The public health implications of T Pyxidis

No, not an unusual pathogen, but a supernova.

This is a belated response to the news that a white dwarf in a binary star system named T Pyxidis is approaching its Chandrasekhar Limit- in other words, it’s close to going ka-boom- and, according to a paper presented at this year’s American Astronomical Society meeting, at 1000 parsecs, it’s closer to us than originally thought.

If we assume that the supernova of a nearby star of would “destroy all life on Earth”, this makes for an attention-grabbing story. Cue headlines like this. The end of the world angle gets talked up, before we get the reassuring news that this is unlikely to happen for 10 million years or so.

Could this really be true? Leaving the unpredictable behaviour and location of T Pyxidis to one side, I spent a few minutes link-and-google sleuthing to try to get to the bottom of the supernova-destroying-life-on-Earth bit. Writing it up will no doubt take longer. My rigorously scientific search strategy led me to the following conclusion: while it’s all very uncertain and (of course) low probability, high impact stuff, it’s perhaps a little more interesting than the headlines suggest.

It doesn’t seem like many people have tried to quantify the public and environmental health risks of supernova explosions, but here’s someone who’s had a go. I’ll summarise and extrapolate as best I can.

The type and proximity of the supernova determines the risk. Type 1a supernovae produce the highest energy radiation, and can therefore be considered the most dangerous. The bad news is that T Pyxidis would be a Type 1a. The good news is that (surprise surprise) the dangers have been hyped. X-rays and gamma rays are identified as the main hazards to life on Earth, with a Type 1a supernova needing to be within 60-200 parsecs- much closer than T Pyxidis- for the X-ray emissions to equal those of a solar flare. Given that solar flares generally pose little risk to life on Earth- the magnetosphere and atmosphere do a good job of shielding all but the highest energy flares- it follows that a supernova would have to be much closer even than this for X-rays to pose a real risk to human/biosphere health, although unshielded satellites and astronauts might be in trouble.

So that leaves gamma rays. The danger here is that a significant gamma ray flux can break apart nitrogen (N2) molecules in the upper atmosphere, forming reactive nitrous oxides which damage the ozone layer. After humanity’s recent brush with ozone disaster, there’s a reasonable understanding of the potential dangers of a diminished or absent ozone layer, caused by unfiltered ultraviolet radiation from the Sun damaging plants and animals. More on this in a minute. First, though- could a T Pyxidis Type 1a supernova, at the stated 1000 parsec distance, really completely destroy the ozone layer? The short answer is no. In this regard, the stories linked above don’t correspond with the relevant sections in Michael Richmond’s piece, and I’ve just stumbled on the reason why: the author of the AAS Pyxidis paper mixed up his figures, and a misleading statement found its way into the paper’s press release. In fact, a Type 1a supernova would need to be ten times closer than T Pyxidis for it to strip away the ozone layer, and it would take an even rarer and more unlikely high-energy gamma ray burst from a hypernova to manage the feat at that distance.

Even if the ozone layer were stripped, would all life on Earth be utterly destroyed? This seems unlikely. After all, life first evolved on Earth before the ozone layer formed. Nonetheless, unfiltered ultraviolet light would cause significant damage to humans, animals and plants. In the initial stages of such an event, humans would be better off than our planetary co-dwellers. Awareness of the danger allows behavioural changes, such as staying indoors, wearing protective clothing, very high-strength sunblock and dark glasses. This might limit the direct health impacts of high levels of UVR such as sunburn, skin cancer, immunosuppression, inflammatory eye disease and cataracts- at least for the wealthier sections of the global population. Of more concern would be the effects of high ultraviolet levels on the rest of the biosphere, with inevitable consequences for humans. Widespread UV-induced destruction of plant life on land and at sea (phytoplankton in the surface layers of the ocean would die too)- would lead to global collapse of food chains, with those consumers at the top- including us humans- ultimately suffering. With world stockpiles of food currently good for only a couple of months or so, mass starvation would follow.

While the gamma ray emissions from a supernova would likely only last several weeks, the nitrous oxides would persist in the upper atmosphere for longer, with full recovery of the ozone layer taking 10 years or more. Nonetheless, deep ocean and buried microbial life would survive, detritovores would have plenty to eat, seeds would persist to sprout when conditions improve, and perhaps a small breeding population of humans (and other consumers) might be preserved until such time as plant growth recovers.

This is the worst case scenario. Michael Richmond’s paper references a figure of ‘considerably more than 100,000 erg/cm^2’ for the power of gamma ray flux needed to strip away the ozone layer, while calculating that a Type 1a supernova at 1000 parsecs- T Pyxidis distance- would produce a flux of less than half that, about 40,000 erg/cm^2.

So what seems more likely than the worst case scenario is a situation where the ozone layer might be damaged, but not completely destroyed [see footnote below]. This means that there’s everything to play for. Crop yields might decline, but not necessarily collapse, and regional differences in pre-existing ozone layer thickness and cloud cover may become vitally important in deciding where to grow food. With a good chance of avoiding apocalypse, managing the health effects mentioned above becomes important; certainly a public health crisis, but one amenable to good planning and co-ordinated, global measures to reduce UV exposure and secure food supplies. With a favourable initial response, more subtle health impacts may even become apparent, such as increased risk of infectious disease from immune system suppression, or in some cases Vitamin D deficiency from successful sun avoidance measures.

While solar flare-like effects of supernova radiation on satellite communications and power grids might impede human efforts to manage the situation, there’s also the hope that at sub-apocalyptic gamma ray levels, the Earth system itself might be able to mitigate some of the worst impacts through feedback mechanisms, such as UV-stressed phytoplankton seeding cloud formation with DMS, and plants increasing their nuclear DNA levels to resist high UV levels.

That’s probably enough idle speculation for now.

One reaction to these kind of scenarios is just to ignore them- “it’s almost certainly never going to happen, and if it does we’ll all die a horrible death anyway, so why waste time thinking about it?”
While the probability might be small compared to everyday risks, this example shows how poorly constrained the uncertainty can be. While there’s nothing to lose any sleep over, we can’t be certain that the risk is infinitesimal- in fact, there’s some evidence that this sort of thing has happened before (admittedly 450 million years ago). Although the worst-case outcomes are very unlikely- and probably not amenable to our efforts to alter events- what’s more to the point is that there is a range of less-disastrous but potentially-mitigatable possibilities. Space hazards like asteroids, comets, supernovae, supernovae-derived comets, or gamma ray bursts are certainly at the extreme end of the scale, but perhaps the same principle applies to more down-to earth dangers– such as pandemics, global warming, and earthquakes.

[Note- if not already clear- the supernova would have to be much closer than T Pyxidis. And this paragraph sort of assumes that the relationships between gamma ray flux, ozone damage and UV radiation are linear-ish in the ranges we’re talking about, which I don’t know. Anybody more knowledgeable/numerate- please feel free to comment]


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: