Supernovae, Supernova Remnants and Young Earth Creationism FAQ

Contents

1. Introduction
2. What are Supernovae?
3. What are the different types of Supernovae?
           3.1 Type I Supernovae
           3.1 Type II Supernovae
4. Examples of Supernovae
           4.1 Past Supernovae
           4.2 Potential Supernovae Candidates
5. What are Supernova Remnants?
           5.1 The Life Cycle of a Supernova Remnant
6. What are the different types of Supernova Remnants?
7. Examples of Supernova Remnants
8. Supernovae and Us
           8.1 Could our Sun turn into a Supernova?
           8.2 What would happen if a Supernova occurred near to Earth?
           8.3 Is it true that the Earth wouldn't exist if it weren't for Supernovae?
9. What are Hypernovae?
10. Are Supernova Remnants evidence of a young Universe?
           10.1 The YEC Methodology
           10.2 The Rate of Supernovae Occurrence
           10.3 Numbers of Supernova Remnants
           10.4 The Age of Supernova Remnants
           10.5 Third-Stage Supernova Remnants
           10.6 The Age of Stars
           10.7 Distance to Supernovae and Supernova Remnants
           10.8 Outdated References
           10.9 Misquoting and Paraphrasing
           10.10 Conclusion
11. Notes
12. References
           12.1 Books
           12.2 Technical Papers
13. Credits

1. Introduction

ver the course of the last couple of centuries, scientists have amassed large amounts of evidence which have led them to conclude that the Universe is about 12-14 billion years old and was formed in the primordial event that scientists now call the Big Bang.

However, in the last fifty years, an offshoot of Fundamentalist Christianity has grown up (mainly in, but not limited to, the US) called Young Earth Creationism. Adherents, called Young Earth Creationists (YECs), vehemently reject most of modern science, on the basis that it contradicts their own version of Christianity, which is based on a strict literal interpretation of the Bible (and in particular, the early chapters of Genesis). Perhaps their most strident (and famous) opposition is to Darwin's Theory of Evolution.

YECs believe that the Universe, and therefore the Earth and all on it, including Humanity, were created by the Biblical God, Yahweh, in only six days approximately 6,000 years ago.

While the majority of YECs are involved in trying to refute the findings of modern science in biology and geology, a few look to astronomy and cosmology to support their beliefs. One of their approaches deals with supernova remnants, the remains of the exploding stars known as supernovae. YECs make two claims regarding supernova remnants:

• There are not enough supernova remnants observed in our Galaxy to support an ancient Universe - the numbers observed actually are indicative of a young Universe.
• There are no ancient supernova remnants thus the Universe is actually young.

This FAQ is written with the dual intention of both providing a general introduction to supernovae and supernova remnants, and addressing the YEC claims. To this end, it is split into six main parts:

1. Sections 2 to 4 provide a general introduction to supernovae, along with a detailed description of what actually happens in a supernova, and also some examples of past supernovae.
2. Sections 5 to 7 provide information on supernova remnants, the by-products of supernovae. Once again, examples of supernova remnants are provided.
3. Section 8 looks into the relationship between supernovae and us, and the dangers to the Earth posed by supernovae.
4. Section 9 dives briefly into the strange world of the phenomena known as hypernovae.
5. Section 10 deals with the assertions of the YECs.
6. Sections 11 to 13 detail notes, references and other materials used in preparing the FAQ.

2. What are Supernovae?

That's it. Literally, a supernova is an exploding star. The star explodes in a massive explosion, resulting in an extremely bright and short-lived object that emits vast amounts of energy, typically as much as an entire galaxy. As well as visible light (i.e. optical radiation), supernovae emit huge amounts of various types of radiation: X-rays, ultraviolet, infrared, gamma rays, neutrinos, cosmic rays and radio waves. The remains of the matter that is exploded away from the star during the supernova is known as a supernova remnant. Supernovae were first proposed as a distinct class of objects in 1934 by the astronomers Fritz Zwicky and Walter Baade.

3. What are the different types of Supernova?

The taxonomy of supernovae is quite complicated. Astronomers use observational criteria, not theoretical criteria, to type supernovae. Type 1 supernovae do not have hydrogen lines in their spectra¹, but Type II do. Each Type is broken down into further subclasses, depending on their light curves (Figure 1), progenitors and location - Type I into Types 1a, 1b and 1c, and Type II into Types IIL and IIP (Cappellaro & Turatto 2000).

As with most other classifications, there are exceptions. The spectra and/or light curves of a few supernovae differ sufficiently from the standard types to lead astronomers to suggest several new subclasses (Panagia et al. 1986; van Dyk et al. 1993; Baron et al. 1995, Benetti et al. 1998; Lentz et al. 2000; Filippenko 2000; Li et al. 2000; Howell 2000).

3.1 Type I Supernovae

Type Ia supernovae occur in a binary system, where one component is a white dwarf ². The gravitational attraction of the white dwarf is so intense that it is capable of siphoning off material from its companion star (Hachisu & Kato 2001). This causes the star to exceed its limit of stability - the Chandrasekhar limit³ - causing it to go into thermonuclear instability. At this point, thermonuclear incineration of the white dwarf ensues, although how exactly this occurs is still under debate, as the physics of thermonuclear burning in the degenerate matter that makes up a white dwarf is complex and still not completely understood, although there is much research going on in this area (e.g. Woosley & Weaver 1994; Branch et al. 1995; Hillebrandt & Niemeyer 2000; Hillebrandt et al. 2000; Branch 2000; Ghezzi et al. 2001). Whatever the exact mechanics, however, the result is a massive explosion that produces an extremely massive outburst of energy, some 10⁵¹ ergs, with an absolute magnitude of about -19.5 (Sandage et al. 1996; Saha et al. 1996)⁴. The star literally blows itself to bits, leaving nothing behind except a rapidly expanding remnant.

Type Ib and Ic supernovae are actually similar to Type II supernovae (they were named before astronomers really understood what they were). They occur as a giant star of about 20 solar masses evolves and loses its hydrogen envelope (the outer layers of the star) via either stellar winds (the extremely weak flow of charged particles, consisting of mainly protons and electrons, which stream from a star's outmost layer into interplanetary space) or to a binary companion (van Dyk et al. 1996); then the exposed helium core explodes. As with Type II supernovae, the explosion is triggered by the collapse of its iron core. Type Ib and Ic supernovae are (slightly) less spectacular than Type Ia supernovae. Type Ib supernovae have strong Helium lines in their spectra, whereas Type Ic supernovae have weak or no Helium lines in their spectra (Baron et al. 1996). The relationship between Type Ib and Ic supernovae and Type II supernovae is such that several Type II supernovae have been observed to have transformed into Type Ib/Ic supernovae (e.g. Finn et al. 1995; Matheson et al. 2001).

The standard Type Ia supernovae light curve shows an early peak followed by a sharp drop, then a linear decline after 50 days at a rate of 0.015 magnitudes per day. The light curves of Type Ib supernovae, though being dimmer than a Type Ia at maximum, show a similar sharp drop. However, the subsequent exponential decline differs markedly from that of Type Ia supernova, with the rate of decline less for Type Ib supernova than Type Ia, being about 0.010 magnitudes per day. The light curves of Type Ic supernovae are identical to that of Type Ib supernovae.

3.2 Type II Supernovae

These occur when a high mass star (greater than approximately 7.6 solar masses) no longer has enough fuel for the fusion process⁵ in the core of the star to create the outward pressure which combats the inward gravitational pull of the star's great mass. When this occurs, the star will swell into a red supergiant... at least on the outside. On the inside, the core yields to gravity and begins shrinking. As it shrinks, it grows progressively hotter and denser. This allows a new series of nuclear reactions to occur, forming new elements which in turn fuse to form further new elements, and so on. This allows the star to temporarily keep on shining (Table 1). All these differing reactions take increasingly shorter periods of time, and release progressively less amounts of energy⁶. As these new reactions take place, the structure of the star becomes similar to an onion - there are shells of progressively less heavy chemical elements surrounding the core.

Once the star fuses silicon into iron it hits a major snag. As can be seen, the above reactions create energy (an exothermic reaction). But to convert iron into heavier elements requires energy (an endothermic reaction, which requires approximately 2 MeV per Nucleon). Thus, fusion halts. At the very high temperatures now present in the core of the star (much greater than 10⁹ K), a process known as photodisintegration⁷ occurs. Due to the loss of energy that occurs due to photodisintegration, the core starts to rapidly collapse. Different parts of the core collapse at different rates, with the result that the inner core decouples from the outer core, leaving it behind. During the collapse, speeds can reach 7,000 km s^-1 in the outer core, and within about one second, a volume the size of Earth has been compressed down to a radius of 50 kilometres. As a result the rest of the star is left in the precarious position of being almost suspended above the catastrophically collapsing core. This collapse of the inner iron core continues until the density there exceeds approximately 8 x 10¹⁷ kg m^-3. At this point, the material that now makes up the inner core stiffens (as a result of the nuclei of the atoms present becoming repelled by each other), with the result that the inner core now rebounds somewhat, sending pressure waves outwards into the infalling material of the outer core. These pressure waves, when they reach the local speed of sound, form a shock wave that starts moving outwards.

As the shock wave propagates outwards, it encounters the falling inner iron core. The extremely high temperatures that occur as a result of this cause further photodisintegration, robbing the shock of most of its energy⁸. If what's left of the iron core is not too massive (less than 1.2 solar masses), the shock will fight its way through the rest of the outer core - which takes about twenty milliseconds, and collide with the remainder of the outer layers of the star. On the other hand, if the iron core is massive enough, the shock stalls, becoming nearly stationary, with infalling material now accreting onto it. At this point, the neutrinos now streaming from the core (due to the conversion of the iron core into essentially a neutron core) superheat the material beneath the shock wave; the resulting plumes of hot material push the shock wave outwards and allow it to continue its march towards the surface, driving all before it (Janka 2001). As the shock encounters material in the star's outer layers, the material is heated, fusing to form new elements and radioactive isotopes (Meyer et al. 1995; Thielemann et al. 1996). The shock then propels the various outer layers of the star out into space, leaving the inner core behind. The total energy in the expanding material is in the order of 10⁵¹ ergs (or less). Vast amounts of photons are released, resulting in a spectacular optical display, the equivalent of 10⁹ suns, giving an absolute magnitude of about -18. Due to the radioactive decay of heavy elements produced in the explosion (Mochizuki & Kumagai 1998; Hernanz 2000; Wanajo et al. 2001), it then starts to slowly fade, at rate of approximately six to eight magnitudes each year. Type II supernovae are not as luminous as Type Ia supernovae, by a factor of at least three. The mechanics of this type of supernovae are dealt with in detail by Bethe (1993), Wallerstein et al. (1997), Mezzacappa (2000) and Liebendoerfer et al. (2001).

If the mass of the core remnant is beneath about three solar masses, it will become a neutron star⁹ (rapidly rotating neutron stars are known as pulsars¹⁰). If it exceeds about three solar masses, it continues to contract. The gravitational field of the collapsing star is so powerful that neither matter nor light can escape it. The "star" then collapses to a black hole (Balberg & Shapiro 2001), a singularity or point of zero volume and infinite density, hidden by an event horizon at a distance called the Schwarzschild radius¹¹. Bodies crossing the event horizon, or a beam of light directed at such an object, would seemingly just disappear - pulled into a "bottomless pit". In either case, the creation of these rather exotic objects is accompanied by a tremndous production of neutrinos, the majority of which escape into space with a total energy approaching 3 x 10⁵³ ergs¹².

The majority of Type II supernovae are split into II-L (linear) or II-P (plateau) subclasses, depending on their light curves - Type II-P display a plateau soon after maximum luminosity.

4. Examples of Supernovae

4.1 Past Supernovae

Since the earliest days of mankind looking up into the skies, we have saw many bright points of light in the sky which appeared suddenly and then slowly faded away over the course of many months. Most of these "guest-stars", as the ancient Chinese called them, were novae of various sorts, but some were genuine supernovae. The most reliable records come from Asia, where Korean, Japanese and Chinese astronomers kept suprisingly accurate records dating back as far as 1400 BC - Wang (1986) reported that there were 90 probable novae and supernovae listed in Chinese records between 1400 BC and 1700 AD. In Europe, on the other hand, the earliest known observation of what we now know to be a supernovae was not until the 11th century AD. As a result of intensive study of these records, and later reports by European astronomers like Tycho and Kepler, astronomers are now aware of quite a few Galactic supernovae having occurred in the last couple of thousand years (Table 2)¹³.

The first extragalactic supernova ever discovered was SN 1885A near the nucleus of M31 (the famous "Andromeda Galaxy") on 20 August 1885. SN 1885A had an apparent visual magnitude of 5.85 - it would have been just barely visible to the naked eye had not the glow from M31 overwhelmed it (de Vaucouleurs & Corwin 1985).

Probably the most famous extragalactic supernova was observed on the 24th February 1987 in the Large Magellanic Cloud A blue supergiant star (of about 20 solar masses) called Sanduleak -69 202 (its previous apparent visual magnitude was a lowly 10.2) exploded in a burst of light visible to the naked eye (when discovered on a photographic plate by the astronomer Ian Shelton of the University of Toronto at the Las Campanas Observatory in Chile, it was magnitude 4.5 - it later peaked at magnitude 2.8 before fading slowly over time (Shelton 1993)) - thus it was a Type II supernova. It was designated SN 1987A¹⁷. In the subsequent years, a bright supernova remnant was seen to form around the star in the form of an expanding shock wave. Only now, years later, is the shock wave reaching rings of previously existing gas surrounding the now dead star (Chu 2000)¹⁸. This is causing the knots of gas to glow brightly. There are many images of SN 1987A available on the WWW. Perhaps the definitive review of SN 1987A is Arnett et al. (1989), although this does not cover more recent developments.

SN 1987A was extremely important to astronomers as it was the first supernova which astronomers could study in great detail with modern astronomical instruments. It confirmed a whole host of predictions that astronomers had made regarding supernovae, including:

• the production of radioactive isotopes, for example ⁵⁶Ni and ⁵⁷Ni and their subsequent decay to ^57,56Co and ^57,56Fe, or ⁴⁴Ti which decays to ⁴⁴Sc and then to ⁴⁴Ca, which are largely responsible for the shape of the light curves of supernovae (Arnett et al. 1989; Li et al. 1993; Knodlseder 2000; Lundqvist et al. 2001)
• the production of neutrinos (Bionta et al. 1987; Hirata et al. 1987; Arnett 1988; Burrows 1988; Arnett et al. 1989)

As observing techniques and equipment improve, more and more supernovae are discovered each year (Figure 2). 172 supernovae were discovered in 2000. At the time of writing, 102 supernovae have been discovered in 2001. A complete list of all supernovae discovered since SN 1885A is maintained by the Central Bureau of Astronomical Telegrams (CBAT).

4.2 Potential Supernovae Candidates

The three nearest candidates for supernovae sometime in the near future (astronomically speaking) are all nearby (again, in astronomical terms) red giants: Betelgeuse (in Orion) at 430 light years, Antares (in Scorpius) at 600 light years, and Rasalgethi (in Hercules) which lies 380 light years from Earth. These will all be Type II supernova. There is a closer red giant - the star Scheat in Pegasus, this is 200 light years away and although this is currently a red giant, the progenitor star is almost certainly not large enough to go supernovae, instead, the outer layers will slowly drift off into space forming a planetary nebula, and leaving a white dwarf behind.

However, it is more likely that the next Type II supernova in our Galaxy might either be the highly evolved orange supergiant HD 179821 (Jura et al. 2001) or the blue supergiant Sher 25. Even though both stars are extremely luminous, they lie at considerable distances from Earth, and thus are not visible to the unaided eye.

Sher 25 has an age of about three million years, but has a mass of about 120 solar masses, which makes it one of the most massive stars ever observed. As it dies, it is blowing parts of its own outer envelope away at speeds of 20-83 km s^-1. As with Sanduleak -69 202, a bubble of gas has formed surrounding the star, which are seen as filaments and a ring shaped structure (Petersen 1999). Indeed, both the star and the surrounding material resemble Sanduleak -69 202 closely, although there are some minor differences, most probably due to differences in the environment surrounding the star (Brandner et al. 1997b). The gases and dust around the star are nitrogen enhanced - the sign of an evolved, very hot star rapidly burning through its hydrogen and helium and forming other elements in the process (Brandner et al. 1997a). Perhaps in a few tens of thousands of years, or maybe even tomorrow, Sher 25 will explode like Sanduleak -69 202, and provide another spectacular display of cosmic fireworks.

Recently, astronomers have suggested that the binary star KPD 1930 + 2752 is a future candidate for a future Type Ia supernova event . The primary star in this system is a subdwarf-B star, and it has an unseen companion star that is almost certainly a white dwarf. The orbital period is only 2 hours 17 minutes. The total mass of the system is 1.47 Solar Masses, above the Chandrasekhar limit. Maxted et al. (2000) proposed that the binary will merge within approximately 200 million years due to orbital shrinkage and the evolutionary expansion of the primary), and when this happens, because of accretion of helium and other elements heavier than hydrogen onto the white dwarf, a Type Ia supernova would occur. Some other astronomers have disputed this scenario, claiming that because the B-star would form a white dwarf before merging with its companion, the total mass of the system would be beneath the Chandrasekhar limit , thus no Type 1a supernovae would occur (Ergma et al. 2001).

It has also been determined that supernovae can be responsible for the production of runaway stars (or at least a proportion of them). These are stars that were originally part of a multiple star system. Sometime in the past, one of their companions went supernovae, and the force of the blast pushed the star off into space at a very high velocity (Blaauw 1961; Hills 1983; Stone 1991; Kaper et al. 1997; Hoogerwerf et al. 2000, 2001).

5. What are Supernova Remnants?

A supernova remnant (usually abbreviated to SNR) is the remains of the matter that is exploded away from a star when it goes supernovae. This ejection of matter is much more violent than occurs in the planetary nebula that mark the end of a low mass star, giving expansion speeds of 1000-10,000 km s^-1. The ejected matter sweeps up surrounding gas and dust as it expands producing a shock wave that excites and ionises the gas, which results in the production of X-rays, and radio waves in the form of synchrotron radiation. This plasma may reach temperatures of 1,000-1,000,000 K, but with densities of only about a million particles per cubic metre. Gradually, the expansion rate slows down, seeding the local neighbourhood with heavy elements, but not before the remnant occupies an area of space dozens or hundreds of light years in diameter.

5.1 The Life Cycle of a Supernova Remnant

In the classical model of SNR evolution (Woltjer 1972; Gull 1973; Chevalier 1977), there are four stages or phases:

• In the first phase, known as free expansion, the front of the expansion is formed from the shock wave interacting with the ambient interstellar medium (ISM). This phase is characterised by constant temperature within the remnant and constant expansion velocity of the shell. This phase can last anywhere from 90 years to over 300 years.
• During the second phase, known as the Sedov or Adiabatic Phase, the remnant material slowly begins to decelerate and cool. In this phase, the main shell of the remnant is unstable, and the remnant's ejecta becomes mixed up with the gas that was just shocked by the initial shock wave. This mixing also enhances the magnetic field inside the remnant shell. This phase can last anywhere from 100-100,000 years.
• The third phase, the Snowplough or Radiative phase, begins after the shell has cooled down to about 1,000,000K, so the shell can more efficiently radiate energy. This, in turn, cools the shell faster and thus making it shrink and become denser. This makes it cool faster still. Because of this snowball effect, the remnant quickly develops a thin shell and radiates most of its energy away as light. The velocity now decreases fairly rapidly. Outward expansion stops and the remnant starts to collapse under its own gravity. This phase can last hundreds of thousands of years.
• The fourth phase, known as Dispersal. Here the shell breaks up when the velocity of the "snowplough" becomes subsonic, and what's left of the remnant dissipates into the ISM.

This is the theoretical model, in reality astronomers have found that most SNRs do not follow this standard model (Harrus et al. 2001). Some of the reasons for this are:

• The ISM in which supernovae occur is rarely isotropic or of a uniform consistency and density, which leads to asymmetry and differences within the remnant (Dohm-Palmer & Jones 1996; Maciejewski & Cox 1999; Slavin et al. 2000).
• If a supernova occurs in a pre-existing bubble of interstellar material surrounded by a massive shell of gas then the Sedov phase will not necessarily occur (Wheeler et al. 1980; Franco et al. 1991; Franco 1994; Gvaramadze 2000), indeed, the SNR may not be detectable at all in this scenario unless it hits the walls of the shell (Fich 1986; Koo & Heiles 1995; Chu 1997).
• If the density of the medium in which the SNR is located is low enough, it is possible for the SNR to finish its life by merging with the ISM before cooling becomes important (Asvarov 2000).
• Different stages can occur simultaneously in different locations within a single remnant (Cioffi et al. 1988; Tenorio-Tagle et al. 1990; Franco et al. 1994; Jones et al. 1998; Asvarov 2000; Bykov et al. 2000; Reynoso & Mangum 2001).
• If the ISM is strongly magnetized, then the evolution of the SNR will differ in terms of the length of the various phases and the overall shape of the remnant (Insertis & Rees 1991).

6. What are the different types of Supernova Remnants?

There are three generally accepted types of SNRs. Note that the categories are not set in stone - SNRs have been observed in the process of gradually transforming from one type to another (Sakhibov & Smirnov 1982; Lazendic et al. 2000). The three types are:

• Shell-type remnants: As the shock wave from the supernova explosion ploughs through space, it heats and stirs up any interstellar material it encounters, thus producing a big shell of hot material in space. A ring like structure in this type of remnant is seen because at the edge of the shell there is more hot gas in our line of sight than in the middle. Astronomers call this phenomenon limb brightening. The vast majority of SNRs are of this type.
• Crab-like remnants: These remnants, also known as "plerions" (a term first suggested by Weiler & Panagia (1978), and from the Greek word meaning "full") are similar to the Crab Nebula. They are similar to shell-type remnants, except that they contain a pulsar in the middle that blows out jets of very fast moving material. These remnants look more like a "blob" than a "ring."
• Composite Remnants: These remnants are a cross between the shell-type remnants and crab-like remnants. They appear shell-like, crab-like or both depending on what part of the electromagnetic spectrum one is observing them in. There are two kinds of composite remnants -- thermal and plerionic. Thermal composites appear shell-type in the radio waveband (synchrotron radiation). In X-ray wavelengths, however, they appear crab-like, but unlike the true crab-like remnants their X-ray spectra have spectral lines, indicative of hot gas. Plerionic composites appear crab-like in both radio and X-ray wavebands, however they also have shells. Their X-ray spectra in the centre do not show spectral lines, but the X-ray spectra near the shell do have spectral lines.

Rho & Petre (1998) proposed a fourth class of SNRs - the so-called "mixed-morphology SNRs". These remnants are classified as shell-type remnants at radio wavelengths, but the X-ray morphology is centrally peaked. In addition, the X-ray emission is thermal which comes from the ISM, not the ejecta making up the SNR. And finally, there is no prominent, central, compact source in radio or X-ray bands (i.e. there is no pulsar).

7. Examples of Supernova Remnants

Some of the more famous SNRs (easily visible in small telescopes) include:

• The Crab Nebula in Taurus,
• The Veil Nebula in Cygnus,
• The Vela SNR in Vela,
• The Puppis A SNR in Puppis.

SNRs are officially designated according to their Galactic coordinates. Thus the SNR at 184.55^o Galactic Longnitude and -5.78^o Galactic Latitude, commonly known as the Crab Nebula, is designated as G184.6 - 5.8.

There are many galleries of SNR images on the WWW - perhaps the two most extensive are the ROSAT X-ray Satellite Gallery and the Chandra X-ray Satellite Gallery. According to Green (2000), there are 225 confirmed SNRs in our Galaxy, with another 61 possible or probable remnants, with more are being discovered all the time (e.g. Bhatnagar 2000; Crawford et al. 2000; Combi et al. 2001; McClure-Griffiths et al. 2001). There are hundreds known in other galaxies (e.g. Danziger et al. 1979; van den Bergh 1983; Peimbert et al. 1988; Long et al. 1990; Braun & Walterbos 1993; Gordon et al. 1993; Muxlow et al. 1994; Yang et al. 1994; Huang et al. 1994; Cowan et al. 1994; Magnier et al. 1995; Matonick & Fesen 1997; Dunne et al. 2000; Schlegel et al. 2000; Rosado et al. 2001). There are nine known plerions in our Galaxy, and twenty-three known Composite remnants, the rest are Shell-Type remnants (Green 2000), although the proportion of plerions is expected to rise in the future as SNRs which are currently classified as Shell-Type or Composite are examined more closely (Gaensler 2000). Since the "mixed-morphology" category was proposed, some astronomers have been re-examining existing SNRs with a view to recategorising them as mixed-morphology SNRs, so far about nine have been identified (Yoshita et al. 2001).

8. Supernovae and Us

8.1 Could our Sun turn into a supernova?

Relax, the answer is an emphatic no! Our sun is nowhere near massive enough to become a Type II supernova and there's no white dwarf companion to become a Type Ia supernova. Besides, it will take another five billion years before our sun's supply of hydrogen is depleted. At that time it will begin its dying process and eventually become a white dwarf with a surrounding shell of material much like the Ring Nebula (M57) in the constellation of Lyra, i.e. a planetary nebulae. This is still small consolation for us on Earth however, as in another billion years or so the Sun will have increased in luminosity so much that the Earth will likely become totally uninhabitable.

8.2 What would happen if a Supernova occurred near to Earth?

In short, life on Earth would be in big trouble. Depending on the distance and the type, the massive amounts of radiation emitted by supernovae could mean possibly all or most of the life on Earth would be fried. From an article by Michael Richmond on the risks to Earth from nearby SNRs:

I suspect that a Type II explosion must be within a few parsecs of the Earth, certainly less than 10 parsecs (32.6 light years), to pose a danger to life on Earth. I suspect that a Type Ia explosion, due to the larger amount of high energy radiation, could be several times farther away. My guess is that the X-ray and gamma ray radiation are the most important at large distances.

Interestingly, there is a possibility that a supernova exploded close to earth (within 100 light years) about five million years ago (Ellis et al. 1996; Fields & Ellis 1999). Could this have caused an extinction event? Quite possibly. Was there one that correlates to this supernova? Probably not¹⁹.

There is also some evidence of another supernova occurring within 600 light years of the Sun within the last couple of million years and which was responsible for the nearby shell of gas known as the North Polar Spur (Cruddace et al. 1976; Hayakawa et al. 1977; Davelaar et al. 1980; Heiles et al. 1980; Egger & Aschenbach 1995), although there are alternative non-supernova explanations (Sofue 1977). There are other similar large shells of gas in the general Galactic vicinity (Nousek et al. 1981).

The famous Geminga pulsar (aka 2CG195+4), which lies close to the Crab Nebula in the sky, has also been proposed as the remains of a supernova that occurred 300,000 years ago. Gehrels & Chen (1993) proposed that this supernova is the cause of the Local Bubble²⁰, whereas Cunha & Smith (1996) proposed that the supernova was the cause of the loop of gas surrounding the star Lambda Orionis in the constellation of Orion, again, roughly 300,000 years ago. Innes & Hartquist (1984) also proposed that the Local Bubble was the result of a past supernova, whereas Smith & Cox (1998, 2001) have suggested that instead of being the result of one supernova, the Local Bubble is the result of repeated supernovae.

Incidentally, Geminga lies approximately 510 light years from Earth (Caraveo et al. 1996a; Caraveo et al. 1996b). Although it was discovered in 1975 as a source of high-energy gamma-rays (Thompson et al. 1977; Bennett et al. 1977), it was not until 1992 that astronomers worked out what it actually was (Bertsch et al. 1992; Halpern & Holt 1992). This is reflected in its name - "Geminga" is Milanese for "That which does not exist".

As a brief aside, the Old Earth Creationist Hugh Ross²¹, has stated:

According to Genesis 5 and 6, one of the many changes God decreed at the time of the Genesis Flood was the shortening of human lifespans from an average of 900+ years down to a maximum of about 120 years. Besides protecting us from intensification of evil, this change, which apparently involved a reprograming of our cells, also protects us from certain types of cancer. The change also involved either the removal of some sort of pre-Flood radiation shield or, more likely, an increase in the amount of cosmic radiation showering the Earth.

He identifies the supernova responsible for the Vela SNR as being one possible cause of this change of life span. However, on the assumption that such a change in human lifespans did take place as reported in Genesis, there are still two major problems with this claim:

• The Vela supernova was between 800 and 1,600 light years away (Gvaramadze 2001a)- at this distance the radiation hitting the Earth from the supernova itself is negligible (about the same as several hours normal sunshine at the surface of the Earth), mainly due to distance and the protective effects of the Earth's atmosphere²².
• In any case, such a dose of radiation as Ross proposes would have left traces of isotopes in various sediment layers - no such traces in these layers have been found.

In short, there is no evidence to support this claim, and plenty of evidence against it.

8.3 Is it true that the Earth wouldn't exist if it weren't for Supernovae?

The Big Bang produced very little but hydrogen and helium, with some lithium (Thielemann et al. 2001). Various other elements (heavier than carbon but lighter than iron) are produced by fusion in the red giant stage of stars (Table 3). Elements heavier than iron get produced mainly in supernovae, specifically in the explosive nuclear burning that takes place either during the phase where the shock wave that results from the collapse of the star's core encounters the outer layers of the star (for Types Ib, Ic and II supernovae), or in the general nuclear fireball that Type Ia supernova become. In the aftermath of a supernova event, the local ISM is saturated with these heavy elements. The supernovae and the resulting blast wave heat and stir up the ISM. For stars that don't go supernova, most of their heavy elements get locked up in the white dwarf that they end up as These elements are slowly distributed through the stellar wind and other forms of mass loss (Vink et al. 2001).

Astronomers investigating a class of meteorites known as carbonaceous chondrites (so-called becuse they contain carbon and are characterised by small inclusions or chrondules of molten material within them) have found by-products of short-lived radioactive isotopes which are produced either exclusively or mainly in supernovae (Lee et al. 1978; McCulloch & Wasserburg 1978; Clark 1979; Arnould et al. 1980; Dearborn et al. 1988; Nittler et al. 1996; Ott 1996; Timmes et al. 1996; Amari et al. 1996; Hernanz 2000). As such meteorites are thought to be primeval remnants from the time of the early solar system, approximately 4.6 billion years ago, it indicates that at some point before the formation of the solar system, a supernova occurred.

To sum up, most of the elements that make up the computer you're using to view this article, the world around you, the solar system and your body, were originally produced in a supernova (Cameron & Truran 1977; Harper 1996). As the singer Joni Mitchell put it, "We are stardust...". So the answer is yes - without these supernovae, it is very likely that us (humanity) and everything else on the Earth (and also the Earth itself) would not exist.

9. What are Hypernovae?

It has been proposed that not all massive stars successfully launch supernova events. If the core of a massive rapidly rotating progenitor collapses into a black hole and absorbs the surrounding stellar mantle without producing a neutrino driven explosion, the result is a collapsar, and the massive release of energy produced is called a hypernova. Hypernovae are typically 150-200 more massive than the Sun and explode with a total energy output of more than 10⁵² ergs (Nakamura et al. 2001) - many times more than the energy output of a typical supernova.

Hypernovae have been proposed as a way to explain the existence of Gamma-ray Bursts²³ (Woosley 1993; Paczynski 1997; MacFadyen & Woosley 1999; MacFadyen 1999). Gamma-ray Bursts (GRBs). Astronomers have identified several hypernova-type events that appear to be linked to observed GRBs (Hansen 1999; Bloom et al. 1999; Chu et al. 1999; Filippenko 2000; Iwamoto et al. 2000).

However, other explanations for GRBs have been put forward - these include

• the merger of two neutron stars or the merger between a neutron star and a black hole (Narayan et al. 1992).
• a helium star falling into a black hole (Zhang & Fryer 2001).
• the merger of two helium stars (Belczynski et al. 2000).
• pulsars emitting relativistic jets that precess past our line of sight (Blackman et al. 1996).
• the "supranovae", in which a neutron star gravitationally implodes to a black hole (Vietri & Stella 1998).
• relativistic jets emitted in Type II supernovae (Dar 2000).
• the collapse of a supermassive star (SMS) into a supermassive black hole (SMBH) (Linke et al. 2001)
• a combination of some or all of the above (Chevalier & Li 1999).

Currently, none of the hypotheses have been confirmed to the exclusion of the others - indeed it is now apparent that there are different types of GRBs which could be caused by differing processes (Piro et al. 2000). More detailed discussion of GRBs is beyond the scope of the FAQ, but Meszaros (1999, 2001), Antonelli et al. (2000) and Piran (2001) all give good overviews of our current understanding of GRBs.

10. Are Supernova Remnants evidence of a young Universe?

All of the YEC literature on the WWW concerning supernovae and supernova remnants originates from one article written in 1994 by a Canadian, Keith Davies, entitled "Distribution of Supernova Remnants in the Galaxy". This article is part of the Creation Discovery Project. Various versions and summaries of this article appear on various other YEC web sites including Answers in Genesis (by Jonathan Sarfati), Creation in the Crossfire (by Jon Colley), Creation Online and He Comes....²⁴. According to the Creation Science Association For Mid-America, Davies' article is based on a presentation²⁵ he gave at the Third International Conference on Creationism in 1994.

The first line of the Creation Online article sums up the YEC argument nicely:

The results of observations done by astronomers indicate that there are not enough supernovas to justify an old galaxy. The numbers observed are consistent with a young galaxy of thousands of years old."

Sarfati elaborates further:

...a young universe model fits the data of the low number of observed SNRs. If the universe was really billions of years old, there are 7000 missing SNRs in our Galaxy.

How do the YECs arrive at this conclusion? In Davies' original article, he estimated the numbers of SNRs visible in an ancient Universe (billions of years old) and in a young Universe (thousands of years old), and compared both values with the actual number of observed SNRs (Table 4). He used the following methodology:

1. Assume a rate of supernovae occurrence in our Galaxy of one every 25 years.
2. Assume that the first stage of SNR expansion ends after 317 years, the second after 120,000 years and the third after a million years.
3. Due to radio telescope observational limitations, assume that only 19% of first-stage SNRs, 47% of second-stage SNRs and 14% of third-stage SNRs are observable.
4. Dividing the ages of the various states by the rate of supernovae occurrence, work out the numbers of SNRs of each type visible if the universe was ancient.
5. Calculate the numbers of SNRs of each type visible if the universe was only 7,000 years old.
6. Compare the two results to the actual number of observed SNRs.

Do the claims of Davies and others stand up to scrutiny? As it happens, not very well.

10.1 The YEC Methodology

Davies claims that:

The number of Supernova Remnants (SNRs) observable in the Galaxy is consistent with the number expected to be formed in a Universe that is 7,000 years old.

However, using Davies' own methodology, the actual number of observable SNRs in our Galaxy (225 from Green (2000)) gives a value of 11,970 years, not 7,000 years. The 11,970 is the lowest possible value for the age of the Universe as derived from his methodology. Yet a strict reading of the lineages in Genesis gives the date of creation as being 4004 BC, about 6,000 years ago (as famously calculated by Archbishop Ussher of Armagh in the 17th century).

His calculations are repeated here, using his values (for convenience, any fractions are rounded to the nearest whole number - it makes very little difference to the results):

• Number of first-stage SNRs in total = 317 (length of first stage of an SNR's lifetime) / 25 = 13. Number of observable first-stage SNRs is thus 19% of 13 = 2. This figure is the same for both an old universe and a young universe.
• Number of second-stage SNRs in total with an old universe = 119,683 (length of 2nd stage of an SNR's lifetime) / 25 = 4,787. With a young universe = 6,683 / 25 = 267. Number of observable first-stage SNRs with an old universe thus 47% of 4,800, or 2,250. Number of observable second-stage SNRs with a young universe = 47% of 267 = 126.
• Number of third-stage SNRs in total with an old universe = 880317 (length of 3rd stage of an SNR's lifetime) / 25 = 35213. With a young universe = 0 / 25 = 0. Number of observable third-stage SNRs with an old universe thus 14% of 35213, or 4,930. Number of observable third-stage SNRs with a young universe = 14% of 0 = 0.

But if the second calculation is repeated with the actual number of observed SNRs in our Galaxy, 225, then according to his methodology, the result is 479 SNRs in total, and if the rate of supernovae occurrence is one every 25 years, then 25 x 479 is 11,970 years.

There is a mathematical error in Davies' calculation. Surely if only 47% of second stage SNRs are visible, then the number of visible SNRs in a young universe is not 267/268 but 47% of this = 126? But to quote Davies:

Total [number] of Second Stage SNRs expected to be observed under a 7,000 year old Universe with t* =25 approx. 268.

In fairness to Davies, this is most likely a simple mathematical mistake, but both Sarfati and the author(s) of the Creation Online article correct this error, and give the value of 126 for visible second-stage SNRs, without telling the reader that they have done so, and without pointing out the error in Davies' original article. The other articles propagate the erroneous value of 268.

Davies devotes a large part of his article to calculating the percentages of SNRs at different stages in their lifetimes that should be visible. He gets the results of 19%, 47% and 14% for first, second and third stage SNRs respectively). However, these figures are wrong. One of the components he uses in the calculation is the relationship known as Sigma-D (obtained from Ilovaisky & Lequeux 1972a), i.e. the relationship between the surface brightness at a specific radio frequency and the linear diameter of an SNR. Although it can be used for determining distances to SNRs (Goebel et al. 1981; Huang & Thaddeus 1985; Case & Bhattacharya 1998), it only works for shell-type SNRs that have the same supernova explosion energy and mechanism, and are evolving in identical environments, whereas Davies assumes that it holds for all SNRs. More information on measuring distances to SNRs is given in Section 10.7. As Green (1991) states,

It is not possible to quote a single surface-brightness completeness limit for current catalogs of SNRs, not only because the background emission varies in different regions of the Galactic plane but also because different regions have been surveyed with different instruments.

Davies also uses the radio observational limitations from Ilovaisky & Lequeux (1972a) to give proportions of SNRs in various stages of evolution (19%, 47% and 14%). There are many problems with Davies' approach:

• Davies' proportions are gross simplifications - it is simply not possible to work out accurate figures for the proportions of SNRs visible at various stages in their lifetimes, as there are too many external factors involved (see Section 10.3).
• Others (e.g. Kodaira 1974; Vettolani & Zamorani 1977; Leahy & Wu 1989) have updated and corrected Ilovaisky & Lequex's original findings - corrections Davies has not taken into consideration, despite their papers being published well before 1994.
• Diameters of SNRs are crucial to Davies' calculations. But to calculate the diameter of a SNR, the distance must be known with accuracy. Davies gives the impression that the diameters and distances of Galactic SNRs that he uses are accurately known, when this is not the case (see Section 10.7 and Green (1984, 1991) for more details).
• In radio astronomy, flux density is defined as power received per unit area per unit frequency. The unit of flux density is the Jansky (Jy) and is equivalent to 10^-26 W Hz^-1 m^-2. The amount of radiation in the radio spectrum emitted by a SNR is measured in Jy, and a radio telescope with a low flux density limit is much more sensitive to SNRs with low flux densities. Davies has limited himself to SNRs with a flux density of greater than 10 Jy. Out of the 225 confirmed SNRs in our galaxy, only 90 (40%) have flux densities of 10 Jy or more. When Davies wrote his article, the numbers of SNRs with flux densities of 10 Jy or more was 102 out of 176 in total. The other remnants, 74 in 1991 and 135 today, Davies completely ignores. These SNRs are effectively dimmer and harder to detect, and thus are probably either further away or larger (and thus older). By ignoring these, Davies is claiming that he finds no old remnants, but he is doing so after eliminating large numbers of possible old remnants from his calculations!

10.2 Rate of Supernovae Occurrence

The main source for Davies' value of 25 years for the rate of supernova occurrence in our Galaxy is an estimate made in 1970 by the Swiss astronomer Gustav Tammann. The value Tammann gave was 26 ± 10 years, calculated by comparing our Galaxy to other similar galaxies (in terms of size and luminosity) and working out the rate of supernovae from observing them (Tammann 1970). Poveda & Woltjer (1968) gave a rough estimate of 60 years, whereas Chai & van den Bergh (1970) estimated 100 years, and Ilovaisky & Lequeux (1972b) gave 50 ± 25 years.

In 1994, Tammann revised his 1970 estimate for the rate of occurrence of supernovae in our own Galaxy to 40 ±10 years (Tammann et al. 1994). Cappellaro et al. (1996) suggested that for our Galaxy there should be 4 ± 1 Type Ia, 2 ± 1 Type Ib/c and 12 ± 6 Type II observable supernovae per millennium, which works out at rate roughly half that of Tammann's 1970 value. Weiler & Sramek (1988) conclude that the average interval between supernovae in our Galaxy is between 20 and 50 years and van den Bergh & Tammann (1991) and Turatto (1999) both give estimates in reasonable agreement with this. The values given in Carroll & Ostlie's Introduction to Modern Astrophysics are 36 years for Type I supernovae and 44 years for Type II supernovae. Folgheraiter et al. (1997) gives an average interval of 30 years as being "the currently accepted value".

In the 1940s, 1950s and early 1960s, astronomers found that supernovae occurred at different rates in different types of galaxies, and that the rate of supernovae in spiral galaxies is dependent on the luminosity of the galaxy involved (Tammann et al. 1994). In addition, the rate at which supernovae are observed to occur in other galaxies is dependent on the inclination of the galaxy - a lot more supernovae are detected in galaxies that lie pole-on to us (van den Bergh & McClure 1990; van den Bergh & Tammann 1991). Another determining factor in the rate of supernova occurrence is the amount of progenitor stars - either suitable binary systems (for Type I) or massive giant stars (for Type II) are required.

Dragicevich et al. (1999) have proposed that the Earth is situated favourably within our Galaxy for viewing supernovae, thus the calculated rate for supernovae occurrence is actually high compared with the rate throughout the Galaxy as a whole.

Astronomers are generally quite cautious in inferring rates of supernova occurrence from the amount of supernova remnants. Indeed, to quote from Jones et al. (1998):

People should exercise extreme caution in inferring supernovae rates from counts of mature and old SNRs.

And from van den Bergh & Tammann (1991):

Since the lifetimes of radio supernova remnants (SNRs) depend so critically on environment, it will be very difficult to derive meaningful information on supernova rates from the statistics of SNRs.

However, on balance, Davies uses an acceptable value for the rate of Galactic supernovae occurrence. Incidentally, it is thought that one supernova occurs every second in the whole Universe (Burrows 2000).

10.3 Numbers of Supernova Remnants

YECs claim that not as many SNRs are observed as would be expected in an old universe. Davies uses a value of one million years for the lower end of the typical visible lifetime of a SNR and assumes that all SNRs last this long. He gets this figure from Ilovaisky & Lequeux (1972b). However, on reading the original paper it is noticeable that this value is actually for the theoretical lifetime of the remnant, not the observable lifetime of the remnant. Why is there a difference? Quite simply, SNRs are actually hard to detect. Factors that seriously hinder our ability to detect SNRs (and which Davies almost completely ignores) are:

• SNRs can only be observed in a small proportion of our Galaxy - our view of most of the Galaxy is blocked by large amounts of dust and interstellar matter. Only some younger, radio emitting SNRs would be visible through this dust (Sramek et al. 1992; Gray 1994). This largely explains why there has been no observed Galactic supernovae in the last 300 or so years (Clark et al. 1981; Dawson & Johnson 1994; Hatano et al. 1997), even though we would have expected perhaps 5-10 to have occurred (McKee 2000).
• It is also difficult to identify much older remnants as they either have faded beyond our ability to detect them (they may have merged with the ISM), they have merged with other remnants, or they have faded into the general background "noise" (Nousek et al. 1981; Matthews et al. 1998; Braun et al. 1989; Landecker et al. 1990; Normandeau et al. 2000). Younger SNRs, or SNRs which are still interacting with gas expelled by their progenitors are much more likely to be detected (Jones et al. 1998; Slavin & Cox 1992). Shull et al. (1989) carried out a statistical analysis of SNRs, and found that with isolated SNRs, less than 1% last for longer than 100,000 years, and only 20% are still intact after 50,000 years.
• The make-up of the local ISM that the supernova occurs in is critical to the observability of the resulting SNR (Dohm-Palmer & Jones 1996). SNRs in regions where the density of the ISM is low (Henning & Wendker 1975; Gaensler & Johnson 1995b) or there is little ionised gas present (Heiles et al. 1980) may not be readily visible. Indeed, it may be the case that as few as 15-20% of supernova events cause observable SNRs (Clark & Stephenson 1977; Clark 1979; Kafatos et al. 1980).
• Some young SNRs can be intrinsically faint at radio wavelengths and thus unusually difficult to detect (Gray 1994; Duncan & Green 2000).
• SNRs are obscured by and can be indistinguishable from other interstellar emission nebulae, and their spectra can be similar to powerful distant radio galaxies and quasars (White & Becker 1990; Inglis & Kitchin 1990; Caswell & Stewart 1991, 1992; Williams et al. 2000). In other words, there is a lot of clutter out there, and finding SNRs is often a tricky and difficult task. Indeed, only a minority of SNRs are visible at optical wavelengths (Long et al. 1990).
• The limits of the equipment used to detect SNRs (usually radio telescopes) impinge upon our ability to observe supernova remnants (Green 1991; Kassim 1992; Frail et al. 1994). As this gets better in the future, the numbers of SNRs detected will rise. This can be illustrated by the way astronomers have detected more and more SNRs in our own galaxy over the last few decades - in 1984, there were only 174 Galactic SNRs known, and back in 1971, only 113 (Downes 1971).
• Not all the sky has been surveyed to the same degree - there are still large areas of the sky (mainly in the southern celestial hemisphere) waiting to be surveyed with more powerful instruments (Case & Bhattacharya 1998).

As a result, Davies vastly overestimates the actual number of observable SNRs. Berkhuijsen (1984) suggested that there might be 1,000 to 10,000 SNRs in our Galaxy (depending on the lifetime of SNRs), but this is the only estimate I'm aware of that provides a figure anywhere near Davies', but even then, Berkhuijsen's estimate is for the total number of SNRs, and not for the observable SNRs.

However, Berkhuijsen's value is very much the exception. Most other estimates for the total number of SNRs in the Galaxy are around 1,000 (e.g. Minkowski 1964; Caswell 1970; Li et al. 1991). Leahy & Wu (1989) give a figure for the total possible number of radio observable SNRs in our Galaxy within 50,000 light years of Earth to be 485 ± 60/f₁, where f₁ is the completeness factor for SNR observations within 6,000 light years of the Sun (i.e. if we have only detected 75% of nearby SNRs, then the estimate is 486 / 0.75 or 648). Case & Bhattacharya (1996) gave 486 ± 42 as an upper limit, whereas Trushkin (1999) gives 300-1000 potentially detectable SNRs in our Galaxy.

YECs have also invoked the number of SNRs in the Large Magellanic Cloud to support their assertions. From Sarfati's article:

Not only that, but the predictions for the Milky Way's satellite galaxy, the Large Magellanic Cloud are also consistent with a young universe. Theory predicts 340 observable SNRs if the LMC were billions of years old, and 24 if it were 7000 years old. The number of actually observed SNRs in the LMC is 29.

The number of SNRs observed in the Large Magellanic Cloud in 1999 is actually 37 (Williams et al. 1999), although more are being discovered all the time - indeed it is recognised that, just like our own galaxy, there are many more SNRs yet to be discovered in the LMC (Milne et al. 1980; Dickel & Milne 1988; Chu & Kennicutt 1988). The discrepancy in Sarfati's figures can probably be explained by outdated references, and thus should not be counted against him.

However, both Davies and Sarfati make a more serious error. The estimate of 340 for the total number of SNRs in the LMC is from Mathewson & Clarke (1973). However, Clark & Caswell (1976), Clarke (1976) and Milne et al. (1980) all point out major problems with Mathewson & Clarke's estimate - basically, due to improved observations of SNRs in the LMC, Mathewson & Clarke's estimate is no longer valid. The true number of SNRs in the Large Magellanic Cloud is much, much lower.

Now, Davies has read at least one of these papers (the Clark & Caswell paper), thus he must be aware of the status of the Mathewson & Clarke estimate. Yet he uses this as one of the main supports of his theory, knowing that it is at the very least in serious dispute. When combined with the deliberate misquotation of the Clark & Caswell paper (detailed in Section 10.9), the only logical conclusion is that either Davies is seriously incompetent or he has deliberately set out to deceive (and Sarfati appears to have blindy copied from Davies' original paper, without verifying the original calculation).

10.4 The Age of Supernova Remnants

The other main plank of the YEC argument is the assertion that all SNRs are less than 10,000 years old. This can be best summed up by a section from Sarfati's article:

According to their [Astronomers'] model, the SNR should reach a diameter of about 300 light years after 120,000 years. So if our galaxy was billions of years old, we should be able to observe many SNRs this size. But if our galaxy is 6,000-10,000 years old, no SNRs would have had time to reach this size. So the number of observed SNRs of a particular size is an excellent test of whether the galaxy is old or young. In fact, the results are consistent with a universe thousands of years old, but are a puzzle if the universe has existed for billions of years.

Additionally, from the He Comes... article:

...And if you calculate, using the observed rates of expansion and the present radii, how long ago it was that the shell type supernovae explosions occurred, all the dates are under 10,000 years. Whereas, if the universe were really old, one would expect a distribution of ages, ranging from just a few years to over the million years that we calculate that the expected supernova remnants would still be strong enough to be detected via today's radio-telescopes.

This claim is widely propagated in YEC literature²⁶. However, it is completely false. While one way of measuring the age of SNRs is indeed to look at the radii and the rate of expansion, and thus calculate the age, this can only be done for younger SNRs - it is not applicable to older, more evolved SNRs, the ages of which are measured in different ways. In fact, the population of observed SNRs does show a wide distribution of ages, from young ones to really ancient ones.

For example, one of the most famous SNRs, the celebrated Veil Nebula in the constellation of Cygnus is approximately 14,000 years old (Levenson et al. 1998). G89.0 + 4.7 is 19,000 years old (Leahy & Aschenbach 1996); G6.4 - 0.1 is 58,000-110,000 years old (Kaspi et al. 1993). The remnant G69.0 + 2.7 is at least 77,000 years old (Koo et al. 1990) and G166.2 + 2.5 is 150,000 years old (Kim et al. 1988). There are many other ancient remnants (Woltjer 1972; Fich 1986; Storey et al. 1992). Duncan et al. (1995) report on G279.0 + 1.1, which they estimate could be half a million years old (it is an extremely large and faint remnant). And older SNRs are not confined to our own Galaxy. The remnant SNR 0450-709 in the Large Magellanic Cloud, which is 340 x 245 light years in size, is several hundred thousand years old (Jones et al. 1998). And with newer and improved equipment and detection techniques, astronomers are finding more and more ancient SNRs. It has even been suggested that the large-scale structure known as the Origem Loop is an ancient SNR in a very advanced stage of evolution, and which is approximately a million years old (Hanbury Brown et al. 1960; Berkhuijsen 1974; Kahn 1976).

As mentioned before, as time goes by, a SNR becomes more difficult to detect, as it increases in size and the material in the remnant gets thinner and spread out more, and distorted by the ISM - Davies completely ignores this. The values typically accepted by astronomers for the average maximum visible lifetime of a SNR range from 60,000 years to upwards of 500,000 years (Cox 1972; Jones 1975; McCray & Kafatos 1987; Leahy & Wu 1989; Dorfi 1994; de Grijs et al. 2000). From Clark (1979):

...within a few tens of thousands of years most of the extended remnants which have survived to 'middle-edge' are expected to merge with the interstellar medium and be unrecognizable.

Perhaps two of the most famous pulsars are those within the Crab Nebula and the Vela SNR (Lorimer & Ramachandran 1999). Astronomers are also attempting to relate other pulsars with various SNRs . Because the age of a pulsar can generally be computed accurately²⁷, if it can be associated with an SNR then the age of the SNR can also be calculated (Furst et al. 1993; Caraveo 1993; Gaensler & Johnson 1995b).

10.5 Third Stage Supernova Remnants

One of the most important assertions that the YECs make is that there are no third-stage, i.e. SNRs in the radiative stage Indeed, the very presence of just one third-stage SNR would completely destroy the YEC argument for a young Universe, as the amount of time a SNR takes to reach this stage is way beyond anything that the YEC time scale allows.

So, are there any actual third-stage SNRs? There have been dozens of papers published over the last several decades examining and discussing actual radiative SNRs - quite an achievement considering how, according to YECs, they don't actually exist! Despite what the YECs say, radiative SNRs do actually exist. A brief reading of the relevant literature reveals the following Galactic SNRs that are in the radiative phase (and there are others):

• G69.0 + 2.7 (Sarfi-Harb & Ogelman 1995).
• G166.2 + 2.5 (Routledge et al. 1986).
• G180.0 - 1.7 (Furst & Reich 1986).
• G189.1 + 3.0 (Oliva et al. 1999).
• G279.0 + 1.1 (Duncan et al. 1995).
• G290.1 - 0.8 (Rosado et al. 1996)²⁸.

In addition, Matthews et al. (1998) reported on an extremely faint SNR called G55.0 + 0.3:

G55.0+0.3 is amongst the faintest SNRs known. This SNR could be just one member of a larger population of faint, old remnants that are not currently detectable at radio frequencies. If a significant fraction of SNR survive longer than 50,000 years, further imaging of the Galactic plane with high surface sensitivity and high angular resolution should reveal more old SNRs.

The most massive stars (the ones most likely to end up as Type II supernovae) are found in clusters. Thus most Type II supernovae will not have been the first one to occur in the vicinity, but more likely occur in a medium that has been disturbed by the action of previous supernovae (Chu 1997). The typical lifetime of a massive star that is likely to end in a supernova (a few 10⁶ years) is not long enough for the ISM to backfill the cavity left over by previous supernovae (Jones et al. 1998). Single or multiple supernovae (in the latter case, in the same general vicinity) can result in the formation of a superbubble, up to hundreds of light years across over a time scale of one to twenty million years (Heiles 1984; McCray & Kafatos 1987; Oey & Clarke 1997; Ehlerova et al. 2001). There are many examples of these superbubbles both in our own Galaxy and in other galaxies (e.g. Blades et al. 1980; Fich 1986; Meaburn & Laspias 1991; Hunter 1994; Bomans & Chu 1997; McClure-Griffiths et al. 2000; Bond et al. 2001). Indeed, it is likely that the Sun is located in one of them (Hughes & Routledge 1972). Maciejewski et al. (1996) describe a structure they have named the "Aquila" supershell, which lies about 8,500 light years from Earth, with a radius of over 520 light years, which they calculate is about ten million years old, and the result of 10-100 supernovae. It contains several star-forming regions. Incidentally, there is one SNR associated with this structure, G34.7 - 0.4, with a calculated age of approximately 20,000 years (Wolszczan et al. 1991; Shelton et al. 1999).

In addition, Davies's assumption that the adiabatic phase of SNR evolution (i.e. the "second stage") always lasts 120,000 years and that the radiative phase always lasts 880,000 years is also completely wrong. As was mentioned in Section 5.1, the evolution of SNRs varies enormously.

10.6 The Ages of Stars

YECs such as Davies claim that the universe is about 6-7,000 years old. However, the life cycle of stars which turn into supernovae is of the order of a few tens of millions of years for high mass stars (Type II supernovae) and at least a billion years (and usually much, much more) for lower mass stars (Type I supernovae). The oldest known stars are approximately 12.5 billion years old (Cayrel et al. 2001; Qian & Wasserburg 2001), which is consistent with the latest estimate of the age of the Universe as a whole at about 13.5 to 14 billion years old (Lahav 2001; Ferreras et al. 2001).

Indeed, supernovae do play an important part in the birth of new stars - when a supernova explodes near a molecular gas cloud, the expansion of the shock front into the cloud can:

• accelerate relativistic particles
• heat and compress the molecular gas
• change its chemistry
• produce turbulent mixing.

The condensed clumps of interstellar gas created by this mechanism eventually end up as new stars (Assousa et al. 1977; Huang & Thaddeus 1986). The classical example of this happening is a group of stars embedded in a reflection nebula approximately 3,000 light years away in the constellation of Canis Major called CMa R1, in which there are two very young stars (Z CMa and HD 53367) which are the same age (about 300,000 years old) as an expanding ring of gas which appears to be an SNR (Herbst & Assousa 1977; Shevchenko et al. 1999). This scenario been disputed by some astronomers however as the identification of the ring of gas as an SNR isn't confirmed completely. A much stronger case for supernova induced star formation is the remnant G349.2 + 0.7 which is interacting with a larger shell of molecular gas, which is likely to be an extremely ancient (four million years old) SNR. This hypothesis is supported by the presence of IRAS 17147-3725, a clump of dust with similar characteristics to the proposed SNR, being ionised by an object with the spectral characteristics of a newly-formed star (Reynoso & Mangum 2001)

10.7 Distance to Supernovae and Supernova Remnants

How are the distances to supernovae and SNRs measured? Well, there are several methods available to astronomers. As the amount of energy released by a Type 1a supernova is quite accurately known, many astronomers have suggested that they can be useful in measuring distances in space, rather like a cosmic yardstick (Riess et al. 1996; Saha 1997; Riess et al. 1998; Perlmutter et al. 1998a; Regnault 2000). It has been discovered however, that not all Type Ia supernovae are identical and therefore not all have the same intrinsic brightness (Cappellaro et al. 1997; Filippenko & Riess 1999; Canal et al. 2000; Hatano et al. 2000; Howell et al. 2001; Howell 2001; Garnavich et al. 2001), and some astronomers have disputed the whole usage of Type Ia supernovae as standard candles (Drell et al. 2000), although others contend that the differences involved are not sufficient enough to rule out their use in measuring the universe (Gibson & Brook 2000; Gibson & Stetson 2001; Richtler et al. 2001).

The Sigma-D relationship (already referred to in Section 10.3), can be used to work out the true luminosity (and thus distance) of some shell-type SNRs. The emissions from the optical filaments at the shock front of a SNR can be examined to produce the true velocity at which they are moving at, and thus the distance can be worked out. The X-ray emission from a SNR in the Adiabatic phase can be measured, and from this, the actual diameter can be worked out, and thus the distance. There are other methods as well, including measuring the redshift of distant supernovae, locating known adjacent objects to which the distances are already known, and many more. For a detailed description of these methods, see Green (1984)²⁹. Even though these methods cannot give us the exact distances - indeed, we don't know the exact distances to most SNRs (indeed, Green (2000) only gives distances to a quarter of galactic SNRs), all of them tell us that apart from a few close to Earth, every supernova and SNR known is more than 7,000 light years away. In fact, astronomers have been able to directly measure the distance to SN 1987A via trigonometry. It works out at about 167,000 ± 4,000 light years (Panagia et al. 1991; Panagia 1999)³⁰. Astronomers have also have recently detected traces of supernovae billions of light years away (Perlmutter et al. 1998b, Riess et al. 1998, Riess et al. 2000). That it has taken about 167,000 years for the light from SN 1987A to reach Earth scuppers any idea of a 7,000 year old Universe.

So how do YECs respond to this? Well, it has been proposed that all the light from supposedly distant objects didn't come from those objects, but was created by Yahweh at the time of Creation in transit - in other words, those distant objects don't actually exist and are only illusions. This is an extension of the Omphalos argument (the Omphalos argument, first expounded in a book of that name by Philip Henry Gosse (1857), argues that the universe was created young but with the appearance of age). Omphalos is unfalsifiable, untestable and totally unscientific. Plus, it would relegate Yahweh to the role of a cosmic deceiver - creating objects and events that we observe (e.g. SN 1987A) which don't actually exist. Due to the obvious theological difficulties of this argument, many YECs have abandoned it (though many haven't) and proposed hypotheses based on alternate cosmologies which allow light to travel billions of light-years in a short period of time³¹, or a variable speed of light. Perhaps the chief exponent of this latter idea is the YEC Barry Setterfield, who has postulated that the rate of the speed of light is variable, and was much higher in the past (just after Creation), thus allowing for objects to appear farther away than they really are.

However, all the evidence is that the speed of light has not changed in this manner (Goldstein et al. 1973; Baum & Florentin-Nielsen 1976; Tubbs & Wolfe 1980; Gruber et al. 1981; Ellis et al. 2000)³². The most distant supernovae (and therefore the most ancient) show the same time scale of radioactive decay of the elements produced (taking into account the relativistic time dilation observed caused by the speed of the expansion of the Universe). This confirms that there has been no notable changes in decay rates between then and now, which is consistent with the idea of an ancient, expanding Universe of vast size (Leibundgut et al. 1996; Riess et al. 1997; Pranztos 1998; Filippenko & Riess 1999; Riess et al. 2000; Ellis & Sullivan 2000; Filippenko & Riess 2000; Turner & Riess 2001)³³.

10.8 Outdated References

Even though Davies' article was written in 1994, the vast majority of his references are from the 1970s with a few going back to the 1960s, with just a few from the 1980s and 1990s. Here are a few examples:

• He uses Mathewson & Clarke's 1973 estimate for the number of SNRs in the Large Magellanic Cloud - which was subsequently questioned in 1976 and then shown to be incorrect by 1980.
• He uses Tammann's 1970 estimate for the rate of supernovae occurrence (this is slightly moot, as the more recent estimates aren't all that different)
• He refers constantly to Illovaisky & Lequeux's 1972 papers, to the exclusion of many other more recent papers which have updated and corrected this paper.

This reliance on old references in itself is somewhat forgivable on its own, but a quick search on the WWW reveals hundreds of scientific papers written before 1994 (the year of publication of Davies' article), which directly contradict Davies' findings (e.g. Leahy & Wu's 1989 paper on the total number of SNRs in our Galaxy).

To make matters worse, the abridgment articles were written even more recently (Sarfati's was originally written in 1997), so there is really no excuse for others to repeat Davies' mistakes.

10.9 Misquoting and Paraphrasing

Davies misquotes several astronomers. For example, he quotes Cox (1986) as saying (in reference to a supposed lack of SNRs in the Large Magellanic Cloud):

The final example is the SNR population of the Large Magellanic Cloud. The observations have caused considerable surprise and loss of confidence.

However Cox was discussing possible models of supernova remnant evolution in the adiabatic phase, and the relevant paragraph from the original paper is actually:

The final example is the SNR population of the Large Magellanic Cloud. The observations (many collected in Mathewson et al. 1983) have caused considerable surprise and loss of confidence in simple models such as those in this paper.

Which is actually saying something completely different from what Davies claims that it says. He also misquotes Clark & Caswell (1976) twice. The first:

Why have the large number of expected remnants not been detected?

is quoted by Davies in such a way to make the reader think his estimate of the number of Galactic SNRs is correct. But in the original paper, this was a rhetorical question, in the context of discussing the 1973 estimate by Mathewson & Clarke that there should be 340 visible remnants in the Larger Magellanic Cloud. Clark & Caswell immediately follow this by giving several reasons why the 1973 estimate is unreliable (the Mathewson & Clarke estimate has been discussed earlier, in Section 10.3). The relevant paragraph from the original paper is:

Thus two anomalies require explanation. Why have the large number of expected remnants not been detected? Is it reasonable that E₀/n should differ so greatly from our estimate for the Galaxy? Both anomalies are removed if we assumed that the N(D)-D relation has been incorrectly estimated owing to the small number of remnants (4) used.

As already mentioned in Section 10.3, Clark & Caswell's suspicion was subsequently proved to be correct. But Davies totally ignores this. The second quote from this paper that Davies uses:

The mystery of the missing supernova remnants

is actually lifted from this sentence in the original paper:

It appears that with the above explanation there is no need to postulate values of E_o/n differing greatly from those in the Galaxy, and the mystery of the missing supernova remnants is also solved.

Both quotes have been lifted out of context and mean something completely different than what Davies says it does. Sarfati uses these two quotations in, what appears to be at first glance, an even more dishonest manner. He states:

As the evolutionist astronomers Clark and Caswell say: 'Why have the large number of expected remnants not been detected?' and these authors refer to 'The mystery of the missing remnants'.

Whilst one could say that this appears to be a deliberate attempt to deceive, it could be the case that Sarfati is just a bad paraphraser. However, at the very least Sarfati is guilty of incompetence, and also guilty of not checking Davies' sources. Davies however, cannot escape so lightly. The only logical conclusion from the above trail of misquotation is that Davies appears to be deliberately deceitful.

10.10 Conclusion

Let's briefly summarise how YECs such as Davies and Sarfati are correct and incorrect. First, in their favour, they are correct on:

• The frequency of supernovae occurrence in our Galaxy

However, that's actually it. This is the only point in their favour. In contrast, they are completely incorrect about the following:

• The number of actual, observable SNRs in our Galaxy.
• The typical observable lifetime of SNRs.
• The evolutionary timescales of SNRs.
• The uniformity (or lack thereof) of SNR characteristics.
• The presence of Radiative SNRs.
• The difficulty of finding SNRs.
• The distance to SNRs.

In others words, just about everything. In addition, the methodology that they use to calculate the numbers of SNRs they claim should be present is hopelessly wrong. They have also engaged in repeated misquotation, selective interpretation of the data and indeed, selective ignorance of data which disagrees with their conclusions. They are quick to assert that astronomers are confused and bewildered about the so-called "mystery of the missing supernova remnants". This however, is an assertion completely without foundation. There is no mystery. The YECs are making the mystery up themselves. They are completely wrong:

11. Notes

1. The spectra of Type I supernovae do not contain prominent hydrogen lines, whereas the spectra of Type II supernovae do.

2. A white dwarf is the tiny, extremely heavy and dense form of a star near to the end of its life. Typical density for a white dwarf is in the range of 10¹⁰ kg m^-3. Matter in a white dwarf is in the form of an electron-degenerate gas, wherein the electrons are all stripped from their parent atoms. Gas in this peculiar state is an almost perfect conductor of heat and does not obey the ordinary gas laws. Such a white dwarf no longer has any source of energy and simply cools down forever, eventually becoming a black dwarf - a cold, dead lump of matter hanging in space.

3. Named after the Indian-born astrophysicist Subrahmanyan Chandrasekhar, who first calculated it in 1930 as a means of passing the time whilst on a boring 18-day boat trip from India to England! It is the maximum possible stable mass for a white dwarf star. It is equal to 1.44 solar masses - the mass of the Sun is approximately 1.9891 x 10³⁰ kg.

4. To illustrate the vast amounts of energy output by supernovae, our Sun has an absolute (how it would appear if it was 10 parsecs, or 32.616 light years away) magnitude of only +4.7 - it would thus appear to the naked eye as a faint star, barely visible to the naked eye. The apparent magnitude of the Sun is -26.8. Even so, excluding neutrinos, optical radiation only accounts for 1% of the total energy output (van den Bergh 1988).

5. The normal fuel of stars is hydrogen. Over the lifetime of the star, this hydrogen is gradually converted (via thermonuclear fusion reactions) into helium by a process called nucleosynthesis. In nucleosynthesis, light atomic nuclei (such as hydrogen) collide with such violence and frequency in the high-temperature and high-density interior of the star that they fuse into heavier nuclei (such as helium) and release vast amounts of energy (like in an H-Bomb). In effect the lighter elements "burn" to form heavier elements.

6. According to Arnett et al. (1989), for a star of 20 Solar Masses, it takes approximately ten million years to complete the hydrogen burning state. Helium burning requires approximately a tenth of this, 950,000 years. Carbon burning takes 300 years, and neon and oxygen burning take 180 and 140 days respectively. Silicon burning is completed in two days. At this time, the temperature in the core is some 3.7 x 10⁹ K.

7. Photodisintegration is the stripping down of nuclei by photons into individual protons and neutrons. This process is highly endothermic (i.e. it requires more energy than it generates). It was first identified by Willy Fowler and Sir Fred Hoyle in the 1960s. Photodisintegration can also occur in the silicon-burning phase.

8. For every 0.1 solar masses of iron that is broken down via photodisintegration into protons and neutrons, the shock loses 1.7 x 10⁵¹ ergs.

9. A neutron star is a star made almost entirely of neutrons, with a density of an atomic nucleus (Horowitz & Piekarewicz 2001) . Such a star contains typically the same amount of matter as there is in our Sun, but packed into a sphere about 10 km across. The maximum mass for a neutron star is approximately three solar masses, the so-called Oppenheimer-Volkoff limit (first postulated in 1939 by Robert Oppenheimer, of A-Bomb fame, and his student, George Volkoff), and the minimum mass about 0.1 solar masses (any lighter neutron star that tried to form would turn into a small white dwarf, as some of the neutrons converted themselves into protons by a process called beta decay). The density of matter in a neutron star is much greater than in a white dwarf - about 10¹⁷ kg m^-3.

10. A pulsar is a rotating neutron star, with a mass similar to the Sun's but a diameter of only about 10 kilometres. The pulses occur because the neutron star is rotating very rapidly: a beam of radio emission produced by synchrotron emission from electrons moving in the very strong magnetic field (about 10⁸ tesla, or a billion times the strength of the magnetic field at the surface of the Earth) of the rotating neutron star sweeps past an observer once per rotation. The pulses are very regular, apart from the occasional glitch, and all single pulsars are gradually slowing down as they lose rotational energy (van der Swaluw & Wu 2001). The time between successive pulses ranges from 1.558 milliseconds for the fastest known pulsar, PSR 1937 + 21 (Xu et al. 2001), to 8.5 seconds for the slowest observed pulsar (Young et al. 1999). The first pulsar was discovered in 1967 by Jocelyn Bell Burnell at Cambridge, England. There are over 1300 pulsars known (Gotthelf et al. 2000; Lorimer 2001), although many more are being discovered all the time (D'Amico et al. 2000; Edwards et al. 2001; McLaughlin et al. 2001) - the most famous being the one at the centre of the Crab Nebular, which has a spin period of 50 milliseconds (Wang et al. 2001). Although it has long been thought that pulsars are the most common form of young neutron star, there have been recent discoveries of other classes of objects (Gaensler et al. 2001). These include: Magnetars - young isolated neutron stars with extremely high magnetic fields (Duncan & Thompson 1992; Paczynski 1992), Soft Gamma-ray Repeaters (SGRs), pulsating X-ray sources with occasional intense Gamma-ray activity but no detectable radio pulsations (Hurley 1999) and Anomalous X-ray pulsars, pulsating X-ray sources which are spinning down slowly (Mereghetti 1999). It may be the case that both SGRs and AXPs are types of magnetars (Thompson & Duncan 1996; Frail et al. 1997; Vasisht & Gotthelf 1997; Kouveliotou et al. 1998) or other unusual pulsars (Marsden et al. 2001b) maybe even some other exotic type of star (Xu et al. 2000). Most SGRs/AXPs appear to be physically associated with supernova remnants (Gaensler et al. 2001).

11. Named after the German astronomer and physicist Karl Schwarzschild, who investigated the concept in the early 20th century. It is the radius below which the gravitational attraction between the particles of a body must cause it to undergo irreversible gravitational collapse. It is equal to 2.95 x (Mass_Body/Mass_Sun) kilometres.

12. Neutrinos are elementary particles with no electric charge and almost no mass, and they interact only very weakly with other matter. Because they hardly interact with matter at all, neutrinos are very difficult to detect. In one type of neutrino detector shown to work successfully, detectors in a large tank of water (located as far as possible underground to block off cosmic rays which interfere with the detection process) pick up Cerenkov radiation generated by the interaction of electrons with solar neutrinos. Detectors of this type made the first observation of neutrinos from a supernova - those from SN 1987A - at 7.36 GMT on the 23rd February 1987 (before the optical light reached Earth). The Kamiokande II detector in Japan recorded 9 neutrinos within 2 seconds, followed by 3 more within 13 seconds, the IMB detector in Ohio in the US detected 8 neutrinos within 6 seconds and the Baksan detector in the then Soviet Union recorded the arrival of 5 neutrinos within 5 seconds.

13. New Scientist Magazine of the 18th September 1999 reported on a supernova that apparently occurred in 1320, but went strangely unobserved at the time. The ROSAT X-ray satellite imaged a supernova remnant in the constellation of Vela, only 640 light years distant, and scientists have observed a spike in the concentration of nitrates in Antarctic Ice Cores corresponding to the year 1320 - similar spikes were observed in 1572 and 1604, when known supernovae occurred (Aschenbach 1998; Aschenbach et al. 1999; Robinson 1999; Burgess & Zuber 2000). Recently however, doubts have been expressed about the recent dating of this SNR, and it has been proposed that the SNR is actually thousands of years older (