jacoby
00
et al M o ers go o d spatial resolution p c the galaxy has
0
R Elmegreen et al
M NGC is one of the most regular two armed grand design
00
spirals in the sky Type Sb cI distance Mp c for p c Schweizer
0 0
R Elmegreen et al The K data of M has allowed
interesting comparisons with M regarding their azimuthal distribution of light
at that wavelength
Observations
Figure shows M at g A The optical images were obtained in
April and UT with the m Nickel telescop e at Lick Observatory
00
The pixel scale is pixel for a eld of view of the Ford CCD binned
0
of Given the large size of the galaxy usually we did not dither more
than a fraction of an arcminute b etween each typically to seconds of
time individual exp osure Frames were bias and dark current subtracted prior
to at elding with twilight ats Bad columns and pixels were masked out and
a constant was t to the ob ject free areas for sky subtraction Individual frames
were registered to the nearest pixel i e without any interpolation Cosmic
rays were identi ed by subtracting a median image from each shifted individual
frame and then masked out The frames were then coadded and the resulting
mosaic was divided by an exp osure map thus getting units of Analog to Dig ita l
Units ADU s at every pixel A nal measure of sky was p erformed at the
edges of the mosaic Since the galaxy extends almost all the way to the edges
of the chip we exp ect that the biggest systematic error in the photometry will
come from overestimating the brightness of the sky at R the galaxy
surface brightness is p ercent of the sky brightness in the g band and
p ercent of sky in the i band
The photometric calibration in the optical bands was done in the Thuan
Gunn system Thuan Gunn Wade Ho essel Elias Kent
This system is calibrated with the primary standard star BD sd
F Since we did not observe BD in April we calculated synthetic
Thuan Gunn magnitudes for the sp ectrophotometric standard star Feige
in our three bandpasses Each bandpass was de ned as the pro duct of the
resp ective lter transmission curve the Q E curve of the Ford CCD we used
and the atmospheric transmission curve for Lick from Hayes the uxes
of BD and Feige were taken from Oke Gunn Massey et
al and Massey Gronwall These synthetic magnitudes were
then used to assign empirical magnitudes to the other standard stars observed
in April and to nd the zero p oint of that night The night of April was
not photometric but we calibrated the nal mosaics by measuring stars in the
th
frames taken on the
0
Figure shows M at K m the arms are marked following the
nomenclature of Iye et al The infrared images were obtained in
March UT with the m telescop e at Kitt Peak the Infrared Imager IRIM
0
uses a x NICMOS detector and has a eld of view of Probst
Heim Buchholz Luce We obtained min frames coadditions of s
0
exp osures at J and s exp osures at K sequentially p ositioni ng the galaxy
in the four quadrants of the array and no dding to sky only every fth frame
The rst step in the reduction was a pixel by pixel linearization The images
0
were then at elded J with sky ats and K with dome ats Sky frames were
multiplicatively scaled to the value at the edges of ob ject frames and subtracted
from them sky subtraction takes care also of the dark current the sky frames
were made by medianing i e averaging the two central values for each pixel
the four frames adjacent to each galaxy frame after masking out bright stars
and extended ob jects Individual frames were then sub pixelated in b oth
00
dimensions and registered to the nearest once again without any additional
interpolation The rest of the reduction was analogous to that of the optical
images Despite the large size of the nal IR mosaic sky subtraction will still b e
the biggest contribution to the systematic error in the photometry estimated
to b e p ercent
The photometric calibration of the IR images was done with faint UKIRT
standard stars Casali Hawarden For the calculation of the zero p oint
0
we assumed K K H K Wainscoat Cowie
0
Figure shows the K band image of M It was obtained in March
at Kitt Peak under non photometric conditions Given the angular size of
M we o set frequently to sky A eld of view smaller than the ob ject and the
changing sky in cloudy conditions add up to a bigger uncertainty in the sky
subtraction however we used rms minimization to improve the scaling of sky
to ob ject i e in each case we used the sky frame that minimized the noise in
the sky subtracted image the sky subtracted frame with the lowest rms was
deemed the reference image and nally we added a constant to each of the
remaining images to shift the level at the edges to the value at the same lo cation
in the reference image Our nal mosaic is bigger and therefore includes regions
farther away from the center of the galaxy than recently published data Rix
Rieke we should then have a b etter estimate of the sky level
Table gives a summary of the observations
Analysis
All the images were co registered using the r image as a reference and
S
depro jected to a face on view Fig For depro jection we rst rotated the
galaxy to make the ma jor axis parallel to the columns i e rotated by an angle
equal to the negative of the PA then we stretched the pixels along rows i e
the minor axis by a factor equal to cos where is the inclination angle
For M we used PA Pho okun Vogel Mundy We
decided to use these numbers for consistency since we use the rotation curve of
M published in this reference However it is worth mentioning that although
Pho okun et al used the pro cedure ROTCUR Begeman to simultaneously
determine the rotation curve PA and they forced the two latter parameters
to remain close to the values that maximize the p ower in the axisymmetric
comp onent PA Iye et al It is not p ossible to determine
kinematical inclinations for galaxies with Begeman and M
must b e very close to that limit as judged from the numbers used by Schweizer
PA With the exception of mo de analysis our conclusions
should not b e signi cantly a ected by the choice of PA and For M we to ok
b oth PA and from Schweizer PA
Next we unwrapped the spiral arms by plotting the images in a vs l og R
map Fig cf Iye et al Elmegreen et al Under this geometric
transformation logarithmic spirals will app ear as straight lines with slop e
ln cot i where i is the pitch angle if the azimuthal angle increases
in the direction of rotation and the spiral arms trail These unwound or
unwrapped images are useful to measure the pitch angle of the arms to measure
the arm interarm contrast by taking cuts along columns and to calculate
the Fourier p ower sp ectrum at each radius Our program allows us to choose
arbitrary smo othing lengths in b oth and R and the output images are in units
00
of ADU s 2 For example it is easy to pro duce a radial light pro le simply
by smo othing in with an angle equal to
To reduce the contamination by p oint sources i e foreground stars and HI I
regions in the unwrapped images we rst masked pixels that were or more
ab ove the average noise of smo oth lo oking arm regions in the unsharp masked g
image and then used the median surface brightness We smo othed the data in
bins with a radial extent of r r and an azimuthal extent of
min max
the separation b etween bins is r r in radius and in azimuth
These smo othing lengths are similar to those used in previous studies Schweizer
Rix Rieke and allow direct comparisons of results At the same
time the use of bin spacings smaller than the smo othing lengths equivalent
to b oxcar smo othing reveals structure at ner scales than the aforementioned
works For the error analysis we smo othed the images in disjoint bins which
are statistically indep endent and used smo othing lengths of the order of the
spatial resolution of the data
Results
Mo de Analysis
Using the unwound images we calculated row by row p ower sp ectra for M
and M We de ne the p ower in each mo de m at each radius as
jH m j
P m
jH j
Here
Z
im
I e d H m
with I b eing the surface brightness we are implicitly integrating over all
pitch angles
The interpretation of Fourier transforms as a prob e of spiral structure can b e
complicated by contamination by foreground stars sky gradients and internal
extinction Elmegreen Elmegreen Montenegro also the distribution of
p ower b etween di erent mo des is highly sensitive to the depro jection parameters
Moreover it is hard to estimate errors in the p ower sp ectra b ecause pixels are
no longer statistically indep endent after taking the Fourier transform However
0
b oth the fact that at all wavelengths and in particular at g and at K Fig
a and b we seem to recover similar information ab out M and the small
false alarm probabilities of p eaks in p erio dograms of the same data cf Press
et al give us con dence in the following general conclusions
We con rm that at all radii Figure M has signi cantly more p ower
in m than in any other mo de Rix Rieke Although M Fig
displays a dominant m mo de b etween and R it shows signi cant
p ower at m and m to o with m stronger inside R and m
p eaking outside this radius The signi cant strength of m is consistent with
a non linear interaction b etween it and m exciting m Bertin a
driving of m by m has also b een suggested by Elmegreen et al
0
The ripples of m are likely real since we observe them b oth at g and at K
it can b e explained as the result of interference b etween the waves that supp ort
the mo de Binney Tremaine Bertin Finally while m shows
0
comparable strength at g and at K Fig there is a trend for higher order
mo des i e m and m to get fainter at longer wavelengths relative to
m Fig a and b this is consistent with mo dal theory which predicts
that higher order mo des will b e stronger in the gas and dust than in the stellar
disk since as stars age their velocity disp ersion increases and smo oths out small
scale features
Azimuthal Color Gradients
A Red Supergiant Sensitive Q Parameter
If the passage of a spiral density wave and the consequent sho cking of
molecular clouds triggers the formation of stars then the mean age t of the
ag e
young stellar p opulation is related to the distance d from the arm at a xed
radius through the di erence b etween the angular rotation sp eed R and
the pattern sp eed i e d R Rt In this case stellar p opulation
p p ag e
synthesis mo dels Charlot Bruzual Bruzual Charlot predict
that the color evolution across an arm is dramatic enough that age gradients can
b e spatially resolved in a nearby spiral even if the ratio of young to old stars is
as small as p ercent by mass Figure shows the mo del Q like reddening free
E R J
parameter Q R J V I R J V I Mihalas
RJ V I
E V I
Binney vs t for an instantaneous burst sup erimp osed on a background ag e
of p ercent of old age yr stars see x the young p opulation
has a Salp eter IMF with M M and M M Q is reddening
low er upper
insensitive for screen absorption assuming the extinction curve of Rieke
Leb ofsky
Q is an excellent tracer of star formation since it has almost the same value
for all except the reddest stars Figure Writing Q in a slightly di erent way
helps to appreciate this fact
I I
V J
Q l og
RJ V I
I I
I R
A combination of small numerator and large denominator in the quotient
i e low V and J surface brightnesses and high R and I uxes will yield a
small and even negative Q a large numerator plus small denominator i e high
V and J surface brightnesses and low R and I uxes will result in a high Q
It is easy to achieve a low value of Q with a single Planckian p eaking around
A Fig a however to obtain a high value of Q it is necessary to have
a double humped sp ectral energy distribution with a valley around A
and increasing intensity b oth redward and blueward of that wavelength Fig
b Therefore Q will have a higher value for a mixture of blue and red stars
than for just ab out any single star Extreme values of Q o ccur when the most
massive stars of the young p opulation b ecome red sup ergiants Indeed the plot
in Figure displays a very pronounced p eak yr after the starburst
due to the app earance of luminous red stars it also shows that Q is indeed
reddening insensitive for screen absorption and negligibly a ected by a mix of
dust and stars with plane parallel geometry and or in a galaxy with
V V
cos which is slightly more edge on than M Bruzual Magris C
Calvet
In general extinction and reddening prop erties are highly dep endent on
geometry The widely used Witt Thronson Capuano mo dels of
radiative transfer within galaxies have spherical symmetry and consist of
mixtures of stars and dust with di erent relative spatial distributions Figure
shows Q for the Witt et al starburst galaxy mo del with
RJ V I V
Q is no longer reddening insensitive for Their dusty galaxy mo del
V
o ers a similar picture with Q changing more than our error for
RJ V I V
These are the most unfavorable i e optically thick Witt et al mo dels that
apply to a galactic disk and it is unlikely that is for a face on galaxy
V
Bruzual et al Peletier Willner Peletier et al so Q should
b e reddening insensitive for the disk of M Notice that Q is bigger at larger
optical depths
Figures and show Q vs t for our set of lters i e
ag e
E r J
S
Q r J g i r J g i using the
r J g i S S
S
E g i
extinction curves of Schneider Gunn Ho essel and Rieke Leb ofsky
for di erent M di erent durations of the starburst and di erent
upper
p ercentages of young stars
Figure displays an unwound Q image of M This image comp ellingly
r J g i
S
shows Q di erences b etween the arms and the interarm regions and more
imp ortantly coherent Q changes through the arms A previous study applying
similar techniques to UBV surface photometry of M Talbot et al had
detected regions either with or without star formation but no intermediate
color We have fo cused our attention on a patch in the N arm at kp c
0
from the center of the galaxy that from the J K image Figure app ears
0
to have little contamination by dust J K is a go o d dust tracer and su ers
0
less contamination from young stars than say g K Even column by column
i e in annuli having width equal to the spatial resolution of the data this
sp ot exhibits an azimuthally asymmetric pro le in Q that is consistent with
r J g i
S
coherent star formation induced by the passage of the spiral density wave to
00 00
increase the signal to no is e we have collapsed the patch from r to
0
Figure shows sup erimp osed the Q pro le and the J K pro le of this
r J g i
S
patch In this region Q increases even as the amount of dust inferred from
r J g i
S
0
J K starts to decrease away from the dust lane This is the opp osite of what
the Bruzual et al and Witt et al mo dels of radiative transfer in mixed stars and
dust predict for a diminishing optical depth and therefore we can rule out that
the change in Q is due mainly to dust Figure shows a full azimuthal cut
r J g i
S
00 0
of Q at R as well as the K azimuthal light pro le for the same radii
r J g i
S
Q changes at the lo cation of every arm and the arm Q pro les are more
r J g i
S
similar to each other than the arm light pro les among themselves a similar
pro cess seems to b e at work in the three arms that is compatible with a change
of age of the stars across them If arms were due to crowding of stellar orbits
exclusively we should not see any change in the Q parameter going from the
interarm regions to the arms
The Stel lar Drift and the Pattern Speed
Figure shows the Q pro le vs distance from the arm for the region
r J g i
S
marked in Fig and and two mo del Q curves with and p ercent
r J g i
S
by mass of young stars resp ectively Both mo dels have b een calculated with
a Salp eter IMF with M M and M M and a duration of
low er upper
the burst of yr past studies on OB asso ciations b oth galactic and
extragalactic suggest that star formation pro ceeds in them for yr
Elmegreen Lada Do om De Greve de Lo ore Massey et al
Regan Wilson Gruendl In b oth mo dels the young burst
is sup erimp osed on a background of stars yr old The choice of
p ercent of young p opulation seems reasonable compared b oth to the number
of times that the region under consideration has encountered the spiral pattern
once every yr in the lifetime of the disk yr and to the amount
of light in the arms contributed by old stars in the B band as estimated by
Schweizer It is also not that surprising that the data is well t by a
young p opulation lacking extremely massive stars b ecause we have masked out
the brightest stars and the HI I regions
We have chosen the origin of the burst d at the lo cation of the
dust lane in the data i e the lo cation of the sho ck Rob erts Huntley
van Albada To transform the time in the mo dels into distance from
the site of star formation we have adopted the at rotation curve with
V km s derived by Pho okun et al from HI data and then
r ot
chosen the di erence b etween kp c and that allows a go o d t of the
p
mo del Q to the data Taking kp c km s kp c yields
r J g i p
S
km s kp c and places R at R This is slightly
p CR
larger than the R found by Elmegreen et al but in b etter
agreement with mo dal theory of spiral structure Bertin et al a which
predicts that the corotation zone is at radial scale lengths i e at a lo cation
where the gaseous mass is plentiful enough to play a signi cant dynamical
00 00
role By tting the radial light pro le of M b etween and we have
0
measured a scale length that varies b etween R at g and R at K
The Star Formation Rate
0
By interpreting the observed K ux in the interarm region downstream of
the sho ck in terms of a p opulation synthesis mo del with the same parameters
used for Q a mixture of p ercent of stars yr old estimated from
their distance from the dust lane and p ercent of stars yr old
we nd a disk stellar mass surface density of M p c With p ercent of
young stars formed in yr these numbers imply a star formation rate of
M yr if integrated over the area inside R and M yr if all the
area inside R is included This result is in the range of M yr found
by Kennicutt for Sc galaxies It is imp ortant to note though that he
derived his result from H ux and UBV colors i e mainly from the output of
stars more massive than M which make up p ercent of the mass of all
stars we are sampling stars less massive that M and therefore a much bigger
fraction of the newly formed stars Combined with a detected color gradient
that is consistent with the evolution in time of a stellar p opulation whose birth
was triggered at the lo cation of the dust lane this general agreement suggests
that at least some of the star formation in M is asso ciated with the spiral
density wave
Arm Interarm Contrast
We measure arm interarm intensity contrasts by taking cuts along columns
in the unwrapped images Arm interarm contrasts at di erent wavelengths
have b een considered a go o d discriminant b etween arm triggering mechanisms
For example o cculent arms due to self propagating star formation would
show large azimuthal intensity contrasts in the blue which traces dust and
hot young stars and very small ones in the infrared which traces the mass
density variations in the disk Elmegreen Elmegreen Rix Rieke
Following Elmegreen Elmegreen we de ne arm interarm contrast as the ratio
of an intensity maximum in an azimuthal cut to the average of the intensity of
the two adjacent minima
Figures a and b show azimuthal brightness pro les of M in g and
0 0 00
K and in r and K resp ectively at R At optical wavelengths the
S
contrast app ears quite altered due to the presence of dust as suggested by direct
0
images and by dust tracers like J K we hence conclude that it is di cult to
measure intrinsic arm interarm contrasts in visible bands When estimated at
0 00
K the arm interarm contrast of M rises from at R to at
00
R then falls back again to a factor of at R it then shows an
0
abrupt jump to b etween and R For M the K arm interarm
00 00
contrast go es from at R to at R then declines
steadily to a factor of at R These numbers are in excellent agreement
with Rix and Rieke Thus for the range of radii where we can carry out
0
comparisons the K arm interarm contrast of the two galaxies is very similar
At A Schweizer measured an arm strength for M that was
almost twice that for M he de ned arm strength as the ratio of arm light
to disk light At each radius disk meant all the light b elow the mean
surface brightness of the two deep est minima separated by more than degrees
in azimuth arms were de ned as the average amount of light ab ove the
disk Although Schweizer tried to avoid dust lanes in the choice of minima
0
the K band is less a ected by dust than optical wavelengths and therefore
b etter for this sort of exp eriment also the arm interarm contrast is likely a
more ob jective estimator of a galaxy s mass spatial distribution than the arm
strength since it do es not involve interpolating across the arms to determine
the old disk surface brightness As already stated ab ove from the p oint of
0
view of their K band arm interarm contrast M and M seem to b e similar ob jects
0
Since we have used the K band to estimate the arm interarm contrast as
well as the star formation rate it is imp ortant to rule out a large contribution
0
from red sup ergiants to the K light i e that instead of actually tracing mass
0
K is seeing a few young red stars with a low M L ratio In the case of M
after masking out HI I regions and p oint sources Rix Rieke inferred
from the strength of the CO m molecular absorption that red sup ergiants
contributed more than p ercent to the K m light only in a small
lo calized region of one of the arms Using the CO index K CO Rhoads
found for NGC that while young co ol sup ergiants do in fact
dominate the m light of active star forming regions the bright HI I regions
we masked out when the smo oth areas of the spiral arms are included the
measured CO index converges to a value consistent with a p opulation
yr old the most massive stars of such a p opulation would b e M no
sup ergiants unfortunately Rhoads do es not show the index for the smo oth
arm regions alone Our results are consistent with the ones just quoted The
t to the observed Q with p ercent young stars by mass and an upp er
r J g i
S
mass limit of the burst of M implies that less than p ercent by mass of
0
red sup ergiants should b e present and their p eak contribution to the K light is
p ercent b etween the ages of and yr This is a short interval
and on average the red sup ergiant contribution will b e much less
0
Discarding a bigger contribution from red sup ergiants to K has signi cant
consequences for disk dynamics An arm interarm mass contrast is to o big
to b e the result of a long lived spiral density wave in an isolated disk i e in
the absence of driving by a bar or companion or of a disk co oling mechanism
like gas infall Sellwoo d Carlb erg Sellwoo d In the case of M
0
therefore the size of the K arm interarm contrast is an indication that the
galaxy is only apparently isolated Indeed HI data shows signatures of gas
infall namely clouds sup erp osed on and b eyond the HI disk at velocities up
to km s from those found for the disk gas Pho okun et al and
excess emission in one of the high velocity wings of the double horned HI pro le
observed with the Arecib o m telescop e Shulman Bregman Rob erts
Summary and Conclusions
As big format infrared detectors have b ecome increasingly available
observations of disk galaxies have revealed spiral arms that although broader
and indeed smo other than in the optical have turned out to b e much stronger
0
and b etter de ned than foreseen even at K Blo ck Wainscoat Blo ck
et al Rix Zaritsky It do es app ear that with the much reduced
obscuration by dust the intrinsic structure of galactic disks is nally b eing
unveiled and with it the signatures of the underlying dynamics Resulting
from the combination of arm strength in the infrared and b etter detectors the
signal to no ise of our data is signi cantly sup erior to previous observations
Fourier decomp osition of the disk light of M reveals that the radial
distribution of p ower dep ends on wavelength which in turn implies that the
spiral structure traced in the visual i e young p opulation I and dust is di erent
from the one detected at m i e old stellar disk The radial mo dulation of
the p ower and the dep endency of p ower with wavelength that we observe are
consistent with mo dal theory of spiral structure
We have identi ed a Q like photometric parameter that is very sensitive
to red sup ergiants Thanks to very deep optical and near infrared images we
have a p ositive detection of an azimuthal gradient in Q due to a stellar
r J g i S
p opulation age gradient across a spiral arm of M From the drift velocity
of the young stars away from their birth site we have measured the angular
sp eed of the spiral pattern and the lo cation of the corotation radius The
p
measured Q implies a star formation rate for M within the range of
r J g i
S
M yr a disk stellar mass surface density of M p c and a maximum
0
contribution of p ercent from red sup ergiants to the K light in a small
region and much less on average According to the mo dels that b etter resemble
our data spiral structure is b oth the pro duct of a density wave that involves
the whole stellar disk hence the close to p ercent by mass of old stars in
the arm and the cause of organized star formation as the wave encounters and
sho cks molecular gas
Without very useful discussions with Joan Na jita and Arjun Dey this
pro ject might have never b een started We are grateful to Giusepp e Bertin for
his comments Gustavo Bruzual and St ephaneCharlot mo di ed their programs
to b etter suit our needs Ron Probst Nick Buchholz and Mike Merrill provided
invaluable assistance for observing with IRIM The Lick Observatory sta was
always there when we needed them We acknowledge the nancial supp ort
received from the Sloan Foundation and from the Packard Foundation R A G
appreciates the supp ort of the Universidad Nacional Aut onoma de M exico
through a DGAPA Do ctoral Fellowship
Table
Observation Log
Ob ject Filter FWHM Exp osure Telescope
ef f
M g A A min Lick m
r A A min
S
i A A min
J m m min Kitt Peak m
0
K m m min
0
M K m m min
REFERENCES
Begeman K A A
Bertin G PASP
Bertin G private communication
Bertin G Lin C C Lowe S A Thurstans R P a ApJ
Bertin G Lin C C Lowe S A Thurstans R P b ApJ
Binney J Tremaine S Galactic Dynamics Princeton Princeton Univ
Press
Blo ck D L Bertin G Sto ckton A Grosb l P Mo orwoo d A F
M Peletier R F A A
Blo ck D L Wainscoat R J Nature
Bruzual G Charlot S ApJ
Bruzual G Magris C G Calvet N ApJ
Casali M M Hawarden T G JCMT UKIRT Newsletter
Charlot S Bruzual G ApJ
Do om C De Greve J P de Lo ore C ApJ
Elmegreen B G Elmegreen D M ApJ
Elmegreen B G Elmegreen D M Montenegro L ApJS
Elmegreen B G Elmegreen D M Montenegro L PASP
Elmegreen B G Elmegreen D M Seiden P E ApJ
Elmegreen B G Lada C J ApJ
Elmegreen D M ApJS
Elmegreen D M Elmegreen B G ApJ
Elmegreen D M Elmegreen B G ApJ
Gruendl R A Ph D thesis Univ of Maryland College Park
Hackwell J A Schweizer F ApJ
Hayes D S ApJ
Heim G B Buchholz N C Luce R W in Instrumentation in
Astronomy VI I I eds D L Crawford E R Craine Bellingham Washington
USA SPIE
Iye M Okamura S Hamab e M Watanabe M ApJ
Jacoby G H et al PASP
Johnson H L ARA A
Kennicutt R C Jr ApJ
Kent S M PASP
Ko ornneef J A A
Lee T A ApJ
Massey P Garmany C D Silkey M Degioia Eastwo o d K
AJ
Massey P Gronwall C ApJ
Massey P Strob el K Barnes J V Anderson E ApJ
Mihalas D Binney J Galactic Astronomy nd ed San
Francisco Freeman
Oke J B Gunn J E ApJ
Peletier R F Willner S P AJ
Peletier R F Valentijn E A Mo orwoo d A F M Freudling W Knap en
J H Beckman J E A A in press
Pho okun B Vogel S N Mundy L G ApJ
Press W H Teukolsky S A Vetterling W T Flannery B P
Numerical Recip es in FORTRAN the Art of Scienti c Computing nd ed
New York Cambridge Univ Press
Probst R G Instrument Manual for the IRIM Infrared Camera System
NOAO
Regan M W Wilson C D AJ
Rieke G H Leb ofsky M J ApJ
Rix H W Rieke M J ApJ
Rix H W Zaritsky D ApJ
Rhoads J E preprint
Rob erts W W Jr Huntley J M van Albada G D
ApJ
Schneider D P Gunn J E Ho essel J ApJ
Schulman E Bregman J N Rob erts M S ApJ
Schweizer F ApJS
Scoville N Z Matthews K Carico D P Sanders D B
ApJ L
Sellwoo d J A Carlb erg R G ApJ
Sellwoo d J A private communication
Talbot R J Jr Jensen E B Dufour R J ApJ
Thronson H A et al ApJ
Thuan T X Gunn J E PASP
Toomre A in Dynamics and Interactions of Galaxies ed R Wielen
Berlin Springer
Toomre A and Kalna js A J in Dynamics of Disc Galaxies ed B
Sundelius Sweden G otesborg Univ
Wade R A Ho essel J G Elias J H PASP
Wainscoat R J Cowie L L AJ
Witt A N Thronson H A Jr Capuano J M Jr ApJ
A
This manuscript was prepared with the AAS L T X macros v E