Huchra (1977), Larson and Tinsley (1978) and Kennicutt et al. (1987) have successfully used models to show that enhancements in the (B-V) colours and H-alpha equivalent widths of close binary and interacting spirals can be explained by short duration (~ few x 10⁷years) bursts involving less than a few percent of the total stellar mass (~ 1 x 10⁹ M_O of gas).

Kennicutt et al. (1987) noted that there is an absence of blue spirals with weak H-alpha equivalent widths. Galaxies with these properties should be seen, however, since the UBV colours of spirals should decay much more slowly than their H-alpha emissions once a burst is finished. Kennicutt et al. (1987) gave two possible explanations for this discrepancy in the observations. Either the stars produced in the burst have a much shallower initial mass function (IMF) compared with the “normal” underlying steady-state star formation or the star formation bursts decay gradually over a period of ~ 10 ⁸ years, following a rapid initial rise.

While the studies of Huchra (1977), Larson and Tinsley (1978) and Kennicutt et al. (1987) show that galaxy-galaxy interactions are one probable cause for bursts in spirals, there is a significant population of non-interacting spirals that also have enhanced (B-V) colours and H-alpha equivalent widths (Kennicutt et al. 1987). These galaxies may be undergoing bursts as well.

If spiral galaxies experience short duration bursts ( ~ 0.5 – 1.0 x 10⁸ years) in their global star-formation rates, their SN II rates would be enhanced for up to ~ 0.85 - 1.35 x 10⁸ years (Mihalas and Binney, 1981) following the onset of the burst [note: these time intervals correspond to the maximum burst duration plus the lifetime of an 8 M_O star - 3.5 x 10⁷ years (Mihalas and Binney, 1981)].

In contrast, the SN Ia rates for these burst galaxies would not become enhanced until:

a) the 8 - 5 M_O stars formed in the burst, evolved into a degenerate carbon/oxygen white dwarfs.

The limiting factor that effectively determines the time scale for SNIa enhancement is the requirement that the white dwarf accrete enough mass so that it eventually reach the Chandrasekhar limit.

Models show that if the accretion rate from the companion is too low, the white dwarf undergoes nova-like eruptions, loosing more mass in the outburst than it gains in the prior accretion. Alternatively, if the accretion rate is too high, a HI rich envelope would form around the white dwarf, ruling it out as a possible progenitor for a SNIa event. Only at moderate high accretion rates (~ few x 10^-7 M_O /year) does the white dwarf accumulate a thin hydrostatic layer of hydrogen upon its surface. This layer continues to build up until stable hydrogen burning is initiated at its base, leading to the steady deposition of helium ash (Hillebrandt and Niemeyer 2000). The problem is how do you achieve this critical accretion rate?

Conventional binary evolutionary models show that when the mass ratio of the donor star to the white dwarf exceeds a critical value, the mass transfer becomes unstable, and a common envelope forms. However, the models of Hachisu, Kato and Nomoto (1996) show that this phenomenon can be avoided if the mass transfer rate exceeds a certain value. When the accretion rates exceed this particular value, the white dwarf begins to emit a strong stellar wind which stabilizes the mass transfer. The wind limits the accretion rate in such a way that the wind loss rates and the accretion rates are almost equal.

However, when evolutionary calculations are done for white dwarf binaries, in which mass transfer occurs through Roche-lobe overflow, and a strong stellar wind is used to stabilize the accretion flow, SNIa progenitors are only produced when the white dwarf’s companion is a main sequence or subgiant star with a mass less than ~ 3.5 M _O (Li and van den Heuvel 1997).

A 3.5 M_O star takes approximately 1.80 x 10 ⁸ years to evolve off the main sequence (Mihalas and Binney 1981). Hence, this effectively ensures that there would be no enhancement in the SNIa rate of the burst spirals until well after the end of the epoch of SNII enhancement i.e. ~ 0.85 - 1.35 x 10⁸ years after the onset of the burst. (Note: Even if the allowable companion mass was increased to 4.5 M _O , it would still take ~ 1.1 x 10⁸ years for a star of this mass to evolve off the main sequence. Hence, it would only marginally shorten the period over which SNII but not SNIa are enhanced .)

Consequently, the spiral galaxies experiencing a burst in their star formation rate, should go through a distinct epoch in which they produce SN II at a much higher rate than they produce SN Ia. This epoch would span the first 0.85 - 1.35 x 10⁸ years following the onset of a burst.

If short duration bursts are a universal phenomenon in spirals, then the fraction of all spirals that are in the burst state (f) would be given by:

                               f = ( t_SNII xN)/ t_{gal                                                                                 (1)}

where t_SNII is the duration of the epoch of enhanced SN II production, t_gal is the typical lifetime of a spiral galaxy and N is the number ofburst during this lifetime. Hence,

                    f = t_SNII /( t_gal / N )= duration of the epoch of enhanced SN II production               (2)
                                                                    length of time between bursts

Table 1 shows the fraction of all spiral galaxies that are predicted to be in the burst state, assuming t_gal = 1.0 x 10¹⁰ years. We can see from this table that even if bursts only occur once every 2 x 10⁹ years, we could expect 5.0 % of all spirals to be in a burst state, provided the burst duration is ~ 1.0 x 10⁸ years.

                                                   TABLE 1

Hence, these simple calculations raise the possibility that we might be able to identify a distinct sub-class of spirals that are preferentially producing SN II but not SN Ia (hereafter, referred to as Burst spirals).

Burst spirals can be identified because of their high current to integrated star formation rates. The best indicators for these rates are the H- alpha equivalent width or the ratio of the 40 - 120 mm infrared luminosity to R-band luminosity (L _IR /L_R)(Young et al. 1996). Unfortunately, while these two indicators are useful distance-independent and luminosity-independent indicators of relative star-formation activity, they are only available for a limited number of spirals galaxies (Young et. al. 1996).

A much more widely available indicator is the ratio of the integrated 40 - 120 mm infrared luminosity to the B-band luminosity
(L/L_B)(de Vaucouleurs et al. 1991). However, the use of the B-band rather than the R-band luminosity, means that this alternate indicator is no longer a direct measure of the current to integrated star formation rate, since young stars can contribute to a galaxy’s B-band fluxes. Nevertheless, the (L_IR/L _B) ratio can still be used to identify burst spirals since these galaxies will have significantly higher values for this ratio compared to spirals that are not undergoing bursts.

A second alternative indicator of the current to integrated star formation rate for spiral galaxies is their infrared surface brightness (SB_IR ) (de Vaucouleurs et al.1991). Persson and Helou (1987) find a relatively strong correlation between the infrared and H-alpha surface brightness of spiral galaxies. The simplest explanation for this correlation is that L_IR is a good indicator of the high mass star formation rates of spirals galaxies and that the SB_IR of spirals scale with their current to integrated star formation rates (Young et al. 1989 and 1996).

Persson and Helou (1987) claim that this explanation is too simplistic because of the likelihood that infra-red cirrus emission from the diffuse neutral medium of galaxies will contribute significantly to their overall L _IR . This is particularly true for galaxies with low H-alpha EQW or low values of SB_IR were the contribution of the infrared cirrus emission component to L_IR may become comparable to that contributed by warm dust in the galaxy’s HII regions. Fortunately, this should not be a major problem in our study since we are primarily interested inidentifying burst spirals that have high current to integrated star formation rates.

Thus, our aim is to see if short duration bursts are a relatively common phenomenon in spiral galaxies. If this is the case, then we should find that spiral galaxies with high (L_IR/L_B) and (SB _IR ) are preferentially producing SN II but not SN Ia.

Parts II outlines some of the limitations in using (L _IR/L_B ) and (SB_IR) as indicators of the current to integrated star formation rate. Part III shows that (L _IR/L_B) and (SB_IR) can be used to determine the (relative) current to integrated star formation rates of Sbc-Sdm spirals. Part IV lists our galaxy sample that includes all of the typed SN between supernovae 1885A and 2000BS. Part V shows that there is a population of Sbc-Sdm spirals with high (L_IR/L_B) and (SB_IR) that are preferentially producing SN II but not SN Ia. Part VI is the analysis of our results and part VII summarizes our conclusions.