2MASS Galaxy Morphology:
JHKs Images sorted by Hubble Type
T. Jarrett, IPAC
-
See also the PASP paper
Near-Infrared Galaxy Morphology Atlas
for analysis and postscript figures
The following galaxy images represent a subset of the Fall/Winter 1999 2MASS public data release. They correspond to NED cross-identifications with known morphological (optically-deciphered) "Hubble" type (see the classic "Tuning Fork" and de Vaucouleurs explanation of Hubble's classification scheme. The jpeg images are derived from a JHKs "fits data cube" product created by the 2MASS Extended Source Processor, ranging in size between 21 and 101 arcsec in width. The images are annotated with the NED catalog name, the morphological type, the size of the image (in arcsec), the 2MASS name, the Ks mag and the J-Kscolor. More details are given at the end of this document under notes.
See also the classic 1959 paper by De Vaucouleurs on Galaxy Classification and Morphology that set the standard for decades to come. The PASP paper Near-Infrared Galaxy Morphology Atlas presents a full statistical analysis of the atlas.
- go to
Ellipticals
Spirals :
Lenticulars normal
normal/barred barred
Dwarfs Compact
Peculiar AGN
Composite
Notes
| Spirals |
NGC 3363: 2MASX 1045094+220441 |
-
Elliptical - Spiral Transition
| S0/SA0 (lenticulars) |
NGC 5010: 2MASX 1312263-154752 |
| Sa/SAa |
NGC 4868: 2MASX 1259089+371837 |
| SAb/Sb |
UGC 06680: 2MASX 1143021+193859 |
| SAc/Sc/Sd (late-type spiral) |
NGC 3140: 2MASX 1009278-163741 |
| Sd only (extreme late-type spiral) |
UGC 08702: 2MASX 1345272+475522 |
Normal - Barred Transition Spirals
| SABa (normal/barred, early-type) |
NGC 3788: 2MASX 1139445+315552 |
| SABb (normal/barred, intermediate-type) |
NGC 4017: 2MASX 1158456+272709 |
| SABc/SABcd (normal/barred, late-type) |
NGC 5595: 2MASX 1424133-164322 |
Barred Spirals
| SB0/SBa (barred early-type spiral) |
NGC 5915: 2MASX 1521330-130530 |
| SBb (barred intermediate-type spiral) |
NGC 3110: 2MASX 1004019-062829 |
| SBc/SBd (barred late-type spiral) |
NGC 5260: 2MASX 1340199-235129 |
| Ellipticals/Spheroidals |
NGC 5223: 2MASX 1334251+344125 |
| E0 |
NGC 4283 : 2MASX 1220207+291839 |
| E1 |
NGC 4687 : 2MASX 1247237+352107 |
| E2 |
NGC 5354 : 2MASX 1353267+401809 |
| E3 |
NGC 3158 : 2MASX 101350+384553 |
| E4 |
NGC 3605: 2MASX 1116466+180101 |
| E5 |
NPM1G+32.0347: 2MASX 1325026+323310 |
| E6 |
NGC 5028: 2MASX 131458-130232 |
| E7 |
NGP9F270-14581: 2MASX 1325151+324016 |
| Dwarf Ellipticals/Spheroidals/Spirals |
NGC 3101: 2MASX 1217196+115636 |
| Compact and BCD |
IC 3017: 2MASX 1209249+133429 |
| IRR/IM/Sm/Pec |
NGC 4004: 2MASX 158051+275243 |
| AGN/Seyferts/LINERS |
MRK 0463E: 2MASX 1356028+182218 |
We may now derive an analogous Hubble "tuning fork" composite morphology chart appropriate to the near-infrared, 1 to 2 microns, using 2MASS image data and optically-identified morphological type. The representative sample ranges in brightness between 10 - 11th mag in the Ks band, and their orientations are generally close to face-on (i.e., disk axis ratio near unity for spirals).
Notes
2MASS extended sources (e.g., galaxies) are identified and characterized with a "pipeline" software package called GALWORKS. An overview and description of the algorithms can be found at 2MASS Extended Source Catalog. For a more general overview of the 2MASS pipeline reduction and resultant data products, see the Explanatory Supplement.
Fits Data Cube
-
Each extended source in the 2MASS database includes a small image data cube, representing
the J, H and Ks "postage stamp" images (planes 1,2 and 3, respectively),
derived from the full resolution Atlas images.
The size of the cube is dynamically selected, scaled according to the measureed "size" of the extended
source (see below), with minima/maxima of 21 and 101 pixels (each pixel is 1 arcsec
in width). The variable image size was needed for purely practical reasons (limited
disk storage). For the largest galaxies (typically Ks < 9th mag) the image size
is not large enough and the outer parts of the galaxy are clipped off. Note that the
2MASS survery is *not* complete for galaxies larger than ~1' in width (the reason is that
the overlap between "scans" is ~50"; hence galaxies falling near a scan edge will be
clipped and may not be extracted, depending on the severity of the flux loss).
-
The jpeg images show a 3-color "rgb"=="JHKs"
representation for each galaxy, overlayed with J-band contours. The
contours start with the intensity level equal to 3*sigma (thrice the
background sky noise) or (if larger) the level equal to 5% of the J-band peak.
Each successive contour is multiplied
by a factor of two (i.e., 20%, 30%,...).
The images were "color stretched" generated using the following recipe. We would like to differentiate the true color of the galaxy. Hence, we stretch the color scale relative to the color of a "normal" galaxy, which turns out to be J-K = 1.0 and H-K = 0.28 (plus or minus 10-20%). We first get the Ks peak of the galaxy (and convert it to a mJy flux unit). We then deduce the J and H peaks using the color of a "normal" galaxy (again, we have converted the mags to a calibrated flux unit). The J,H Ks flux peak represents the upper color-stretch limit. The lower limit is chosen as follows: take -7.5% of the peak; here put a floor on the respective peaks of 5 dn ("data number", representing the raw uncalibrared pixel value) -- this is needed for the low surface brightness galaxies.
We assigned the "blue" intensity scale to the J-band image; "green" to the H-band image, and "red" to the Ks image. The resultant color-stretch will allow sufficient contrast between the galaxy core and the low surface brightness emission in the disks. Normal galaxies (J-K ~1) will appear to the eye as white or yellowish in appearance. Red galaxies (J-K > 1.2) will appear to the eye as "reddish". Blue galaxies (J-K < 1; late-type spirals may have this color) will appear blue or greenish. For faint sources (or alternatively, LSB galaxies), the background noise becomes appreciable (compared to the galaxy flux) and the contrast between the cores and background is greatly increased (the background is brighter with the ubiquitous H-band airglow more apparent).
A figure caption is included with the jpegs: NED name, morphological type, size of image (arcsec), 2MASS name, Ks mag and J-Ks color. The 2MASS name maps into the equatorial position of the source (hhmmsss+-ddmmss).
Galaxy Size and Photometry
-
The galaxy Ks flux and J-Ks color shown in the jpeg images correspond
to the "K-band fiducial" 20 mag/arcsec2 elliptical isophotal photometry.
Here the elliptical orientation (position angle and ellipticity)
of the galaxy is determined from the Ks image (at roughly the isophote corresponding
to 3 times the background RMS "noise"). The radius, or semi-major axis, is derived from
the Ks 20 mag/arcsec2 elliptical isophote (which is approximately
equal to 1 times the background RMS "noise"). It is this "radius" that represents the
size of the galaxy. The JHKs photometry is then performed in an aperture given by the
Ks position angle and ellipticity (derived from the 3-sigma isophote)
and the semi-major axis (derived from ~1-sigma isophote); hence the term "Ks fiducial",
allowing comparison between bands and color computations.
Caveats and Known Problems
-
Some galaxies (typically the largest ones) are located near a scan (or Atlas image) edge.
Hence, the measured flux is systematically too small (i.e., the Ks mag represents an upper
limit). Note that the J-Ks color is only minimally affected by the flux loss (since it occurs
nearly equally for all bands).
On rare occasions, the true center of a large galaxy is offset from the image center. This occurs from poor catalog (via NED) coordinate positions. GALWORKS cannot adequately "locate" the galaxy center. Under these circumstances, the measured flux (and of course, coordinate position) are incorrect. In a related case, there may be a bright foreground star near the core of the galaxy that fools the extended source processor and uses the star as the central position.
A few galaxies (usually faint) may not have a Ks flux due to any number of reasons, from scan-edge proximity to bright-star masking. Under these circumstances, the Ks mag is set to 99.
There are only a few examples of AGN-type galaxies due to the fact that these objects tend to be distant (redshift > 0.1) and hence unresolved (to 2MASS) -- they are not "extended" sources.