PDA

View Full Version : Another year, another new type of supernova?



Jerry
2013-Feb-05, 06:25 AM
A quick search of the board shows that we introduce a quirky supernova-like event about once a year. From yesterday's Fun papers in Archives:

New Type SN?* http://arxiv.org/abs/1302.0009 This event, at a redshift of ~1.39 looks like more local, superluminous events; except that the light curve is unacceptably short for a superluminous, hydrogen poor event. Extremely bright, it resembles a more local type IC; which generally have very long light curves, but there is at least one example of a relatively local superluminous event with an average light curve.

So what to make of it?

First of all, distance event searchs would not normally identify such a rapid event - it has always been assumed that in the time-dilated redshifted world, nothing would decompose this quickly. Now that one has been found; search alogrithyms will be amended to help shed light on such activity.

Second, if there is anyone other than Jerry who is not sold on whether-or-not very distant events are in fact time-dilated; this is the kind of evidence they should be looking for.

If this one-of-a kind is not really an unusual event; but represents the norm at great distance; maybe it really is turtles all the way down.

Shaula
2013-Feb-05, 07:19 AM
So the options are:
1) All the other events are anomalous, this one event that supports your beliefs is actually the norm (just haven't seen any before now)
2) This is an anomaly.

Think 2 looks like a more sensible one.

Jerry
2013-Feb-06, 03:10 AM
So the options are:
1) All the other events are anomalous, this one event that supports your beliefs is actually the norm (just haven't seen any before now)
2) This is an anomaly.

Think 2 looks like a more sensible one.

There isn't a sensible answer at this time - the data is too sparse. We haven't had search algorithyms looking for this family of events.

The sample of 'normal' Type Ia supernova events that we have at high redshift that have greater UV colour than more local events; and this hotter color is more consistant with the overluminous local sample with long light-curves. It is difficult at this time to determine which distant events correlated with which local events; and whether or not the families change with distance.

As I said in the initial post, we keep finding outliers - and now that we know there are distant events with extremely short light curves (after correction for time dilation), the search algorythms will be changed; so that we screen for shorter events. It is only after a mature sample of short light-curve events has been searched for that we can draw any conclusions about whether or not this is a rare exception, or part of a boatload of stuff nobody even tried to identify before now.

Shaula
2013-Feb-06, 06:56 AM
Well the big issue with "Data set too sparse" is that it would make our own galaxy an anomalously quiet backwater with next to no supernovae going on. In order for your purported short duration SN to actually dominate.

Yes we keep finding outliers - and you keep on making a huge deal about them when that is what they are. Outliers. It is ridiculous to ignore 90% of the bell curve and scream that the 10% that agrees with your ideas is the significant bit.

Jerry
2013-Feb-09, 03:33 AM
About a year ago, you could almost close out my argument; because the first hypernova-like event was found with a normal light-curve. (Before this one was identified, all of the hypernova at low redshift were found to have longer-than-normal light-curves.) But this "hyper-nova" in deeply redsifted with a radically short light curve space does not have a local counterpart. Even more telling (as the authors discus) is that the luminosity fades too rapidly to be associated with a type Ia (Ni Fe) event; even though it contains the spectral signature of one. Radiation decay follows specific half-life features; so a highly luminous event that fades to quickly challenges basic physical constraints.

Outliers become more and more significant when there is an absense of events we should be observing: Highly redshifted events with redshift corrected light-curves as long as the longest light-curves observed in more local events.

The James Webb Telescope has the potential to clear this up: Now that we know we have to search for shorter light-curves at high redshift; we have a real chance of obtaining a high redshift sample across a range of real magnitudes that we can compare with more local events.

It'll come.

Shaula
2013-Feb-09, 07:24 AM
We shall see. You are dealing with sparse data sets and I have never seen you do even a basic confidence assessment to try to support you assertions about what we "should" see. It has always been "we should see this because I think so" - oh and leaping upon single data points with a sort of frenzied hope that they will 'prove' something. I also see no checks for reporting bias or any attempts to delve into larger data sets yourself. It is snippets of papers here, snippets of papers there, a quote here, a new report there. Not the best way to prove anything other than selection bias. Ironic as that is then what you claim is going on.