The interacting triplet M81, M82, and NGC 3077- An ultra deep 219 hour collaboration detailing the nuanced interaction remnants and galactic cirrus.

This incredibly detailed image of the interacting triplet M81, M82, and NGC 3077 was created from more than 216 hours of exposure across three continents —  Europe, USA, and Oceania. We combined 4019 subframes and 12993 minutes of exposure to reveal faint details previously masked by lower amounts of data. The high exposure time also allowed us to sharpen fine features, giving the image more contrast and revealing more fine structures.

On the top of the image lies NGC 3077, a small starburst dwarf galaxy with a starforming core. Below lies M81, a grand design spiral about 12 million light years away. To the right is M82, a starburst galaxy with a huge superwind triggered by interaction with M81 and NGC 3077. All around the image is the galactic cirrus, dust lit by the glow of the milky way. I highly recommend looking at the image with HI emission overlaid, which displays what of the background is neutral gas from the interactions and what is dust from the milky way.

The images below are zoomable! Use your mouse or fingers to zoom and pan.

Image:

Full image here: https://live.staticflickr.com/65535/52246306840_153c2f9cb7_o.png

Image with HI emissions:

Full image here: https://live.staticflickr.com/65535/52246096294_f4f7c241af_o.png

Annotated Image:

Full image here: https://live.staticflickr.com/65535/52245829093_77c09f4620_o.png

Interesting Features

The details noted here were inspired by the Big Amateur Telescopes image of this region. You can see more here: https://www.astrobin.com/users/Big_Amateur_Telescope/

Below is a comparison of an inverted luminance image to show the cirrus surrrounding these galaxies (black and white) and VLA’s HI emission obervations (red). As you can see, some of the IFN structure corresponds to the HI emission, especially around M81 and M82. Most likely these IFN-like structures are neutral/slightly ionized gas remnants from the original interaction between M81 and M82. It is also interesting to note the volumes of gas stripped by NGC 3077 from M81. Without better velocity maps it is hard to make a good analysis of this interaction, however the most likely explanation is that the initial M81-M82 interaction freed gas from the spiral arms of M81, resulting in the striking gas bridge. Blok et al. (2018) contains a strong analysis of these velocities: https://arxiv.org/pdf/1808.02840.pdf.

A closer look at M82 shows some interesting detail. The first interesting detail is a stellar stream coming off M82 at a 60 degree angle. It’s hard to know the origin of this stream – it could either be an interaction remnant or a shocked starburst stream. Most likely, this is a secondary result of a galactic low-velocity shock caused by interaction with M81. In short, the shock heats and compresses the gas to optimal star-forming conditions. The outflows then eject in an AGN-like manner, resulting in the signature dual emission. This physical process is a likely way that E+A galaxies are formed (which I am currently researching!). If you want to learn more about outflows, consider reading this paper: https://ui.adsabs.harvard.edu/abs/2010ApJ…721..505R/abstract

The second interesting feature is a H-alpha “cap” off to the bottom right of the galaxy. Devine and Bally (1999) found that it could either be a bowshock formed by the impact of the superwind, or it could trace inoized lymann alpha emission leaking from the M82 nuclear region through the hot bipolar cavity produced by the starburst-driven superwind. They concluded that the most natural explanation was the impact of the superwind, however it was hard to tell without more detailed spectroscopic analysis. If you want to learn more, you can do so here: https://iopscience.iop.org/article/10.1086/306582

Finally, while often overlooked, there are hundreds of tiny galaxies in the background of this image. In fact, any spot of light that does not have a white core in this image is most likely a galaxy! Below is an unnamed galaxy cluster around the galaxy PGC 2732338. See if you can find it in the annotated image above! (Hint: it’s near M82)

Data Credits:

Acquisition: Andre Vilhena (@the.cosmic.arena), Dominic (@domnuch), Jens Unger (@jazz.astro), Justin P. (@justadudewitha_camera_), William Ostling (@the_astronomy_enthusiast), Steve Gill (@parkesburg_observatory), Ovidiu Dascalu (@ovidiudascaluwu)

HI data: W. J. G. de Blok et al 2018 ApJ 865 26

Image processing: William Ostling (@the_astronomy_enthusiast), Justin P. (@justadudewitha_camera_)

Data Details:

Total: 219 hours 33 minutes, 4019 subframes

Ha: 56 hours 50 minutes, 608 subframes

L: 80 Hours 15 minutes, 2237 subframes

R: 27 hours 53 minutes, 361 subframes + 119 OSC

G: 27 hours 56 minutes, 351 subframes + 119 OSC

B: 28 hours 8 minutes, 352 subframes + 119 OSC

Processing:

Stacking
I want to give a huge thank-you to Justin (@justadudewitha_camera_), who spent over 62 hours stacking this dataset!
- OSC data was split into RGB channels with VNG debayer
- The subframes were star aligned with drizzle data
- All subframes were weighted using the following formula:
33*((PSFSignalWeight-PSFSignalWeightMin)/(PSFSignalWeightMax-PSFSignalWeightMin))
+ 18*(1-(Eccentricity-EccentricityMin)/(EccentricityMax-EccentricityMin)))
+ 18*(1-(FWHM-FWHMMin)/(FWHMMax-FWHMMin))
+ 21 *(SNRWeight-SNRWeightMin)/(SNRWeightMax-SNRWeightMin)
+ 10
- A normalization reference was created for each channel by integrating the top 15 (lum) or 10 (HaRGB) weighted subframes, then ABE func 2 was applied.
- Local norm with a scale of 768 was applied to all the subframes
- Subframes were integrated using linear fit clipping
- Subframes were drizzled with a scale of 2, drop shringk 0.78, grid=16, varshape 1.5
Linear Processing
- All the master frames were cropped
- A broadband RGB image was created
- Background Extraction
      - The RGB image background was flattened based on techniques outlined here: https://pixinsight.com/tutorials/multiscale-gradient-correction/index.html
      - L and Ha were flattened using DBE
      - The RGB image was then photmetrically color calibrated to a SA reference
- H-alpha addition
      - Red image was extracted from the RGB image
      - Red image was denoised using NoiseX
      - H-alpha image had stars removed
      - H-alpha image was denoise using NoiseX
      - The denoised H-alpha image was contiuum subtracted
      - The subtracted H-alpha was added to RGB and L
- Deconvolution of L_Ha
      - A PSF was created with the following values:
N	50 (679)
B	0.000053
A	0.025351
sx	9.564812
sy	9.564137
theta	0.000000
beta	3.018283
r	0.999929
FWHMx	9.72
FWHMy	9.72
MAD	0.000353 (0.000005 .. 0.009717)
      - The stars were removed from the luminance image
      - The starless luminance image was deconvolued using the Regularized Van Cittart algorithm
      - The stars were added back
- The HaRGB image was color-calibrated again using the same settings
- The Repaired HSV separation script was applied to the HaRGB image
- Noise reduction
      - The HaRGB image was denoised using noisex (targeting luminance) and MMT (targeting chrominance)
      - the HaL image was denoised using noisex
Non-Linear Processing
- RGB was stretched using GHS, targeting RGB/K and color
- Luminance was stretched using GHS targeting RGB/K
- LRGB combination
- Clipped the tail of the histogram
- Masked saturation adjustments
- Color-corrected HDR targeting the galaxies
- More masked saturation adjusments
- Color balancing using curves and a large-scale mask
- Starless enhancement
      - A starless copy of the image was created
      - The starless copy was brightned using exponential transform in Screen Mask Invert mode
      - the starless copy was screen-blended back into the image
- Slight non-linear noise reduction using noisex
- Slight Color balance and saturaiton adjustments
- Color-corrected HDR
- Background neutralization and levels set to .11
- Curves transformation