Positive detection of the Uranus ring using a 0.5m amateur telescope
Phil Miles, Rubyvale QLD Australia
Anthony Wesley, Murrumbateman NSW Australia
December 13 2016
Last update: January 10 2017
Background
During 2016 Phil and I had many conversations about Uranus and the possibility for amateur-class instruments
like his 508mm newtonian scope to directly image its extremely faint ring system. This seemed very
optimistic as the smallest scope to have done this previously was an 800mm aperture, and that was done at a time
when it was thought that the rings would be brighter than normal. [1]
The ring system was edge-on to the earth in 2007, and this time would probably have been the best for trying
to directly image it as the reflected light would be most concentrated. By this year (2016) the rings have
opened appreciably, lowering their apparent surface brightness and making detection by smaller instruments
more difficult. The rings will continue to open and dim until around the year 2029 when the ring system will
be fully face on to the earth and then they will start to close up and increase in brightness after that time.
However, after seeing the excellent images taken at the Pic Du Midi 1.06 metre telescope in 2014/16 [2] [3] [4]
we started to think that it was possible to directly image the rings even in their current position.
The Pic Du Midi images were taken with similar filters and a comparable camera to Phils and comparing the
brightness of the rings to moons such as Miranda made us believe that it was within reach, given rare
near-perfect seeing.
A lot of time and effort was spent over the last 18 months by both Phil and myself to optimise his scope and
imaging setup in ways that we hoped would maximise our chances of capturing this very faint target. In particular
we hoped that the newtonian design of his scope with various "planetary" optimisations would minimise scattered
light. To this end he has a small secondary mirror (only 22% of the primary diameter) and no coma correctors or field
flatteners, so his scope is tuned toward high resolution, narrow field, planetary imaging.
On December 13 2016 Phil found an evening of excellent seeing and recorded eight video sequences on
Uranus with various near-IR filters in an attempt to capture the ring. He set the gain much higher than normal
and exposure much longer to maximise the captured light. As a result the planet is very over-exposed but
in later analysis the ring system is detected in every one of these recordings, in both filters.
Data reduction and analysis was done by me (Anthony Wesley) on the following days. It was immediately clear that
we had captured *something*, but it took a while longer to gain confidence that it was indeed the ring system
that we were seeing, almost lost in the glare from Uranus.
Equipment
Phil Miles 508mm (20") f/4 planetary newtonian
5x Televue Powermate + Prostar focal reducer
Astrodon 700nm longpass filter
Baader 610nm longpass filter
PGR GS3-U3-32S4M video camera (imx252)
Results
Data captured using Firecapture as follows:
The 700nm filter sequences were recorded for 15 minutes each @ 5fps with gain 48db (max).
The 610nm filter sequences were recorded for 15 minutes each @ 5fps with gain 36db
Here are sample raw frames from each of these recordings. Uranus is approximately 3.5 arc-seconds in diameter
and measures approximately 44 pixels across in these frames. This gives an image scale of about 12 pixels per
arc-second.
 |  |
| 700nm NIR longpass single video frame | 610nm NIR longpass single video frame |
The video data was analysed in several ways as we didn't know if the ring would be visible or how faint it
might be compared to the glare from the planet. Our initial analysis was done by combining several
video runs into a single sequence to maximise any possible signal.
Analysis 1: Combining 3 x video sequences through the 700nm filter for a total of 45 minutes data
Analysis 2: Combining 5 x video sequences through the 610nm filter for a total of 75 minutes data
Analysis 3: Processing each video sequence individually to find the one with the single clearest result
All of these analysis techniques show the ring, however the atmospheric seeing plays a dominant effect
on the clarity of the result so in the end we found the clearest detection in just one of the 610nm
individual sequences that was recorded in the steadiest seeing.
Each processed image is compared to a simulated view that shows the location and orientation of the
ring system as well as the moons Ariel, Miranda and Umbriel as reference. Each simulated view used here was
generated for the time of the respective images. Here is the simulated view for 1136UTC.
These reference images are generated using this online resource from the SETI institute.
Analysis 1 - 45 minutes of data @ 5fps, 700nm longpass filter
 |
45 minutes of data @ 5fps, 700nm longpass filter.
Stacked and enhanced result. 10925 frames |
 |
Above image overlaid with the reference image to show ring position.
Simulated image has been scaled and rotated using the moons for reference. |
Analysis 2 - 75 minutes of data @ 5fps, 610nm longpass filter
 |
75 minutes of data @ 5fps, 610nm longpass filter. Stacked and enhanced result. 18623 frames |
 |
Above image overlaid with the reference image to show ring position.
Simulated image has been scaled and rotated using the moons for reference. |
Analysis 3 - Single best recording, 15 minutes of data @ 5fps, 610nm longpass filter
 |
15 minutes of data @ 5fps, 610nm longpass filter. Stacked and enhanced result. 4500 frames |
 |
Above image overlaid with the reference image to show ring position.
Simulated image has been scaled and rotated using the moons for reference. |
Notes
* The ring is detected in each of these analyses. It is most clearly seen in analysis #3.
* Due to the proximity of the ring to Uranus and its extremely low brightness only the ends of the
ring (ansae) are detected, with the interior lost in the glare and scattered light from Uranus.
* It is noticeable that the lower-left ansae (at 7 o'clock position) is more easily seen than the
opposite end at 1 o'clock. This is not fully understood, but might be explained by either the
known brightness variability in the ring and/or some optical scattering asymmetry in the instrument.
Many of the online images that we consulted for reference (including those captured by HST) show a
marked unevenness in brightness around this ring. It may be that we found a night where the brightest
part of the ring was close to the lower left ansae, increasing its visibility.
* The approach used here (NIR longpass filters) is different to the more generally suggested system of
attempting to image the rings using a narrowband CH4 filter to reduce glare from Uranus itself. The CH4
filter method is used successfully on larger telescopes, however amateur-class instruments generally
do not have enough light gathering ability (ie aperture) for this to be viable as the resulting images
are extremely dim. The long exposure required will result in increased image blur due to movement
in both the instrument and the atmospheric seeing, with the resulting loss in effective resolution possibly
greater than the benefit in contrast from the CH4 filter.
* The successful detection of the rings using NIR pass filters shows an alternative that should be viable for
smaller aperture scopes. Given the results shown here it seems reasonable to conclude that smaller instruments
should be able to detect this ring under conditions of near perfect seeing. At a guess it might be possible for
telescopes down to about 14 inch aperture to successfully detect the very outer edges of the ring using the
generally available NIR longpass filters (610nm / 685nm / 700nm).
references
Note all images are (C)Copyright to their respective owners
[1] Previous detection of Uranus ring using an 800mm telescope in 2011
[2] Uranus ring detection by JL.Dauvergne / F.Colas using the Pic Du Midi 1.06m scope in 2016
[3] Uranus ring detection by M.Delcroix / C.Pellier / JP.Cazard / F.Colas using the Pic Du Midi 1.06m scope in 2016
[4] Uranus ring detection by M.Delcroix / F.Colas using the Pic Du Midi 1.06m scope in 2014
A second positive detection from Damian Peach
A second positive detection has come from Damian Peach. Here is his image and description:
"After seeing the email from Anthony Wesley and Phil Miles regarding detection of the rings of Uranus I had a look back
at my best session near opposition in 2015. Seeing was excellent and I spent some time collecting data on the night.
Turns out they were also detected on this session.
I'd never thought to look for them before until now as just assumed they were too dim. The new IR sensitive
cameras however are now bringing such difficult targets into view.
This is around an hours worth of data through an RG610 filter. Around 60,000 frames @ 10fps with ASI224."
Footnote: Damian was using a 14" SCT at a time close to opposition when the rings would likely have been
brighter than normal.
A second detection by Phil Miles, January 9 2017
On January 10 2017 Phil in Rubyvale once again found a period of very good seeing coinciding with Uranus
at a high elevation. He recorded several runs through a 685nm longpass filter but only one of them (the first
one) was in good seeing and shows the ring quite clearly. Once again this brings home to us just how
sensitive this feature is to good seeing.
First up is a link to a heavily processed image showing the ring:
2017-01-09-0952_9-PMRQA-685.png
This is the best 85% of a 15 minute recording made at a speed of 5 frames per second with gain set for a
max histogram of about 90%.
To avoid any "blink bias" I'd recommend that you take some time to examine this static image and decide
for yourself if it shows part of the ring, and where it is in the image. (Hint, it's very close in to the
planet, and both faint and narrow, extending barely out beyond the scattered light from Uranus itself).
Once you think you know where it is, use this blink animation to check:
2017-01-09-0952_9-PMRQA-685-blink.gif
Notes about the ring
We've both been reading more about the ring, in particular its peculiar aspect in showing more clearly
on one side of the planet than the other. The best explanation I've come across is given on the "Rings of Uranus"
Wikipedia page. The ring we see here is the e-ring (epsilon ring) which is both highly eccentric and also
very nonuniform in width. At its closest point to Uranus (periapsis) it is about 20km in diameter and
very dim due to mutual shadowing of the ring particles, while at the furthest point from Uranus (apoapsis)
the ring widens to about 96km across and appears about 3x brighter than the periapsis segment.
If you look at various images of Uranus on the net that show the ring then this brightness disparity is
clear to see in almost all of them.
Notes about the instrument setup
After the initial detection on Dec 13 last year Phil has added significantly more blackening and baffling
inside the scope to reduce scattered light, particularly around the primary mirror. We've also
changed barlows from the TV 5x Powermate + 0.8x Focal reducer that was in the scope on Dec 13 to now
having just a 4x TV powermate. This has changed the image scale slightly but means removal of some glass
elements and so a bit less scattered light.
Phil also has rotated the camera by about 15 degrees compared to its previous orientation after some
requests from people in the amateur community to be sure the ring isn't some artifact generated in the camera.
Phil used a 685nm longpass filter for this image, compared to the 610nm and 700nm longpass filters
used previously, and also lower gain to give a histogram of about 90% when capturing.
I hope people will understand that these changes mean that this second detection is very unlikely to be
caused by artifacts in the scope, equipment or camera. Particularly as this detection seems more obvious
than the earlier one.
