Category Archives: Profesional

A 2dF night at the Anglo-Australian Telescope

One of the most complex astronomical instruments nowadays available is the Two Degree Field (2dF) system at the Anglo-Australian Telescope (AAT, Siding Spring Observatory, NSW, Australia). The main part of 2dF is a robot gantry which allows to position up to 400 optical fibers in any object anywhere within a “two degree field” of the sky.

The 2dF instrument attached to the primary focus of the AAT. Note that the mirror of the telescope is opened. This image was chosen to be part of the Stories from Siding Spring Observatory Photo Exhibition the AAO organized last year.
Credit: Á.R.L-S.

392 optical fibers are fed to the AAOmega spectrograph, which allows to obtain the full optical spectrum of every object targeted by an optical fiber. The remaining 8 optical fibers are actually fibre-bundles and are used to get an accurate tracking of the telescope while astronomers are observing that field, which may last up to 3 hours. 2dF possesses two field plates: one located at the primary focus of the telescope and another at the position of the robot gantry. While a field is being observed in one plate, 2dF configures the next field on the other plate. A tumbling mechanism is used to exchange the plates. 2dF was designed at the AAO in the late 90s and, since then, it has been used by a large number of international astrophysicists. In a clear night, 2dF can obtain high-quality optical spectroscopic data of more than 2,800 objects.

Indeed, this sophisticated instrument has conducted observations for hundreds of astronomical projects, including galaxy surveys such as the 2dF Galaxy Redshift Survey, the WiggleZ Dark Energy Survey, and the Galaxy And Mass Assembly (GAMA), survey which is still on going and in which I actively participate. The optical fibers of 2dF can be also fed the new HERMES spectrograph, which is now starting the ambitious Galactic Archaeology with HERMES (GALAH) survey at the AAT. GALAH aims to observe around 1 million galactic stars to measure elemental abundances and measure stellar kinematics.

Frame of the time-lapse video “A 2dF night at the Anglo-Australian Telescope”. The 2dF robot gantry moving and positioning the optical fibers. Credit: Á.R.L-S.

How does 2dF move and position the optical fibers? A very nice way of explain it is using the time-lapse technique, that is, taking many images and then adding all to get a movie of the robot while moving and positioning the fibers. That is why in 2012 I decided to create the video, A 2dF night at the AAT, which assembles 14 time-lapse sequences taken at the AAT during September and November 2011 while I was working at the AAT as support astronomer of the 2dF instrument. Actually, this time-lapse video shows not only how 2dF works but also how the AAT and the dome move and the beauty of the Southern Sky in spring and summer. The time-lapse lasts for 2.9 minutes and combines more than 4000 frames obtained using a CANON EOS 600D provided with a 10-20mm wide-angle lens.

Time-lapse video “A 2dF night at the AAT”. I recommend to follow the link to YouTube and watch it at HD and full screen in a dark room. Credit: Á.R.L-S.

The video consists in three kinds of sequences created at 24 frames per second (fps). The first 3 sequences show how the 2dF robot gantry moves the optical fibers over a plate located at the primary focus of the telescope. Although in real life 2dF needs around 40-45 minutes to configure a full field with 400 fibers, the time-lapse technique allows to speed this process. The first 2 sequences have been assembled taking 1 exposure per second, therefore 1 second of the video corresponds to 24 seconds in real life. The third sequence considers an exposure each 3 seconds, and hence it shows the robot moving very quickly. The next four sequences show the movement of the telescope and the dome. All of them were obtained taking 2 images per second (a second in the movie corresponds to 12 seconds in real life). The long black tube located at the primary focus of the telescope is 2dF. The remaining sequences, all obtained during the night, were created taking exposures of 30 seconds, and hence each second in the video corresponds to 12 minutes in real life.

Frame of the time-lapse video “A 2dF night at the Anglo-Australian Telescope”. The AAT telescope, with 2dF (the long, black tube) attached at its primary focus, is prepared to start observing. Credit: Á.R.L-S.

Astronomical time-lapse videos allow to see the movement of the Moon, planets and stars in a particular position in the Earth, something that conventional videos cannot achieve. In particular, dim stars and faint sky features, such as the Milky Way with its bright and dark clouds and the Magellanic Clouds, can be now easily recorded. As in my first time-lapse video, The Sky over the AAT, I set the camera up at the beginning of the night, let it run, and check on its progress occasionally. I used at focal of f5.6 and an ISO speed of 1600 ISO for the night sequences.

Frame of the time-lapse video “A 2dF night at the Anglo-Australian Telescope”. The Magellanic Cloud rise while the Milky Way sets over the Anglo-Australian Telescope at Siding Spring Observatory on 3 Nov 2011. Some kangaroos can be seen in the ground. Credit: Á.R.L-S.

However, the procedure that took more time was processing the hundreds of individual photographies included in each sequence. In many cases, I needed more than 12 hours of computer time, including 3 or 4 iterations per sequence, to get a good combination of low noise and details of the sky, plus “cleaning” bad pixels or cosmic rays. In particular, for this video I tried hard to show the colours of the stars, a detail which is usually lost when increasing the contrast to reveal the faintest stars. In the last sequence of the video, Aldebaran and Betelgeuse appear clearly red, while the stars in the Pleiades and Rigel have a blue color.

Frame of the time-lapse video “A 2dF night at the Anglo-Australian Telescope”. A dark night at The Anglo-Australian Telescope (23 Sep 2011). Orion constellation is seen over the AAT dome. The red colour of Alderaban and Betelgeuse and blue colours of Pleiades and Rigel are clearly distinguished. Credit: Á.R.L-S.

As I did for my previous time-lapse, here I also included a sequence which shows the trails created by the stars as they move in the sky as a consequence of the rotation of the Earth. This sequence shows the Celestial Equator and stars at the South (top) and North (bottom) Celestial Hemisphere. Note that star trails have indeed many different colours. Other details that appear in this time-lapse video are clouds moving over the AAT, satellites and airplanes crossing the sky, the nebular emission of the Orion and Carina nebulae, the moonlight entering in the AAT dome, and kangaroos “jumping” in the ground.

Frame of the time-lapse video “A 2dF night at the Anglo-Australian Telescope”. Startrails over the Anglo-Australian Telescope on 23 Sep 2011. The colours of the stars are clearly seen in this image, which stacks 1h 6min of observing time. Credit: Á.R.L-S.

Finally, I chose an energetic soundtrack which moves with both 2dF and the sky. It is the theme Blue Raider of the group Epic Soul Factory, by the composer Cesc Villà. Actually, all sequences were created to fit the changes in the music, something that also gave me some headaches. But I think the result was worth all the effort.

SN2014J in M82 observed at the William Herschel Telescope

A week ago, on January 21st, the English astrophysicist Steve Fossey gave a telescope workshop for a group of undergraduate students (Ben Cooke, Tom Wright, Matthew Wilde and Guy Pollack) belonging to the University College of London (UCL). As usually happens in the British capital, the sky was practically covered by clouds. However, Fossey and his students used the automatic 35 cm telescope of the University of London Observatory to spot the famous starburst galaxy M 82. Located at 12 million light-years away in the constellation of Ursa Major (The Big Dipper), the galaxy M 82 hosts an intense star-formation burst, being its light dominated by young, hot, massive, blue stars. As consequence of this frenetic activity, M 82 possesses long jets of hot gas that has been expelled from the center of the galaxy. Therefore, it is not casual that the students chose this galaxy as a target for their assignment. While Fossey was centering the galaxy in the field of the telescope he realized that there was a bright star which should not be there. They checked that this new star was real using another telescope of the Observatory. As clouds were approaching, they quickly took some few images in different filters. The first analysis was doubtless: they had just discovered a supernova in the galaxy M 82.

Discovery image of type Ia SN2014J in the starburst galaxy M82 (below) compared with an older image of the galaxy before the supernova exploded (top). The discovery image was obtained at 19:20 UT, 21st January 2014 using the automatic 35 cm telescope of the University of London Observatory. Credit: UCL/University of London Observatory/Steve Fossey/Ben Cooke/Guy Pollack/Matthew Wilde/Thomas Wright.

In just one day, amateur astronomers and professional astrophysicists used their telescopes to study M 82. These observations soon confirmed the discovery made by Fossey and his students. Actually, some astronomers even found that they had taken data of the galaxy and the supernova a week before the official discovery, but the new exploding stars was unnoticed by them. A couple of days after the discovery, a group of astrophysicists led by Yi Cao (Caltech) got the first optical spectrum of the supernova using the 3.5m ARC Telescope at Apache Point Observatory (New Mexico, USA). The analysis of this spectrum showed that the progenitor of the supernova was a white dwarf, and hence the explosion was classified as a type Ia supernova. The official name of this exploding star is SN 2014J. It has not reached its maximum brightness yet: when Fossey and his students discovered the supernova, it was 2 weeks before when we expect this happens. Right now it is so bright (around 10th magnitude) it is very easy to spot using a small amateur telescope. Perhaps even it can be seen using binoculars when the supernova reaches its maximum brightness in a week or so!

Hence, it is not difficult to understand that SN 2014J and M 82 have been the main astronomical news in the last week. Using the 4.2m William Herschel Telescope (WHT), which is part of the Isaac Newton Group, located at the Roque de los Muchachos Observatory in the beautiful island of La Palma (Canary Islands, Spain), the astrophysicists Manuel Moreno-Raya (CIEMAT, Spain) and Lluís Galbany (DAS/UC, Chile) have observed with great detail both the supernova and the galaxy. Between Thursday 23rd and Sunday 26th January they used the ISIS spectrograph, as well as the ACAM instrument (Auxiliary-Port Camera), of the WHT to get images and spectra of the supernova. I was continuously in touch with them as I’m part of their research team (actually, I’m co-supervising the PhD thesis which is conducted by Manu). I originally planned to travel to La Palma to be helping on these observations, however this was colliding with my support activities at the Anglo-Australian Telescope (Siding Spring Observatory, NSW, Australia). Manu and Lluís sent me the data as they were coming from the WHT, and I was reducing, combining, and getting the preliminary images and spectra of this object!

The image below shows the supernova SN 2014J and the galaxy M 82 using the data obtained with ACAM. I tried to get all the important details of this puzzling object: the dust lanes crossing the disc (dark-yellow), the strong star-formating bursts (blue) and even the filamentary structure of the super-galactic wind that M 82 possesses (in red). This feature is hot, ionized gas which has been expelled from the center of the galaxy and here it is seems perpendicular to the galactic disc. SN 2014J very brightly shines at the west (right) of M 82 galactic center.


Colour image of starburst galaxy M 82 with the type Ia supernova SN 2014J. M 82 lies at 12 million light years from us, in the Ursa Major constellation. The supernova is marked with two white lines. The data needed to get this image were taken using the ACAM instrument located at the Cassegrain focus of the 4.2m William Herschel Telescope (WHT) (Roque de los Muchachos Observatory, La Palma, Canary Islands, Spain). We got data in u, g, i, r, and Hα filters. Data coming from the u filter (2 x 200 seconds exposures) are colour-coded in blue; data in the g filter (3 x 100 seconds exposures) are colour-coded in cyan; data in the i filter (3 x 100 seconds exposures) are colour-coded in green; data in r filter (3 x 300 seconds exposures) are colour-coded in red. The majority of the data were obtained last 24th January, at 04:40 UT. Data in r and u filter were taken on 25th January, at around 06:00 UT. The Hα data (4 x 300 seconds exposures), which are colour-coded in red, were taken on 26th January at 06:30 UT. Data coming from the Hα filter clearly reveals the super-galactic wind of M 82. All data were reduced and combined using standard IRAF routines. The colour composition was obtained using Photoshop. The field of view is 8 arcminutes and the resolution 0.25 arcsec/pixels. However, the seeing was not too good, between 2 and 5 arcsec.
Credit: Observers: Manuel E. Moreno-Raya (CIEMAT, Spain) & Lluís Galbany (DAS / UC, Chile). Data processing and color image composition: Ángel R. López-Sánchez (AAO / MQ, Australia). Support astronomer: Chris Benn (ING, UK), Telescope Operator: José Norberto González (ING, UK). Research Team: Manuel E. Moreno-Raya (CIEMAT, Spain), Mercedes Mollá (CIEMAT, Spain), Ángel R. López-Sánchez (AAO / MQ, Australia), Lluís Galbany (DAS / UC, Chile),Aurelio Carnero (ON, Brazil), Inma Domínguez (UGR, Spain), & Pepe Vílchez (CSIC / IAA, Spain).

In addition, we have already analyzed the low-resolution spectrum of the SN 2014J obtained using ACAM. This spectrum gets all the optical range, between 3500 and 9500 Angstroms, and clearly identifies the object as a type Ia supernova. The main features are absorption bands of iron (Fe II and Fe III), magnesium (Mg II) and silicon (Si II) between 4000 and 5000 A. These bands actually are blends of absorptions due to these metallic elements. Indeed, astrophysicists expect the intensity of these bands will be changing as the supernova evolves, as the chemical abundances and ionization of each species vary as some elements are converted into others and more material coming from the center of the dead star is observed. Even so, it is a surprise to find these absorption bands almost 10 days before the supernova reaches its maximum brightness. The spectrum also shows absorptions of sulfur (S II) at 5240 and 5450 A, a strong absorption by silicon (Si II) at 6150 A, and absorptions of calcium (Ca II), sodium (Na I) and oxygen (O I). Some features are actually created in the Earth atmosphere and hence they do not belong to the supernova, these are labelled as “Tel” (from “Teluric lines”). However, the feature which interested us most was the carbon absorption (C II) at 6580. This line indicates that the progenitor of the supernova was a white dwarf composed by carbon and oxygen (as it happens in the majority of the white dwarf). However, it is uncommon to observe this line in type Ia spectra. This suggests that the surface of the white dwarf has not been completely burnt during the explosion. All absorption lines are found “blue-shifted”, that is, at shorter wavelengths that those expected. That is a consequence of the high speed at which the material is moving, expanding fast away from the dead star. The measurement of the C II and S II lines observed in our ACAM optical spectrum indicates that this material is moving at around 15 000 km/s!


Low-resolution optical spectrum of the type Ia supernova SN 2014J discovered in the galaxy M 82 obtained using the ACAM instrument at the Cassegrain focus of the 4.2m William Herschel Telescope (WHT) (Roque de los Muchachos Observatory, La Palma, Canary Islands, Spain). The intensity or relative flux (“Arbitrary Flux”, vertical axis) is plotted versus wavelength (“colour”, horizontal axis). The main features, which includes absorption lines of iron, magnesium, silicon, sodium, calcium, oxygen and carbon, are labelled. The spectrum combines two expositions of 200 seconds each using the ACAM V400 grism. The data were obtained last 25th January at 7:10 UT, which approximately corresponds to Epoch -11 days. It is expected the supernova reaches its maximum brightness in that time. The reduction of the data and the wavelength calibration was performed using standard IRAF routines.
Credit: Observers: Manuel E. Moreno-Raya (CIEMAT, Spain) & Lluís Galbany (DAS / UC, Chile). Data processing and color image composition: Ángel R. López-Sánchez (AAO / MQ, Australia). Support astronomer: Chris Benn (ING, UK), Telescope Operator: José Norberto González (ING, UK). Research Team: Manuel E. Moreno-Raya (CIEMAT, Spain), Mercedes Mollá (CIEMAT, Spain), Ángel R. López-Sánchez (AAO / MQ, Australia), Lluís Galbany (DAS / UC, Chile),Aurelio Carnero (ON, Brazil), Inma Domínguez (UGR, Spain), & Pepe Vílchez (CSIC / IAA, Spain).

Interestingly, the project that Manuel Moreno-Raya (CIEMAT, Spain) and his research team, composed by Mercedes Mollá (CIEMAT, Spain), Lluís Galbany (DAS / UC, Chile), Aurelio Carnero (ON, Brazil), Inma Domínguez (UGR, Spain), Pepe Vílchez (CSIC / IAA, Spain) and myself, was observing at the WHT was focused in obtaining deep, high-quality data of galaxies hosting type-Ia supernova. The idea is to quantify the physical and chemical properties of these host galaxies with the final aim of getting a better understanding of the parameters which control the brightness of these supernovae and apply these new measurements to improve the accuracy to very distant galaxies. This research is the main part of the PhD thesis project that Manu is conducting. Besides the observations of M 82 and the SN 2014J, we also got deep intermediate-resolution optical spectroscopy data of around 20 galaxies. These data still have to be analyzed in detail, something that will take months.

SN 2014J is the type-Ia supernova closest to the Earth since that Johannes Kepler observed in 1604. The Kepler’s Supernova actually exploded in our Galaxy, at just 20 thousands light-years from us, and it was so bright it was seen with the naked eye, being the brightest object in the sky after the Sun and the Moon. The type Ia supernova SN 1972e was also very close to us, as it exploded in the dwarf galaxy NGC 5253 (*). NGC 5253, which lies at a distance of 13 million light years, is in some way a similar object to M 82, as it also hosts a very powerful star-formation event. SN 1972e became the prototype object for the development of theoretical understanding of Type Ia supernovae, but this position may change with all the data that are coming from SN 2014J. What surprises will provide this new supernova? Can the new data be used to get a better understanding of the type Ia supernovae as a cosmological distance estimators and help to discover the nature of the mysterious dark energy which induces the expansion of the Universe? This research has just started.

UPDATE: Part of the information included in this post was used to prepare a telegram for ATel, The Astronomer’s Telegram, number 5827, Broad and narrow band imaging and spectroscopic follow up of SN2014J in M82, published on 28 Jan 2014; 18:30 UT.

(*) I should tell you many more things about the dwarf galaxy NGC 5253… It was my nightmare for some few years and after performing a very complete and detailed multi-wavelength analysis of this weird object I’m still not sure what is happening in there!

NEA 2012 DA14 observed from the AAT

Yesterday I explained we are resuming the scientific observations at Siding Spring Observatory (NSW, Australia) since the recent bushfires on 13 January 2013. Today we have used this telescope to observe the Near Earth Asteroid (NEA) 2012 DA14, which is getting its closest encounter to our planet today (it will be at only 27000 km from the surface of the Earth). Here you have the details:


The path of the NEA 2012 DA14 from the AAT. Observed during the evening twilight on the 15 February 2013 using the FPI camera of the 2dF instrument at the 3.9m Anglo-Australian Telescope located at Siding Spring Observatory (NSW, Australia). 25 frames, 2 seconds exposure each, 5 seconds between frames. First frame obtained at 09:53:54 UT, last one at 09:56:22.
Support Astronomers: Lee Spitler (MQ/AAO) & Andy Green (AAO).
Night Assistant at the AAT: Steve Chapman (AAO).
Edition of the data and movie: Ángel R. López-Sánchez (AAO/MQ).
Link to the video in my Flick.

The data were obtained by Lee Spitler (MQ/AAO), Andy Green (AAO) and Steve Chapman (AAO) during the evening twilight while waiting the sky was dark enough to start the scheduled scientific observations. We will try to get more data of this object during the morning twilight. More to come soon!

Update at 14:45 AEST, 16th Feb 2013

I’ve checked there are some problems to watch the video using smartphones and tablets such iPhones and iPad. Just try this link and it should work, although you may get a low resolution version of it. I’m creating a new video to be uploaded to YouTube. At the same time, I’ve just finished this image showing a mosaic with 13 of the frames obtained for the video.

Mosaic with the path of the NEA 2012 DA14 from the AAT. Observed during the evening twilight on the 15 February 2013 using the FPI camera of the 2dF instrument at the 3.9m Anglo-Australian Telescope located at Siding Spring Observatory (NSW, Australia). Each frame has an exposition time of 2 seconds.Times are shown in Universal Time (UT), add +11 hours to get the local time in NSW.
Support Astronomers: Lee Spitler (MQ/AAO) & Andy Green (AAO).
Night Assistant at the AAT: Steve Chapman (AAO).
Edition of the data and movie: Ángel R. López-Sánchez (AAO/MQ).


Finally, I confirm we also observed the NEA at the AAT during the morning twilight.

Update at 16:55 AEST, 18th Feb 2013

An updated version of the video, which is now in HD and includes the celestial coordinates, has been included in this excellent article published today in The Conversation by Simon O’Toole (AAO).


The path of the NEA 2012 DA14 from the AAT. Observed during the evening twilight on the 15 February 2013 using the FPI camera of the 2dF instrument at the 3.9m Anglo-Australian Telescope located at Siding Spring Observatory (NSW, Australia). 25 frames, 2 seconds exposure each, 5 seconds between frames. First frame obtained at 09:53:54 UT, last one at 09:56:22.
Support Astronomers: Lee Spitler (MQ/AAO) & Andy Green (AAO).
Night Assistant at the AAT: Steve Chapman (AAO).
Edition of the data and movie: Ángel R. López-Sánchez (AAO/MQ).

Back observing at the Anglo-Australian Telescope

On 13th January 2013 the Siding Spring Observatory and the beautiful Warrumbungle National Park near Coonabarabran (NSW, Australia) were terribly affected by the worst bushfire in NSW in the last decade. Although the astronomical facilities have not experienced any severe damage, the bushfire has destroyed some houses at the Observatory (including the Lodge), burnt tens of houses and destroyed the majority of the trees in the National Park.

However, tonight Thursday 14 February, after a month and a day since the bushfire, astronomers are recommencing observing with the 3.9-m Anglo-Australian Telescope at Siding Spring Observatory. I’m one of these astronomers who are now performing the observations remotely from the Australian Astronomical Observatory (AAO) headquarters in North Ryde, Sydney, supported by technical staff at the telescope. The AAO has made public today a press release informing that astronomers are back to work at the AAT!.

The Spindle Galaxy with the AAT. It is an edge-on lenticular galaxy classified as NGC 3155 or Caldwell 53. The data were obtained on 14 February 2013 using the FPI camera of the 2dF instrument at the 3.9m Anglo-Australian Telescope located at Siding Spring Observatory. 8 x 20 s + 5 x 40 s + 1 x 60 s integration time (460 s), combined with IRAF. Colours derived using U, V and I images obtained at the 2.5m Cerro Tololo International Observatory by Kuchinski et al. (2000).
First astronomical observations after the bushfires on 13 January 2013.
Credit: Ángel R. López-Sánchez (AAO/MQ) & Lee Spitler (MQ/AAO),
Night Assistant at the AAT: Steve Chapman (AAO)
.

Although we are using tonight the Two Degrees Field (2dF) instrument with the AAOmega spectrograph, which allows the acquisition of up to 392 simultaneous spectra of objects anywhere within a two degree field on the sky, we have also used the auxiliary camera that 2dF possesses, the Focal Plane Imager, to take some images of the Spindle galaxy, also known NGC 3115 or Caldwell 53, a lenticular (S0) galaxy located at around 32 million light years from Earth.

However, tonight’s observations are having the AAT looking up to a billion light-years out into space to test our ideas about the still mysterious Dark Energy.

The Crescent Nebula

A very nice example of a nebula surrounding a Wolf-Rayet star is the so-called Crescent Nebula (NGC 6888, Caldwell 27, Sharpless 105). Located in the northern constellation of Gygnus, The Swan, it lies at around 5000 light years from us. The Crescent Nebula has been formed by the strong stellar winds of the Wolf-Rayet star WR 136 (HD 192163), which is located in the center of the nebula. This is an image of the Crescent Nebula I took in 2004 using the 2.5m Isaac Newton Telescope (INT) at the Roque de los Muchachos Observatory (La Palma, Spain) while I was still preparing my PhD Thesis at the Instituto de Astrofísica de Canarias (IAC, Tenerife, Spain) about the properties of dwarf galaxies hosting Wolf-Rayet stars. Actually, the image was taken during the twilight, when sky is still dark enough the get details in the narrow-band filters.

Crescent Nebula using narrow-band filters, by Angel R. Lopez-Sanchez

Image of the Crescent Nebula (NGC 6888) obtained by the author combining data using the broad-band optical B filter (blue) and the narrow-band optical filters [O III] (green) and Hα (red) obtained using the Wide Field Camera (WFC) attached at the 2.5m Isaac Newton Telescope (INT) at the Roque de los Muchachos Observatory (La Palma, Spain). The size of the image is around 22 x 22 arcminutes, just slightly smaller than the field of view of the full moon in the sky (30 arcminutes in diameter). Credit: Ángel R. López-Sánchez

The complex structure of the Crescent Nebula is a consequence of the interaction of the strong wind of the Wolf-Rayet star with material ejected by the star in an earlier phase, probably while it was a red supergiant. The actual loss-mass rate of the WR136 is around 0.00001 solar masses per year, which means the star losses the equivalent of the Sun’s mass every 10,000 years.

The image clearly shows ionized gas (nebular emission) with very different conditions: while red-color (Hα emission) is tracing the normal, emitting ionized gas, the green colour ([O III] emission) indicates regions with high excitation of the gas, meaning higher temperatures probably because of shocks. In just some few hundreds of years the star will explode as type-II supernova and destroy all the nebula, although it will create a new object: a supernova remnant.