Introduction
Gamma-ray bursts (GRBs) have been recognized as the most powerful explosions in the universe (see, e.g., Meszaros Reference Meszaros2001; Piran Reference Piran2005; Bloom et al. Reference Bloom2009). There appear to be two classes of GRB, based on the duration of the event, which is also correlated with spectral hardness – short-duration bursts tend to have harder spectra but less overall energy (Kouveliotou et al. Reference Kouveliotou, Meegan and Fishman1993; Zhang & Choi Reference Zhang and Choi2008). There may also be different populations within these classes, based on luminosity (Chapman et al. Reference Chapman, Tanvir, Priddey and Levan2007; Berger et al. Reference Berger2007; Cenko et al. Reference Cenko2008; Chapman et al. Reference Chapman, Priddey and Tanvir2008; Virgili et al. Reference Virgili, Liang and Zhang2009). While still under study, long bursts are associated with core-collapse supernovae (Meszaros Reference Meszaros2001; Piran Reference Piran2005; Campana et al. Reference Campana2008), while short bursts may be the result of mergers between compact objects such as neutron stars and black holes (O'Brien & Willingale Reference O'Brien and Willingale2007; Levan Reference Levan2008).
Starting in the 1990s it was recognized that GRBs, like supernovae and other astrophysical sources of ionizing radiation, could pose a threat to life on Earth over long time scales (Thorsett Reference Thorsett1995; Scalo & Wheeler Reference Scalo and Wheeler2002; Melott et al. Reference Melott, Lieberman, Laird, Martin, Medvedev, Thomas, Cannizzo, Gehrels and Jackman2004). The radiation from a GRB is highly beamed, and therefore the Earth must fall within that beam in order to be impacted. It has been estimated that the nearest likely long-duration burst pointed at the Earth in the last billion years would be of the order of 1 kpc distant (Melott et al. Reference Melott, Lieberman, Laird, Martin, Medvedev, Thomas, Cannizzo, Gehrels and Jackman2004; Dermer & Holmes Reference Dermer and Holmes2005; Thomas & Melott Reference Thomas and Melott2006).
The author and collaborators (see http://kusmos.phsx.ku.edu/~melott/Astrobiology.htm) have performed extensive simulations of the likely impact on the Earth's atmosphere and biosphere by a ‘typical’ long-duration GRB. Here we will briefly review this work, describe some recent efforts and discuss some outstanding issues in more accurately quantifying the impact of GRBs on life on Earth. Full details of past work can be found in Thomas et al. (Reference Thomas2005a, Reference Thomasb), Melott et al. (Reference Melott, Thomas, Hogan, Ejzak and Jackman2005), Thomas & Melott (Reference Thomas and Melott2006), Thomas & Honeyman (Reference Thomas and Honeyman2008) and Melott & Thomas (Reference Melott and Thomas2009). In addition, we have explored a wider range of event duration and spectral parameters in order to draw more general conclusions about the impact of astrophysical sources of ionizing radiation (Ejzak et al. Reference Ejzak, Melott, Medvedev and Thomas2007).
Modeling parameters
Our simulations have been performed using the Goddard Space Flight Center two-dimensional atmospheric chemistry and dynamics model, developed by Charles Jackman and others. This code has been used extensively to study ozone changes due to a variety of effects, including supernovae (Geherls et al. Reference Gehrels, Laird, Jackman, Cannizzo, Mattson and Chen2003). The model has 18 latitude bands and 58 altitude bands in log pressure. Further details of the model may be found in Thomas et al. (Reference Thomas2005b).
The effects discussed below will primarily focus on results obtained for a ‘standard’ long GRB with power 5×1044 W, with a duration of 10 s, at a distance of 2 kpc, delivering a fluence of 100 kJ m−2. We have used the Band spectrum (Band et al. Reference Band1993), with E 0=187.5 keV, to compute the atmospheric ionization, which is then used as a source of nitrogen oxides in the atmospheric model. In Ejzak et al. (Reference Ejzak, Melott, Medvedev and Thomas2007) we have investigated a broader range of event duration and spectra. These results will be discussed in a general way below.
Terrestrial effects of a GRB
A GRB within a few parsecs that is directed at the Earth will impact one hemisphere of the planet with a short, but intense blast of high-energy photons. Gamma-rays and X-rays are highly attenuated by the Earth's atmosphere. Therefore, the ground-level effects are primarily indirect. A small fraction of the incident energy reaches the ground as dangerous ultraviolet (UV) radiation (Smith et al. Reference Smith, Scalo and Wheeler2004), but this is limited in time to the duration of the event, which is at most tens of seconds for a long burst, and is less than a second for a short burst. While it is possible that this flash would affect some organisms, it seems unlikely that a biological catastrophe would result from this effect alone. Of course, for planets with thinner atmospheres the energy deposited at the ground would be greater and more serious effects may be expected (Smith et al. Reference Smith, Scalo and Wheeler2004; Galante & Horvath Reference Galante and Horvath2007). We are concerned here with effects on life on Earth and so will concentrate on the longer-term impacts.
There are three potentially harmful long-term effects of a GRB that follow from changes in atmospheric chemistry (Reid & McAfee Reference Reid and McAfee1978). High-energy photons cause dissociation, ionization and ionizing dissociations of N2 and O2 in the atmosphere. Subsequent reactions lead to the formation of nitrogen oxides, most importantly NO and NO2. These compounds catalytically deplete ozone (O3) in the stratosphere, leading to increases in surface-level solar UV over long time periods (years). Secondly, NO2 itself is a brown gas that absorbs strongly in the visible. This may potentially have a climatic effect by reducing solar insolation at the ground, thereby leading to cooling. Third, the atmosphere returns to normal via the removal of nitrogen oxides by way of precipitation of nitric acid (HNO3).
While nitric acid rain can have a negative effect on a variety of organisms (including amphibians), it appears that the amount deposited following a GRB would be small enough so as to not have a serious impact (Thomas & Honeyman Reference Thomas and Honeyman2008). There is a possibility that the deposited nitrate may actually benefit some organisms, particularly land plants that may be nitrogen starved.
It is currently unclear whether cooling due to NO2 opacity is likely to be important. Our results indicate a global average reduction of about 1% in solar visible light fluence, lasting a few years (Melott et al. Reference Melott, Thomas, Hogan, Ejzak and Jackman2005). Larger reductions would occur at the poles, where compounds produced by the burst tend to accumulate (due to poleward transport in the stratosphere). It is possible that this reduction in sunlight could initiate a global cooling event, particularly if the climate is near a ‘tipping point’. To date no simulations of the climate effects have been performed, but our data is available for such work.
The primary effect, then, is the enhanced solar UV at the surface that would follow destruction of stratospheric ozone following a GRB. Globally averaged ozone column density reductions of 25–35% are possible, depending on where the burst occurs in latitude and at what time of year. Reductions of up to 75% at a given location and time are seen for some event cases. The atmosphere recovers on the time scale of years. For our standard fluence, full recovery takes just over a decade. Our results may be compared to values associated with recent anthropogenic reductions in O3. Globally, reductions of 3–5% have been seen, with short-lived (weeks) reductions of up to 70% under the so-called ozone hole.
As reported in Thomas et al. (Reference Thomas2005b) and Ejzak et al. (Reference Ejzak, Melott, Medvedev and Thomas2007), O3 depletion scales roughly as the cube root of the fluence of the event. Additionally, the overall depletion is much more sensitive to the hardness of the spectrum and the total energy deposited than to event duration. For a given fluence and spectrum, the maximum globally averaged ozone depletion is roughly constant for a wide range of duration, from fractions of a second to years, although when that maximum occurs varies.
Biological effects following O3 depletion
A variety of effects on organisms may be expected due to the enhanced surface level solar UV following severe depletion of ozone. The UVB (290–315 nm) part of the spectrum is most affected by O3 levels and is known to damage DNA molecules directly but can also have effects on other biomolecules as well (Vincent & Neale Reference Neale and de Mora2000). UVB effects appear at the organism level in ways ranging from lowered metabolic rates and photosynthetic capacity, low growth rate, delayed or arrested cell division and/or death. In multicellular organisms, the effects may cause developmental delay and abnormalities, altered tissue composition and cancer. Biological weighting functions (BWFs) specify wavelength dependence of effectiveness (Coohill Reference Coohill1991; Neale Reference Neale and de Mora2000), and they are used to weight spectral irradiance to obtain total effective irradiance.
In our work to date, we have used the Setlow DNA damage weighting function (Setlow Reference Setlow1974; Smith et al. Reference Smith, Baker, Holm-Hansen and Olson1980) to quantify the biological impact of enhanced UVB irradiance following a GRB. We compute the surface-level UVB using the solar spectrum, O3 column densities from the modelling results and the Beer–Lambert relation, which assumes that absorption is more important than scattering.
Using this approach, we find DNA damage of up to 16 times the annual global average of pre-burst values. This level of damage lasts a few months in some areas, particularly at mid-latitudes during the summer. Those locations may experience five to seven times the normal damage level for a period of several years.
Current and future work
It has been proposed that a GRB may have initiated the late Ordovician mass extinction (Melott et al. Reference Melott, Lieberman, Laird, Martin, Medvedev, Thomas, Cannizzo, Gehrels and Jackman2004). This proposal was based largely on the water-depth dependence of this extinction, which fits well with a radiation event, since UV is attenuated rather strongly in water and hence would affect organisms which dwell at the top of the water column more strongly than those that live towards the bottom. We have recently completed a study investigating the latitude dependence of our predicted DNA damage values compared to observed latitude dependence of extinction from this period (Melott & Thomas Reference Melott and Thomas2009). We find that only a south polar burst can fit the observed trend in extinction with latitude. This allows us to predict that (yet to be measured) extinction rates at northern latitudes should be much smaller, since the effect of a polar burst is almost completely contained within that hemisphere.
Obtaining better estimates of the biological damage following a GRB or other ionizing radiation event is an area of on-going effort. While the Setlow function is fairly generic and a good representation of damage to bare DNA, there are many factors that make the actual effect on organisms and ecosystems much more complicated. Organisms may have shielding, repair mechanisms, avoidance strategies, etc. In addition, extrapolation from DNA damage to ecosystem-level effects can be problematic due to the complex inter-relationships involved. We are currently working to improve estimates of damage at the organism level and then to connect that to larger scale effects. These efforts include: performing more accurate calculations of solar UV irradiance in the ocean water column under severely depleted ozone; applying a wider variety of weighting functions; and attempting to model broader ecological effects. We are currently focusing our efforts primarily on phytoplankton in the ocean basins. The ocean accounts for about half of global primary productivity, and, accordingly, half of CO2 fixation and O2 evolution (Behrenfeld et al. Reference Behrenfeld, Boss, Siegel and Shea2005). Thus, even modest effects on oceanic productivity could have major repercussions throughout the whole marine food chain and, through atmospheric effects, the global climate.
Along with other sources of ionizing radiation (such as supernovae and soft gamma repeaters), gamma-ray bursts present a real threat to life on Earth and other terrestrial planets in the long term. Such events have likely played a role in the development of life on Earth, and a better understanding of their impacts and rates will be an important component of improving our view of the past, and future, of life on Earth. In a broader context, this understanding can also help to constrain concepts such as the Galactic Habitable Zone and potentially guide efforts such as the Search for Extraterrestrial Intelligence (SETI). More detailed effort is now needed to improve estimates of the rates of GRBs and other events, and to more precisely determine the ecosystem-level effects from a given event.
Acknowledgements
The author's travel costs to attend the ESLAB symposium were funded by an International Travel Grant from Washburn University, as well as the College of Arts and Sciences travel fund. Much of the work reported here was supported by NASA Astrobiology grant NNG04GM41G.
