The Interacting Galaxy Pair Arp 82
  Mark Hancock's page
SB&T page
Arp 82 Press Release

For more details see `Large-scale Star Formation Triggering in the Low-mass Arp 82 System: A Nearby Example of Galaxy Downsizing Based on UV/Optical/Mid-IR Imaging'   Hancock, M., Smith, B. J., Struck, C., Giroux, M. L., Appleton, P. N., Charmandaris, V., & Reach, W. T. 2007, AJ, 133, 676

Composite image of Arp 82 (NGC 2535 is the big galaxy and NGC 2536 is the smaller companion). The blue color is GALEX far-UV and near-UV, the green color is SARA Hα and R band, and the red color is Spitzer IRAC 3.6, 4.5, 5.8 and 8.0 μm. In this image north is up and east is to the left. Notice the kink in the blue tidal tail to the north.

From top (left to right): GALEX FUV and NUV, SARA Hα and R, and Spitzer 3.6 μm, 4.5 μm, 5.8 μm, 8.0 μm, & 24 μm. In all the images north is up and east is to the left. The scale bar is 30 arcseconds and corresponds to 8.2 kpc. The large galaxy to the north is NGC 2535 and the smaller companion to the south is NGC 2536.

Expanded FUV (top left) and 8.0 μm (top right) images with several star forming regions (clumps) identified. Clumps 27-30 lie along the HI arc and are shown on the FUV image. Zoomed in HI map (Kaufman et al. 1997), FUV, and NUV images of the HI arc (bottom left, middle, and right respectively). Seeing this arc clearly in the UV bands indicates the presence of hot young stars.

[3.6]-[4.5] gradient. Open squares represent clumps in NGC 2536, x's represent clumps in the bridge, stars represent clumps in the spiral and filled squares represent clumps in the counter tail. Clumps 3 and 24 are plotted as upper limits. The mean [3.6]-[4.5] color for Whitney et al. (2004) field stars is -0.05, while the predicted value for interstellar dust is -0.35 (Li & Draine 2001)

[5.8]-[8.0] vs [4.5]-[5.8]. The solid black triangle represents the entire Arp 82 system. The solid pentagons represent predicted IRAC colors for interstellar dust (Li & Draine 2001), open pentagons represent the Hatziminaoglou et al. (2005) colors of quasars, and the open 5 point stars represent the mean colors of M0 III stars from M. Cohen (2005, private communication) and field stars from Whitney et al. (2004). The median uncertainty is also shown. Clump 24 is plotted as a lower limit.

[4.5]-[5.8] vs [3.6]-[4.5]. The symbols are the same as above. Notice that clumps 23 and 26 have IR colors similar to field stars or quasars and may not be part of Arp 82.

Log[L(Hα)/L(8.0μm)] vs 8.0μm mag. The symbols are the same as before. There is a fair amount of scatter in this ratio in spite of the fact that both Hα and 8μm trace star formation.

To test whether the variations in L(Hα)/ L(8.0μm) are due in part to extinction of the Hα emission, we plot the ratio against L(IR)/L(FUV), a good proxy for extinction.

Log[L(Hα)/L(8.0μm)] vs Log[L(IR)/L(FUV)]. The symbols are the same as before. The mean uncertainty for the horizontal axis reflects only the scatter in the L(IR) calibration.

A weak anti-correlation may be present, in that the regions with lower dust obscuration tend to have higher L(Hα)/L(8.0μm) ratios. There is, however, a lot of scatter in these plots, suggesting that some of the variation in the L(Hα)/L(8.0μm) ratio is intrinsic. This scatter may be due to PAH excitation by non-ionizing photons contributing to the 8.0μm emission.

Log[L(IR)/L(FUV)] gradient. The symbols are the same as before. The mean uncertainty for the vertical axis reflects only the scatter in the L(IR) calibration.

From the L(IR)/L(FUV) ratio, it can be seen that the dust extinction is greatest near the nuclei of the two galaxies, NGC 2535 and NGC 2536, and much less in the bridge and tail regions.

FUV/R vs FUV/NUV. The symbols are the same as before. The Starburst99 models (Leitherer et al. 1999) are: solid blue line and filled blue triangles E(B-V)=0.0, dot-dashed green line and filled green triangles E(B-V)=0.2 and dashed red line and filled red triangles E(B-V)=0.6. The SB99 models were generated assuming instantaneous star formation, solar abundances, and reddened with the Calzetti, Kinney, & Storchi-Bergmann (1994) reddening law. The filled triangles plotted along the blue, green, and red curves represent ages of 1, 10, 50, 100, and 300 Myr respectively, with the youngest being at the upper right. The median uncertainty in the clump's flux ratios is shown. The top and right axes are magnitudes.

We have determined the ages and extinctions of the clumps with further population synthesis analyis with these SB99 models, the FUV/NUV, FUV/R ratios, and the EW(Hα).

Clump age vs distance from NGC 2536. The symbols are the same as before. Blue symbols are the broadband ages and red symbols are the ages determined from EW(Hα). The uncertainties in ages reflect only the uncertainties in the measured fluxes.

The clumps have mean broadband age and E(B-V) of about 9 Myr and 0.4 mag. respectively. With the exception of the two oldest clumps in the galactic nuclei (clumps 2 and16), the broadband ages for the Arp 82 clusters range from about 1 to about 30 Myr. Clumps 2 and 16 are considerably older. The mean broadband ages of NGC 2536, the bridge region, the spiral region, and the tail region are 30 Myr, 5 Myr, 15 Myr and 8 Myr respectively. From the broadband age analysis it appears that the youngest clumps are generally, but not exclusively, in the bridge and tail regions.

log[L(IR)+L(FUV)] vs distance from NGC 2536. The mean uncertainty reflects only the scatter in the L(IR) calibration.

The IR luminosity is the re-emission of absorbed UV photons, so the sum of the UV and IR luminosities is a good proxy for the total UV emission and therefore is proportional to the star formation rate (SFR). From this figure it can be seen that the SFR is greatest in the spiral region of NGC 2535 and in NGC 2536, with much less star formation in the bridge and tail regions. The lower SFR of the clumps away from the central regions could be due to the fact that less gas was dragged there as a result of the interaction. Less gas spread over a still large volume implies lower densities, lower compressions, and larger clump masses to pull together gravitationally. We have determined the SFR of the clumps using the relation SFR(Msolar yr-1) = 4.5 x 10-44 L(IR)(erg s-1) in Kennicutt (1998). The clumps have a total SFRIR of 2.0+/-0.8 Msolar yr-1. For comparison, we used the relation SFR(Msolar yr-1) = 7.9 x 10-42 L(Hα)(erg s-1) (Kennicutt 1998). The total SFR of the clumps is 1.3+/-0.2 Msolar yr-1, consistent with the SFRIR. The SFRIR of the entire Arp 82 system is 4.9+/-2.0 Msolar yr-1 vs 2.4+/-0.4 Msolar yr-1 for SFR. The total clump SFRIR accounts for about 40% of the entire system SFRIR.

From the measured R band fluxes and the ages and extinctions implied by the broadband colors, we have determined masses for each of the clumps in Arp 82 using our SB99 models. The more massive clumps tend to be found in the spiral region while the least massive clumps are in the tidal features. The clumps have a median massR of about 8 x 107 Msolar. The ten tidal clumps make up only about 3% of the total clump massR. The 2 clumps in the small companion, NGC 2536, and the 2 largest clumps in the nucleus of NGC 2535 (13 and 16) make up about 82% of the clump massR.

MassR vs L(3.6μm). The mass is determined from SB99 and measured R band flux. The symbols are the same as before.

As a test of the validity of our massR determinations, we plot the 3.6μm luminosity against massR. A strong correlation is seen in this figure. The 3.6μm band is dominated by stars, so clump fluxes in this band are a good proxy for mass.

Snapshots of the model at a time near the present. The upper panels show two orthogonal views of the gas particles. The primary's position is fixed at the origin. The dashed curve also shows the orbit of the companion center relative to the primary, starting from the point marked by a cross, and continuing through the present to the future merger. The coordinate values are given in kpc with the adopted scaling. The lower panels show gas particles that have either just turned on star formation feedback (lower right), or have recently turned on feedback, and so heated the gas (lower left). The latter thus gives a cumulative picture of recent star formation, rather than just the immediate star formation.

Shows star formation histories of the model galaxies. Specifically, the thin solid curve shows the number of star forming particles within 24 kpc of the center of the primary galaxy as a function of time. The dashed curve shows the number of star forming particles within 12 kpc of the center of the companion galaxy as a function of time, and the dotted curves show the corresponding quantities for the two galaxies in isolation. The values of the integration radii were chosen to enclose the whole disk of the individual galaxies, without including particles from the other galaxy. The bold solid curve shows the separation between the galaxy centers in kpc as a function of time, multiplied by a factor of 5 in order to use the same y-axis scale. Comparison of this curve to the thin solid and dashed curves allows one to see how the SFR depends on separation. The vertical line denotes the approximate present time.

An animation of the interaction. Red represents cool gas, Green represents hot gas, and blue represents new stars. The animation represents 2 Gyr. Notice at the beginning of the simulation the 2 disks show a low rate of star formation. The SFR has increased dramatically by the end of the simulation. In addition to reproducing a plausible star formation history, the model reproduces several key morphological features: the long tidal arm, the strong bridge, and the ocular structure. See Struck (1997) and Struck et al. (2005) for the model details.


Calzetti, D., Kinney, A. L., & Storchi-Bergmann, T. 1994, ApJ, 429, 582
Hatziminaoglou, E., et al. 2005, AJ, 129, 1198
Kaufman, M., Brinks, E., Elmegreen, D. M., Thomasson, M., Elmegreen, B. G., Struck, C., & Klaric, M. 1997, AJ, 114, 2323
Kennicutt, R. C. 1998, ARA&A, 36, 189
Leitherer, C., et al. 1999, ApJS, 123, 3
Li, A. & Draine, B. 2001, ApJ, 554, 778
Struck, C. 1997, ApJS, 113, 269
Struck, C. et al. 2005, MNRAS, 364, 69
Whitney, B. A., et al. 2004, ApJS, 154, 315

Back to Mark Hancock's page
Back to Beverly Smith's page
Back to the SB&T page