Skip to main content

Supernova Photometry Chat with Dr. Michael Richmond

On 13 September, 2011, we had the good fortune to have Dr. Michael Richmond, Physics Professor and Observatory Director for the Rochester Institute of Technology in New York chat with us about SN 2011fe and supernova photometry in general. Here is the transcript from that chat session.


<DocKQR>    OK, folks. Let's get this started. Good morning (trust me, its morning somewhere!)
<DocKQR>    Today we have with us Dr. Michael Richmond, Professor of Physics at the Rochester Institute of Technology.
<DocKQR>    Michael is also the Director of the RIT Observatory.
<MichaelRichmond>    Hello, everyone.
<DocKQR>    He has been lighting up the AAVSO Supernova Forum now that SN 2011fe has gone off and we invited him here to chat with us this afternoon
<MichaelRichmond>    I've been interested in studying supernovae since I was a grad student
<MichaelRichmond>    (which is far too long ago now)
<DocKQR>    to discuss Supernovas, supernova photometry in general, and his observations of SN 2011fe.
<MichaelRichmond>    and I'm doing my best here in cloudy Rochester, NY to follow SN 2011fe.
<DocKQR>    I'll let Michael talk now. Then we'll want to take some elementary questions.
<MichaelRichmond>    I'd be happy to try to answer questions you might have about this particular SN
<MichaelRichmond>    ("SN" is an abbreviation for "supernova", and "SNe" for the plural)
<Andy>    Has it reached peak?
<MichaelRichmond>    or about other SNe,
<MichaelRichmond>    and I can try to answer questions on the observing end, or the theory end.
<MichaelRichmond>    I'm probably better on the observing end -- theory was never as interesting to me.
<MichaelRichmond>    So, rather than spend a lot of time typing now,
<MichaelRichmond>    I'll refer to you to a couple of places to get some general background info
<MichaelRichmond>    on SN 2011fe. The AAVSO has a web site you can visit:
<MichaelRichmond>    http://www.aavso.org/forums/supernovae-photometry
<MichaelRichmond>    and you might also find a little on the observational side at the pages I create
<MichaelRichmond>    to log my observations each night. Look at the bottom of this list:
<MichaelRichmond>    http://stupendous.rit.edu/richmond/ritobs/ritobs.html
<MichaelRichmond>    I'll keep updating that list with new pages for each night of observations.
<MichaelRichmond>    I haven't added the info for last night's data, in part because it's so crummy. Sigh.
<MichaelRichmond>    There are a lot of other places to find basic info about supernovae on the web.
<MichaelRichmond>    You can exercise your own Google-fu, and I encourage you to do so.
<MichaelRichmond>    If you want to, you might look at the web pages for supernovae
<MichaelRichmond>    in a couple of the courses I teach at RIT:
<MichaelRichmond>    http://spiff.rit.edu/classes/phys230/phys230.html
<MichaelRichmond>    is an introductory course with very little math required
<MichaelRichmond>    while the following has quite a bit more math in it:
<MichaelRichmond>    http://spiff.rit.edu/classes/phys301/phys301.html
<DocKQR>    OK, Michael. Want a question?
<MichaelRichmond>    And one more place to look is on a list of public talks I've given
<MichaelRichmond>    Wait wait -- just one more link
<MichaelRichmond>    http://spiff.rit.edu/richmond/asras/
<MichaelRichmond>    Okay, now I've finished ...
<DocKQR>    Are there any particular difficulties with SN photometry? How badly does the background galaxy interfere and what is done about that?
<MichaelRichmond>    Ah, the curse of the background galaxy.
<MichaelRichmond>    Yes, this can make measurements of supernovae more difficult than ordinary stars.
<MichaelRichmond>    There are three basic ways to deal with it.
<MichaelRichmond>    1) ignore it. Just do what you usually do and hope for the best.
<MichaelRichmond>    Often, this is done at early times, when the SN is bright, and it isn't so bad.
<MichaelRichmond>    At later times, when the background becomes more important, it leads to
<MichaelRichmond>    significant over- or under-estimates of the actual brightness.
<MichaelRichmond>    2) try a clever trick to reduce the background.
<MichaelRichmond>    Many galaxies are rather symmetric. If you take a picture of the galaxy and
<MichaelRichmond>    rotate it by 180 degrees, it will look very much like the original.
<MichaelRichmond>    So, take a picture with the SN in it, rotate by 180 degrees, and subtract from the original.
<MichaelRichmond>    Tadah! The background will often disappear -- mostly -- leaving just the SN behind.
<MichaelRichmond>    3) Wait for the SN to disappear -- a few months or a few years --
<MichaelRichmond>    and then take a set of "template" images.
<MichaelRichmond>    Subtract the template images from images with the SN,
<MichaelRichmond>    to leave just the light of the SN.
<MichaelRichmond>    That's the best way, but it usually means waiting for a long time.
<MichaelRichmond>    So, many impatient people use methods 1 or 2 at the start.
<DocKQR>    Could you use archive images to make the template? That way, if I read you correctly, you wouldn't have to wait for a fade.
<MichaelRichmond>    If you are lucky enough to have taken images of the galaxy
<MichaelRichmond>    with the same telescope + filter + camera combination,
<MichaelRichmond>    then, sure, you can use those pre-explosion images.
<DocKQR>    Ah! That makes sense!
<DocKQR>    OK. Thank you, Michael!
<DocKQR>    Next? (Doc ducks the impending question barrage!)
<MichaelRichmond>    But most people can't do this, and have to wait for later.
<MichaelRichmond>    So, next question?
<MichaelRichmond>    Oh, wait.
<MichaelRichmond>    For SN 2011fe, right now, the background light of M101 is pretty much negligible.
<MichaelRichmond>    So, you can just pretend the galaxy isn't there at all, and still get _very_ good results.
<MichaelRichmond>    Not perfect, perhaps, but for most purposes, good enough.
<MichaelRichmond>    Okay, now for next question.
<MichaelRichmond>    Anyone?
<TimCTX>    Michael, do you dare hazard a guess as to the length of the fade or is it to early to make a WAG?
<MichaelRichmond>    Well, if you look at this picture:
<MichaelRichmond>    http://www.rit.edu/cos/observatory/images/other/sn2011fe_comp94d.small.png
<MichaelRichmond>    you can see my measurements of SN 2011fe compared to measurements of
<MichaelRichmond>    another type Ia SN, 1994D. I just matched them roughly by hand.
<MichaelRichmond>    I'd say that this SN 2011fe is following the earlier one pretty faithfully,
<MichaelRichmond>    but there are some differences: this one looks bluer than SN 1994D, for example.
<MichaelRichmond>    You can see that _if_ this SN follows the usual evolution, it will go back to mag V = 12
<MichaelRichmond>    by around three weeks from now.
<MichaelRichmond>    -o-
<MichaelRichmond>    (that symbol is old-school Unix-speak for "over", as in "over with this one")
<EricBOS>    why is it it bluer? less extinction in the galaxy?
<MichaelRichmond>    Well, that's a good question.
<MichaelRichmond>    The effect of dust is largest in the blue, so we would expect to see the largest effect
<MichaelRichmond>    in the B-band.
<MichaelRichmond>    So, that's not a bad notion.
<MichaelRichmond>    However, the extinction to SN 1994D was pretty small, as I recall ....
<MichaelRichmond>    Hold on, I'll look it up ...
<TimCTX>    B-V in this case is approx ~-.1
<TimCTX>    SN 2011fe
<MichaelRichmond>    Ah, SN 1994D seemed to have less than 0.05 mag of reddening,
<EricBOS>    Is the extinction measured by spectroscopy?
<MichaelRichmond>    so I'd guess that the difference in colors is not due to dust.
<MichaelRichmond>    Perhaps SN 2011fe is just a little bluer intrinsically.
<MichaelRichmond>    The best ways to measure extinction are via spectroscopy, yes.
<MichaelRichmond>    With a high-resolution spectrum, you can measure the amount of interstellar gas
<MichaelRichmond>    between you and a SN. Then, if you can guess the ratio of gas to dust (that's one guess)
<MichaelRichmond>    and guess the effect of the dust on light (that's another guess)
<MichaelRichmond>    you can compute the amount of extinction in different optical filters.
<MichaelRichmond>    I haven't seen high-res spectra of SN 2011fe myself,
<MichaelRichmond>    though I know lots of people must be acquiring them now.
<MichaelRichmond>    Have any of you seen such spectra?
<MatthewTMT>    I confess ignorance: what's the bump that's prominent in I at t+30?
<MichaelRichmond>    Ah, the old secondary maximum.
<EricBOS>    How does one guess the ratio gas/dust?
<MichaelRichmond>    One at a time. Let's stick to gas/dust now, and do the bump later.
<ArneHQA>    http://astrosurf.com/aras/surveys/supernovae/sn2011fe/obs.html
<MichaelRichmond>    So, the ratio of gas to dust.
<MatthewTMT>    (sorry, got out of sync)
<MichaelRichmond>    In the Milky Way, you can measure the amount of gas by look at a cloud with spectroscopy.
<MichaelRichmond>    You can measure the amount of dust by looking at distant galaxies on the other side of the cloud
<MichaelRichmond>    and seeing how much they have been reddened and/or dimmed.
<MichaelRichmond>    So, in the Milky Way, you can get some handle on the ratio of gas to dust
<MichaelRichmond>    in interstellar clouds.
<MichaelRichmond>    It's not a constant ratio -- it varies from here to there, from cloud to cloud.
<MichaelRichmond>    In other galaxies, we have (basically) no good way to measure this ratio.
<MichaelRichmond>    So, we often guess either a) this is just like the Milky Way,
<MichaelRichmond>    or b) well, this galaxy seems to have fewer metals in its stars, so we adjust accordingly.
<MichaelRichmond>    The effect of gas and dust on observations of SNe in other galaxies is a current question
<MichaelRichmond>    about which many people are taking a big interest.
<MichaelRichmond>    If you want to do cosmology with SNe, you need to know the answer to that question.
<MichaelRichmond>    Right now, the effects of gas and dust cause one of the largest uncertainties
<MichaelRichmond>    in our use of SNe for cosmology.
<MichaelRichmond>    -o-
<ArneHQA>    Michael, the url I gave has the spectra
<MichaelRichmond>    Oh, wait, Arne provided a spectrum for us to look at
<MichaelRichmond>    Let's all go look at the spectrum!
<MichaelRichmond>    Okay, I'm looking at it now myself.
<MichaelRichmond>    This is an example of a "low-resolution" spectrum, at least in the form that Arne provided.
<MichaelRichmond>    The entire visible spectrum is squished into a few thousand pixels.
<MichaelRichmond>    In order to study the gas content of clouds between us and the SN,
<MichaelRichmond>    both clouds in our own MW and clouds in the SN's host galaxy,
<MichaelRichmond>    we would need to zoom in on little areas of this spectrum.
<MichaelRichmond>    High-resolution spectroscopy is more difficult to do,
<MichaelRichmond>    because it breaks up the light into smaller pieces
<MichaelRichmond>    (don't quote me on that, I'll get in trouble with my students)
<MichaelRichmond>    and sends fewer photons to each little pixel of one's detector.
<MichaelRichmond>    One needs a BIG telescope to do the high-res spectroscopy properly. Alas.
<MichaelRichmond>    So, I can't use this particular spectrum to say anything about gas and dust.
<MichaelRichmond>    Well, even if it was a good high-res spectrum,
<MichaelRichmond>    I couldn't make the proper measurements while typing all this in a live chat.
<MichaelRichmond>    But we can look at the spectrum for other reasons.
<MichaelRichmond>    Note all the bumps!
<MichaelRichmond>    The spectra of ordinary stars are like a single big hill, with little tiny wiggles.
<MichaelRichmond>    The spectra of SNe are crazy roller coasters of features.
<MichaelRichmond>    One of the reasons for the big difference is the SPEED of the material
<MichaelRichmond>    flying out of a SN.
<MichaelRichmond>    Because the material is blown out at speeds around 10,000 km/s,
<MichaelRichmond>    and at very high temperatures,
<MichaelRichmond>    there are very high motions of the gas within the ejecta.
<MichaelRichmond>    Some is coming straight towards us at 10,000 km/sec,
<MichaelRichmond>    some moving away at 10,000 km/sec,
<MichaelRichmond>    some moving sideways to us and so not receding at all.
<MichaelRichmond>    Gas which moves relative to us will have its spectral emission and absorption lines shifted.
<MichaelRichmond>    In SNe, some of these shifts can be 10 or 20 or 50 Angstroms in size.
<MichaelRichmond>    If one layer of gas is absorbing light at, say, 6200 Angstroms,
<MichaelRichmond>    the gas coming TOWARD us will be blueshifted: line appears at 6100 Angstroms.
<MichaelRichmond>    The gas moving AWAY from us is redshifted: line appears at 6300 Angstroms.
<MichaelRichmond>    The gas moving SIDEWAYS isn't shifted at all: line appears at 6200 Angstroms.
<MichaelRichmond>    Mix all that light together, and you see a single broad line from 6100 all the way to 6300 Angstroms.
<MichaelRichmond>    In ordinary stars, on the other hand, motions are, what, 1 or 2 or 5 km/s at most
<MichaelRichmond>    and usually much smaller.
<MichaelRichmond>    So absorption and emission lines in ordinary stars are very narrow.
<MichaelRichmond>    If you see a spectrum with wide, big lines, it is likely to be a SN.
<MichaelRichmond>    Someone has very politely labelled some of the strongest lines in this spectrum.
<MichaelRichmond>    That's good, because I might be able to point out just 2 or 3 from memory.
<MichaelRichmond>    Does anyone see lines of hydrogen?
<MichaelRichmond>    (waits)
<MichaelRichmond>    (waits)
<MichaelRichmond>    (waits)
<DocKQR>    Errr. No.
<MichaelRichmond>    Right!
<MichaelRichmond>    This was a "type Ia" supernova.
<MichaelRichmond>    That label -- "type Ia" -- simply means that the spectrum shows no signs of hydrogen.
<MichaelRichmond>    The reason that there is no hydrogen in this object is that the star lost it all.
<MichaelRichmond>    Ordinary stars are mostly hydrogen, like our Sun.
<DocKQR>    OK, I'm really going to be dumb here, Michael. That really surprises me seeing that H is the largest part of a star.
<MichaelRichmond>    Near the end of their lives, however,
<MichaelRichmond>    stars like the sun will gently blow their outer layers off.
<MichaelRichmond>    If you have ever seen a planetary nebula (M57, for example),
<MichaelRichmond>    you know how pretty those layers of gas look when they are lit up by the central core
<MichaelRichmond>    of the star.
<MichaelRichmond>    Now, that central core has very little hydrogen left.
<MichaelRichmond>    It's mostly helium, or, for stars which are bit more massive at birth than the Sun,
<MichaelRichmond>    oxygen and carbon.
<EricBOS>    If the lines were not indicated on the SN spectra, could you tell whether the features are absorption lines or emission lines?
<MichaelRichmond>    This leftover core is very dense and very hot,
<MichaelRichmond>    and glows a whitish color: we call it a "white dwarf"
<MichaelRichmond>    (getting to it, Eric)    
<MichaelRichmond>    There's a general result, which I can't explain here,
<MichaelRichmond>    about the physics of these compact white dwarfs:
<MichaelRichmond>    if you can manage to pile up enough mass on their outer layers,
<MichaelRichmond>    you can cause a thermonuclear explosion to occur!
<MichaelRichmond>    A "Type Ia" supernova is the result of such an explosion.
<MichaelRichmond>    Because the star which explodes has lost all its hydrogen,
<MichaelRichmond>    we see only features due to heavier elements in its spectrum.
<MichaelRichmond>    You can see features due to silicon,
<MichaelRichmond>    but also, important, the features due to iron (Fe) and cobalt (Co).
<MichaelRichmond>    I say these iron-group elements are important because the bulk of the white dwarf
<MichaelRichmond>    turns into iron-group elements,
<MichaelRichmond>    and because the radioactive decay of certain isotopes of these elements
<MichaelRichmond>    (especially nickel-56)
<MichaelRichmond>    provide most of the power which causes the SN to shine so bright.
<MichaelRichmond>    If it weren't for the radioactive decay of those elements,
<MichaelRichmond> we would not see the supernovae reach such a bright peak,
<MichaelRichmond> nor last for several weeks or months before fading away.
<MichaelRichmond> So, if you want to understand Type Ia SNe, you need to understand all about
<MichaelRichmond> the fusion of elements into iron-group elements.
<MichaelRichmond> Phew.
<MichaelRichmond> Okay, now for Eric's question:
<MichaelRichmond> could you tell whether the features are absorption lines or emission lines
<MichaelRichmond> The answer is "I don't know".
<MichaelRichmond> In ordinary stars, it's easy to find the "continuum".
<MichaelRichmond> That's just the main top-surface of the spectrum,
<MichaelRichmond> the part down from which absorption lines appear.
<MichaelRichmond> In the case of supernovae, however,
<MichaelRichmond> the star doesn't have a well-defined "surface" the way our Sun does.
<MichaelRichmond> The material in a SN is flying outward at high speeds,
<MichaelRichmond> and possibly in clumps,
<MichaelRichmond> and some of it is opaque at certain wavelengths,
<MichaelRichmond> but transparent at others.
<MichaelRichmond> So, at some wavelengths, you are looking deep into the guts of the exploding star,
<MichaelRichmond> but at other wavelengths, you are seeing only the skin and hair.
<MichaelRichmond> So to speak.
<MichaelRichmond> That means that a SN does not have a single, well-defined temperature,
<MichaelRichmond> unlike ordinary stars.
<MichaelRichmond> (yes, yes, experts will say that even the sun doesn't have a simple, well-defined temperature)
<MichaelRichmond> (humor me here)
<MichaelRichmond> So, in the spectrum of a SN, we can't easily draw that line which represents the
<MichaelRichmond> blackbody-like emission of light from a gas at a single temperature.
<MichaelRichmond> That means that deciding whether some particular feature goes ABOVE the line
<MichaelRichmond> (and so is emission)
<MichaelRichmond> or BELOW the line (and so is absorption)
<MichaelRichmond> is a difficult thing to do.
<MichaelRichmond> I think that most of the main features of a SN at this stage in its development are absorption
<MichaelRichmond> lines, and you can see that the marked features are mostly (entirely?) dips.
<MichaelRichmond> At later times, when the gas thins out, the spectrum is dominated by emission lines.
<MichaelRichmond> -o-
<DocKQR> How are your fingers doing, Michael?
<MichaelRichmond> My fingers are okay now, but I'll need to take a brief break at some point.
<ArneHQA> the lack of a continuum is also why (B-V) is hard to measure, and "blue" is relative
<MichaelRichmond> Perhaps a few more questions.
<MichaelRichmond> Go on, Arne, please.
<DocKQR> Perhaps if someone has a general question we can call on our Director or Sciences Director to pinch-hit for a few minutes.
<EricBOS> Is there a relation between intrinsic color and absolute magnitude of SNIa's?
<MichaelRichmond> Oh, as soon as I quaff a soda, I'll be fine.
<MichaelRichmond> Any other questions?
<DocKQR> Well, right now we have Eric's on the table.
<MichaelRichmond> Eric's question -- can you remind me?
<EricBOS> Is there a relation between intrinsic color and absolute magnitude of SNIa's?
<MichaelRichmond> Yes, there is.
<MichaelRichmond> A long time ago (well, 20 or 40 years ago),
<MichaelRichmond> astronomers had only a small dataset of measurements of SNe.
<MichaelRichmond> At that time, it appeared that Type Ia SNe were very very similar to each other.
<MichaelRichmond> So, for a while, we assumed that all Type Ia SNe were exactly the same:
<MichaelRichmond> same peak luminosity, same absolute magnitude, same color, etc.
<MichaelRichmond> However, over the past 20 or so years,
<MichaelRichmond> with the much larger number of supernovae we've now observed,
<MichaelRichmond> it has become clear that not all Type Ia SNe are identical.
<MichaelRichmond> Sure, they are similar to each other,
<MichaelRichmond> but some may be 2x or 4x as bright as others at peak.
<MichaelRichmond> Hmmmm.
<MichaelRichmond> That's not a good sign for those who wish to use them as "standard candles."
<MichaelRichmond> Fortunately, it turns out that in MOST cases,
<MichaelRichmond> there are correlations:
<MichaelRichmond> Type Ia SNe which are brighter than average are also
<MichaelRichmond>   a) bluer than average at early times
<MichaelRichmond>   b) slower to decline after maximum
<MichaelRichmond> whereas the Type Ia SNe which are fainter than average are
<MichaelRichmond>   a) redder than average at early times
<MichaelRichmond>   b) quicker to decline.
<MichaelRichmond> The connections were first spelled out in the refereed literature by Mark Phillips in
<MichaelRichmond> 1993 or 1994.
<MichaelRichmond> There have been improvements made in the scheme,
<MichaelRichmond> with lots of new ways to measure and characterize SNe,
<MichaelRichmond> but the basic idea remains.
<MichaelRichmond> The bottom line is:
<MichaelRichmond> a subset of Type Ia SNe really are nearly identical,
<MichaelRichmond> another group of Type Ia SNe are different, but can be recognized and "fixed",
<MichaelRichmond> and a third group are just trouble-makers.
<MichaelRichmond> So, it IS possible to do reasonably good cosmology with Type Ia SNe,
<MichaelRichmond> as long as you have lots of events, and lots of extra info on the events,
<MichaelRichmond> like colors and light curves and spectra,
<MichaelRichmond> and as long as you are careful to use only the good ones,
<MichaelRichmond> and "fix" the ones which can be fixed,
<MichaelRichmond> and ignore the ones which can't be fixed.
<MichaelRichmond> Is that answering your question?
<EricBOS> Yes, thanks!!!
<MichaelRichmond> As Bugs Bunny said in "Rabbit of Sevillie",  ..... "Next?"
<MatthewTMT> Did you cover the bump in I?
<MichaelRichmond> Ah, the bump in I.
<MichaelRichmond> Not yet.
<MichaelRichmond> So, why is there a bump in the I-band?
<MichaelRichmond> This one has no easy answer,
<MichaelRichmond> in part because it's a complicated subject,
<MichaelRichmond> and in part because I haven't had to answer it,
<MichaelRichmond> so I don't know the answer cold.
<MichaelRichmond> It has something to do with several factors.
<MichaelRichmond> First, the initial explosion creates a lot of Ni-56 -- or is it cobalt-56?  One or the other.
<MichaelRichmond> Radioactive decay of item one (I _think_ it's the Ni-56) into item two (I _think_ it's cobalt 56)
<MichaelRichmond> produces a lot of energy to power the explosion and light show.
<MichaelRichmond> Then, the item two finally decays into the stable iron-56.
<MichaelRichmond> That provides even more energy over a longer timescale.
<MichaelRichmond> All this transmutation occurs while the ejecta is flying through space.
<MichaelRichmond> So, if you're trying to make a model of how light which is created INSIDE the ejecta
<MichaelRichmond> reaches the OUTSIDE of the ejecta,
<MichaelRichmond> you have to deal with complications: the chemical composition is changing
<MichaelRichmond> as the ejecta expands.
<MichaelRichmond> In addition, as the ejecta expands, its density is decreasing,
<MichaelRichmond> and its temperature is decreasing, too.
<MichaelRichmond> Now, when light is trying to fight its way through a clump of gas,
<MichaelRichmond> what matters?
<MichaelRichmond> Temperature, density and chemical composition.
<MichaelRichmond> Ugh.  This is a complex problem.
<MichaelRichmond> It turns out that at one point, a few weeks after the explosion,
<MichaelRichmond> a large bulk of the material has turned into iron-56.
<MichaelRichmond> At first, the temperatures are so high that the iron is ionized:
<MichaelRichmond> its electrons are stripped off.
<MichaelRichmond> Well, not all 26 electrons -- iron is element 26, right? --
<MichaelRichmond> but some of them.
<MichaelRichmond> In particular, at one point, a large amount of the gas is iron with 2 electrons missing.
<MichaelRichmond> We call that "Fe III".
<MichaelRichmond> Astronomers count the number of missing electrons and add one.  Don't ask why.
<MichaelRichmond> Now, at a certain point in the evolution of the gas, it cools down
<MichaelRichmond> so that one of the two missing electrons can be captured by iron atoms.
<MichaelRichmond> When the atoms capture this electron, their opacity changes.
<MichaelRichmond> Some of the light which used to be blocked by the iron atoms can now escape.
<MichaelRichmond> Reddish light has an easier time escaping from the new, just-one-electron-mission iron gas,
<MichaelRichmond> and so at this point, the supernova emits a large amount of reddish light.
<MichaelRichmond> If you are looking at the supernova through a reddish filter --- like the I-band filter ---
<MichaelRichmond> you'll see the supernova grow brighter for a short time as this energy escapes.
<SeijiTSJ> interesting
<MichaelRichmond> If you're looking a the supernova with a blue filter, you won't see this same glow.
<MichaelRichmond> I am oversimplifying a great deal,
<MichaelRichmond> so much that I almost think I understand myself.
<MichaelRichmond> That's a sign that I'm not doing a good job.
<MichaelRichmond> But perhaps this may give you a vague idea of why a bump may appear
<MichaelRichmond> in some filter, but not in another.
<MichaelRichmond> If you want to see some of the details, check out this presentation:
<MichaelRichmond> http://online.itp.ucsb.edu/online/snovae_c07/kasen/pdf/Kasen_SnovaeConf_...
<MichaelRichmond> -o-
<DocKQR> Whew!
<MichaelRichmond> That's probably all I can say about that question.
<MatthewTMT> Fair enough, thanks.
<MichaelRichmond> Oh, wait
<MichaelRichmond> I
<MichaelRichmond> This "secondary bump" in the I-band
<MichaelRichmond> is a clear sign of a type Ia supernova.
<MichaelRichmond> Other supernovae, the core-collapse type, do not show a secondary maximum.
<MichaelRichmond> If you are studying distant supernovae, so faint that you can't acquire a spectrum,
<MichaelRichmond> this distinctive shape of the light curve can be a very handy tool
<MichaelRichmond> to distinguish between Type Ia and other SNe.
<MichaelRichmond> Okay, now I've used up all my knowledge happy
<DocKQR> Matt, can you mention a bit about what data the AAVSO has received with regard to SN 2011fe?
<MichaelRichmond> Well, if one goes to the AAVSO "WebObs" for this object,
<MichaelRichmond> one will find quite a lot of data.
<MichaelRichmond> It is all "unvalidated" at the moment,
<MatthewTMT> Yes, as I said earlier, we have about 1400+ observations so far.
<MichaelRichmond> so before I discuss it, I'd like one of our the AAVSO experts to explain what that means.
<MichaelRichmond> Arne, Doc?
<MatthewTMT> We have a good mix between visual and CCD.
<MatthewTMT> Unvalidated simply means that a staff member hasn't gone over the data to do a quality check.
<MichaelRichmond> Okay, could you wait for just 30 seconds ...
<DocKQR> Absolutely!
<DocKQR> We do quality checks in a number of ways. With the amount of data now coming in, there are times when we have to be very specific about quality checks with stars getting concentrated quality checks when we know the data will be used for a project.
<DocKQR> Ongoing quality checks are actually done with the help of the observers and members themselves via our ZAP program, which you can look up on the web site.
<MichaelRichmond> Okay, everyone, please take a look at this picture:
<MichaelRichmond> http://stupendous.rit.edu/richmond/temp/foo.gif
<MichaelRichmond> It shows the current AAVSO light curve for SN 2011fe
<MichaelRichmond> CCD V-band observations are shown as green dots, and visual measurements as black symbols.
<MichaelRichmond> What can we learn from this particular graph?
<MichaelRichmond> Well, I'd say we can learn several things:
<MichaelRichmond> first, the supernova has reached its maximum light, and maybe just has started to decline.
<EricBOS> we could do without the CCD obs? wink
<MichaelRichmond> Another thing we can see is that all the CCD data doesn't agree with itself.
<MichaelRichmond> That's one sign that there are small, probably systematic errors in the CCD measurements.
<MichaelRichmond> It's difficult to combine measurements from several observers, if you want them to agree
<DocKQR> Could that be caused by the background light?
<SeijiTSJ> Michael, do you think the second peak will arrive in future?
<MichaelRichmond> at the level of 1 or 2 or 3 or 5 percent.
<MichaelRichmond> Hold on, one at a time!
<MichaelRichmond> The differences are probably NOT due to the background light,
<MichaelRichmond> and the different ways that observers are extracting magnitudes from the images.
<MichaelRichmond> Instead, I think that most of the differences between observers are caused by the
<MichaelRichmond> slightly different filters and cameras that different people have.
<MichaelRichmond> I can purchase a V-band filter, and you can purchase a V-band filter,
<MichaelRichmond> but unless we purchase the filter from the same company, and the same manufacturing run,
<MichaelRichmond> our filters are likely to have small differences.
<MichaelRichmond> For ordinary stars, this can cause small differences in our measurements.
<MichaelRichmond> With the proper technique, we can often remove these differences.
<EricBOS> How does these differences affect the science? Are these important with the "amplitude" of the SN?
<MichaelRichmond> Want to know how?  Read up on "first-order photometric color terms".
<MichaelRichmond> You might even find some documents at the AAVSO web site to help you.
<MichaelRichmond> For example, you might look at this:
<MichaelRichmond> http://spiff.rit.edu/richmond/sne/sn2011fe/color_terms.html
<MichaelRichmond> Now, for supernovae, the problem is worse than for ordinary stars.
<MichaelRichmond> The reason is that that supernovae have such wierd spectra,
<MichaelRichmond> with big ups and downs.
<MichaelRichmond> If a strong emission line lies just INSIDE my V-band filter,
<MichaelRichmond> but just OUTSIDE your V-band filter,
<MichaelRichmond> then there's no way that you and I can measure the same amount of light.
<MichaelRichmond> That's one of the big problems with doing photometry of SNe:
<MichaelRichmond> when you try to combine measurements from different observers,
<MichaelRichmond> you often need to make small corrections, and it's not easy to figure out what they are.
<MichaelRichmond> It's much easier if you just use measurements made by one person,
<MichaelRichmond> but then, of course, one person loses time to clouds and sleep and so forth.
<MichaelRichmond> So, studying supernovae is a no-win situation happy
<MichaelRichmond> What I would like to do is to communicate with people who are contributing measurements
<DocKQR> ...or to rotten Rochester weather (grew up in CNY. I know!)
<MichaelRichmond> of SN 2011fe to the AAVSO
<ArneHQA> the transformation issue also affects novae with the halpha line in the wings of some V filters
<MichaelRichmond> and talk to them -- find out how they are making their measurements.
<MatthewTMT> AAVSO HQ is developing materials to help CCD observers properly.  For now, you can see whether data are on a standard system by looking at the transformed flag in data downloads.
<MichaelRichmond> If I can find 2 or 3 or 5 people who will work with me to
<MichaelRichmond> go over the details of their observations,
<MichaelRichmond> and who will make the EXTRA observations required to determine the color terms
<MichaelRichmond> for their telescopes and cameras and filters,
<MichaelRichmond> then I hope to combine the measurements for those 2 or 3 or 5 people
<MichaelRichmond> into a single, continuous light curve, with no gaps.
<MichaelRichmond> If we can do that for the V-band,
<MichaelRichmond> and for the B-band, and the R-band, and the I-band,
<MichaelRichmond> then we can write a paper and publish it.
<EricBOS> Unfortunately I couldn't observe the SN yet, but I have tranformation coeffs
<MichaelRichmond> Then lots of other astronomers, who have spectra or radio data or whatever,
<MichaelRichmond> can use our measurements in their work.
<MichaelRichmond> Right.  people who are willing and able to determine color coefficients
<MichaelRichmond> for their telescopes are the people I hope to contact over the next few weeks.
<ArneHQA> (note: all AAVSOnet telescopes have known transformation coefficients)
<MichaelRichmond> Okay, that's one thing.
<MichaelRichmond> Let's look at the light curve again.
<MichaelRichmond> What other differences are there between green points and black points?
<MichaelRichmond> Anyone?
<MichaelRichmond> Just yell it out.
<EricBOS> scatter
<MichaelRichmond> Right.  Visual obs have more scattre.
<MichaelRichmond> I also see a systematic difference.
<MichaelRichmond> The visual observations tend to be a bit brighter than the CCD observations at early times.
<MichaelRichmond> That difference may change at later times -- there's a hint of it here ...
<MichaelRichmond> and we may see that at later times, the visual estimates may be a bit fainter than the SN.
<MichaelRichmond> The human eye's response to light is not the same as that of a V-band CCD system.
<MichaelRichmond> For ordinary stars, there might be a small difference between visual and CCD observations.
<MichaelRichmond> I am sure that some of the AAVSO experts will point out examples, if we ask.
<MichaelRichmond> But supernovae are not ordinary stars,
<MichaelRichmond> and one reason is that they change both BRIGHTNESS and COLOR over a few weeks.
<MichaelRichmond> That means that the difference (visual - V-band) can change over several weeks,
<MichaelRichmond> as the supernova's color and brightness change.
<MichaelRichmond> There was some hint of this in the past
<MichaelRichmond> with some supernova .... I think it was SN 1991T ...
<MichaelRichmond> which showed a systematic shift in the (visual - V-band) measurements as a function of
<MichaelRichmond> the color (or temperature) of the SN.
<MichaelRichmond> I would like to use the AAVSO's dataset to look for a similar relationship
<MichaelRichmond> in SN 2011fe.  It should be easier to do,
<MichaelRichmond> because this SN should be a bit brighter than 1991T, I think.
<MichaelRichmond> So, that's another thing to keep in mind over the next few weeks/months.
<MichaelRichmond> -o-
<MichaelRichmond> Next?
<EricBOS> The LC looks rather good tho? How does these differences affect the science? Are these important wrt the "amplitude" of the SN?
<DocKQR> OK, with the data density that Michael is answering these questions
<DocKQR> I think we have time for a couple more.
<MichaelRichmond> Eric asked a question
<TimCTX> Michael, Thanks for sharing so much information with us... it is appreciated
<MichaelRichmond> but I'm going to count to 100 before answering.
<MichaelRichmond> so that someone else may ask one.
<MichaelRichmond> Eric asks a lot of questions happy
<DocKQR> That and you don't get carpal tunnel!
<EricBOS> yep, sorry!
<MichaelRichmond> 20
<MichaelRichmond> 40
<MichaelRichmond> 60
<DocKQR> Actually I think some of us are trying to digest what you've already said, Michael. I know I am!
<MichaelRichmond> 80
<MichaelRichmond> 100
<MichaelRichmond> Okay, here we go.
<MichaelRichmond> Eric asked: How does these differences affect the science? Are these important wrt the "amplitude" of the SN?
<MichaelRichmond> How do the difference affect the science?
<MichaelRichmond> Well, in the most basic term, we care about the peak magnitude of the SN.
<SeijiTSJ> Michael, have you taken some spectra?
<MichaelRichmond> Cosmologists use the peak magnitude to measure distances to very distant SNe.
<MichaelRichmond> Suppose that there were, oh, just making up a number, 0.5 magnitude difference
<MichaelRichmond> between visual estimates and CCD measurements.
<MichaelRichmond> (Seiji -- no, I have no spectrograph)
<MichaelRichmond> If we used eyeballs to measure nearby supernovae,
<MichaelRichmond> and CCDs to measure distant supernovae,
<MichaelRichmond> and then compared the nearby to the distant supernovae,
<MichaelRichmond> we would be making a mistake:
<MichaelRichmond> the nearby ones would all be a little fainter than we thought,
<MichaelRichmond> due to this small, systematic difference between visual and CCD measurements.
<MichaelRichmond> Now, if we didn't know that,
<MichaelRichmond> we'd insert a false extra 0.5 magnitude into our cosmological calculations.
<MichaelRichmond> That might cause the fate of the universe to change from open-forever
<MichaelRichmond> to open-forever-but-not-quite-as-open-as-we-thought.
<MichaelRichmond> Or, in other terms, it might change our notion of the cosmological constant
<MichaelRichmond> from lambda = 0.73 to lambda = 0.65 or something.
<MichaelRichmond> (I am again shamelessly making up numbers)
<EricBOS> Are the uncertainties due to extinction and intrinsic differences not larger than the scatter in the (CCD) observations?
<MichaelRichmond> The important thing is that, if we use different tools to measure nearby and distant SNe,
<MichaelRichmond> we can end up with a SYSTEMATIC error,
<MichaelRichmond> and that systematic error can cause us big problems when we do the cosmology.
<MichaelRichmond> You might not think that this is a realistic example --
<MichaelRichmond> no astronomers would be silly enough to compare visual and CCD measurements, right? --
<MichaelRichmond> but it's not really that silly.
<MichaelRichmond> Astronomers use one set of instruments -- small telescopes on the ground --
<MichaelRichmond> to study nearby SNe.
<MichaelRichmond> Astronomers use a different set of instruments --- big telescopes on the ground,
<MichaelRichmond> often with different filters, and telescopes in space ---
<MichaelRichmond> to measure distant supernova.e
<MichaelRichmond> So it isn't so far-fetched to imagine that there could be systematic
<MichaelRichmond> errors in the comparison of these two sets of measurements.
<MichaelRichmond> If you ever try to compare measurements of a mag V=12 star to a mag V=20 star,
<MichaelRichmond> you'll soon find how hard it is to do.
<MichaelRichmond> Now, Eric asked another question:
<MichaelRichmond> Are the uncertainties due to extinction and intrinsic differences not larger than the scatter in the (CCD) observations?
<MichaelRichmond> How to answer this one?
<MichaelRichmond> "Scatter" in the CCD observations is perhaps a few percent,
<MichaelRichmond> if you look at the scatter in measurements of the same object made by a few different groups.
<MichaelRichmond> Uncertainties due to extinction ... well, if we knew how big the extinction was,
<MichaelRichmond> we could remove it properly.
<MichaelRichmond> The fact that we don't know how big it is means it's hard to estimate its size.
<MichaelRichmond> The really big problem with extinction can occur if dust in the local univers
<MichaelRichmond> does not have the same properties as dust in the distant universe.
<MichaelRichmond> Here in the Milky Way, stars have been creating dust for 10 billion years.
<MichaelRichmond> That means that the dust is a mixture of, well, lots of different elements, from a mix of many stars.
<MichaelRichmond> But the dust in a galaxy at z=1 may be different:
<MichaelRichmond> that galaxy has only existed for, um, a few billion years,
<MichaelRichmond> (sidenote: use this Cosmology Calculator to find out the real age
<MichaelRichmond> http://www.astro.ucla.edu/~wright/CosmoCalc.html
<MichaelRichmond> hmmmm.... turns out to be about 5 billion years)
<MichaelRichmond> so that dust is probably quite different from the dust in our galaxy.
<MichaelRichmond> That means that even if we know how much dust there is in that other galaxy,
<MichaelRichmond> we might calculate the way that it affects starlight in the wrong way.
<MichaelRichmond> That, in turn, means that all of our measurements of SNe in the "nearby" universe
<MichaelRichmond> could have  a small but systematic difference, compared to SNe in the distant universe.
<MichaelRichmond> It is these systematic errors that drive astronomers crazy.
<MichaelRichmond> Sorry, that's not right.
<MichaelRichmond> The systematic errors drive SOME astronomers crazy.
<MichaelRichmond> The astronomers who just want to publish lots of papers in ApJ may ignore this whole idea
<MichaelRichmond> because its takes time to figure out how to deal with the issue.
<MichaelRichmond> Whoops.   Must be careful to avoid snarkiness.
<MichaelRichmond> Ahem.
<MichaelRichmond> Perhaps another question would be a good idea.
<MichaelRichmond> -o-
<DocKQR> There are times when I like snarkiness. Those are the times that get me in trouble, I'm sure.
<MichaelRichmond> Further questions .... ?
<DocKQR> Folks I think we can do one more. I'm afraid of getting Doctor bills from Michael's doctor. Anyone else?
<EricBOS> sorry for asking all these questions
<DocKQR> Don't be sorry, Eric. They were great questions!
<MichaelRichmond> Andy? MORE? rebecca?
<EricBOS> I actually had only one question prior to this session, and I didn't ask it yet...
<DocKQR> Dun Dun DUHHHHH!  happy
<MichaelRichmond> Neil?  Tim?  Tom?  
<MichaelRichmond> Member06?
<DocKQR> You guys are still here, right?
<MichaelRichmond> Well, Eric, you may as well go for it.
<DocKQR> Go for it Eric!
<EricBOS> thanks! Here it comes:
<EricBOS> How does one apply the "template" method practically? Is there a tool in IRAF to convolve images with a Gaussian as described in your 1994D paper? Can you refer to a resource explaining how to do this?
<TimCTX> I am saturated... pun intended
<MichaelRichmond> Whoa, that's a tough question.
<MichaelRichmond> The proper answer to that question is: arrange to take a class with me.
<MichaelRichmond> I _could_ explain to you how I would do this sort of procedure.
<MichaelRichmond> But it would take a long time, and probably I would refer to bits and
<EricBOS> Would like that. Guess I have to pass the exams of this class first happy
<MichaelRichmond> pieces of software that you might not have at hand.
<MichaelRichmond> I don't use IRAF much myself, so it's possible that I might not know about
<MichaelRichmond> some tool in IRAF that does much of the work.
<MichaelRichmond> What is true is that "image subtraction" is used pretty widely these days,
<SeijiTSJ> Michael, what line emission or absorption do you think you can see now?
<MichaelRichmond> not only in the study of supernovae.
<EricBOS> Didn't you apply this method yet?
<MichaelRichmond> So, it's possible that you might be able to find some local expert
<MichaelRichmond> who can explain some of the pieces to you.
<EricBOS> Oops, sorry Seiji
<MichaelRichmond> If you want to ask me about this later, please send me E-mail,
<MichaelRichmond> and we can exchange E-mails on the topic.
<MichaelRichmond> There's no way I can address it adequately during this chat, I'm afraid.
<MichaelRichmond> And no, I haven't been applying that technique to images of SN 2011fe
<MichaelRichmond> mostly because, for my own setup, the background light of the galaxy really
*** AAVSOMember10 [5e04224d@E9CE86C3.DE259A57.97586D97.IP] has joined #aavso
<MichaelRichmond> is negligible compared to the light of the SN, at this time.
<MichaelRichmond> Seiji asked:
<EricBOS> OK, thanks for this excellent course Michael!
<MichaelRichmond> Michael, what line emission or absorption do you think you can see now
<MichaelRichmond> Seiji, could you look at the spectrum Arne mentioned:
<MichaelRichmond> http://astrosurf.com/aras/surveys/supernovae/sn2011fe/obs.html
<MichaelRichmond> You'll see a number of lines are marked.
<MichaelRichmond> I didn't create this graph or take the spectrum -- ask Arne or read the web page for details.
<MichaelRichmond> But that spectrum does give you an idea for the strongest lines in the spectrum now.
<SeijiTSJ> thanks
*** AAVSOMember10 [5e04224d@E9CE86C3.DE259A57.97586D97.IP] has quit [Quit: CGI:IRC (EOF)]
<MichaelRichmond> Most of due to intermediate-mass elements (oxygen, calcium)
<MichaelRichmond> or iron-group elements (iron, cobalt).
<MichaelRichmond> Sumimasen.
*** <EricBOS>_ [user@9C20A0AC.F2448070.B2B3AC89.IP] has joined #aavso
<MichaelRichmond> I guess I might have to go home soon.
<DocKQR> OK.
<DocKQR> Everyone breath.
<DocKQR> We can all give a virtual round of applause
<DocKQR> to our guest Micheale Richmond who
<DocKQR> may have given the most information packed chat we've ever had!!
<DocKQR> Micheal asked me if I had been in advertising when I wrote the blurb for this chat. Now I'm going to ask him
<DocKQR> if he has a background in stenography with the amount of his typing!
<EricBOS> Thank you very much again Michael!
<MichaelRichmond> Nope, no stenography, just lots and lots of coding and writing lectures.
<DocKQR> So, THANK YOU, Michael. This has been a chat to remember!
<MichaelRichmond> I wish I'd taken a typing course, though.
<DocKQR> I don't think you need the typing course!!
<MichaelRichmond> You're welcome.
<DocKQR> OK, folks...
<MichaelRichmond> I hope everyone will look for the SN!
<DocKQR> With Rebecca and Matt's help I'll be assembling a log of the chat for posterity, and it will be up on the website sometime tomorrow morning.
<MichaelRichmond> -o-o-
<DocKQR> Thank you all for coming, and again, many MANY thanks to Michael for coming to the chat and sharing his experience with us!!
 

AAVSO 49 Bay State Rd. Cambridge, MA 02138 aavso@aavso.org 617-354-0484