Apparently the main threat from an EMP is not from the radiation from the actual detonation, but from electrons and positrons emitted by the explosion which then cause secondary wideband radiation. At low altitude the atmosphere will stop these particles within a few tens of miles, and therefore the effect is relatively short-ranged. However at stratospheric altitudes the combination of thinner air and the strength of the earth's magnetic field can carry the particles over an extremely wide area, allowing for heavy EMP over a wide geographical area. Bombs can also be optimized to output more beta/electron radiation to worsen the EMP effect.
The shape of the magnetic field over North America makes EMP especially threatening, in that a powerful EMP-optimized bomb detonated over the Dakotas could knock out the electrical grid and computing resources of almost the entire central US. The coasts would probably be mostly ok, however.
The real danger is that you aren't always able to tell if an attack is going to be an EMP or nuclear.
When you detect the missiles flying towards you, there's no way of knowing that the payload is non-nuclear.
And you may decide to fire nukes back before they will be destroyed, rather than wait and see.
All adversaries with ICBM's know this and are not stupid enough to take the risk.
There is no mechanism by which nuclear fission in a U or Pu bomb would produce electrons and positrons (which strictly speaking are antimatter) as its main products, but rather neutrons and gamma rays - well a lot of EM radiation at different frequencies.
A reasonably lucid explanation of EMP can be found e.g. here: http://www.thespacereview.com/article/1549/1
There is no mechanism by which nuclear fission in a U or Pu bomb would produce electrons and positrons (which strictly speaking are antimatter) as its main products, but rather neutrons and gamma rays - well a lot of EM radiation at different frequencies.
I wouldn't say none. I don't think fission directly produces anything (i.e., within femtoseconds; maybe a rare pair-production from sheer quantum vacuum voodoo?), but the products should make plenty of nasties. Sr90 is a good example of beta-bearing waste; much more will be made in odd nuclear ratios (neutron rich/heavy) which result in relatively quick (microseconds to seconds) weak decays, resulting in both types of 'trons.
EMP effects are unlikely to cause permanant damage to modern cars, which are fairly well protected (due to the electrically awful nature of car electrics they need to be), are enclosed in a metal cage, and are electrically short. Tests where cars have been struck by lighting typically shows no effect or them stalling and needing to be restarted. Despite claims to the contary older Kettering ignition systems will be fine too unless you happen to have a mile of wire trailing out of your car. Of course thousands of cars stopping on a busy road can still be damaging - even if they can be restarted it doesn't stop people driving into each other.
Most aircraft controls should be fine too, but navigation and communication could easily fail so I'd expect some collisions when attempting to land, particularly in poor visibility.
The things that really get hit hard are those with long cables (power grids, telephone lines) or which are sensative by design such as radio receivers. Expect to loose the power grid for some weeks or months. And whilst backup generators are likely to start ok there are likely to be enormous problems with fuel distribution, and also with damage to switchgear on large installations such as hospitals. The main questions I would ask to decide if the 500,000 is sensationalist are:
1) How many deaths could be expected from car crashes resulting from unexpected engine failure and
2) How many Americans (in hospitals or otherwise) rely on electricity for life support in a way that 30 minutes matter.
Of course this is only relevent if the EMP is caused by a solar flare, if it was nuclear then MAD dictates many more deaths from the resulting war.
...her comment was: "Don't Hospitals have backup generators?"
EMP effects are unlikely to cause permanant damage to modern cars, which are fairly well protected (due to the electrically awful nature of car electrics they need to be), are enclosed in a metal cage, and are electrically short. Tests where cars have been struck by lighting typically shows no effect or them stalling and needing to be restarted. Despite claims to the contary older Kettering ignition systems will be fine too unless you happen to have a mile of wire trailing out of your car.
---snip---
A high altitude nuclear explosion (that creates HEMP) produces three major energy components that arrive in sequence, and which have measurably different effects that can be cumulatively damaging to electronic equipment. The first energy component is the initial energy shockwave which lasts about one microsecond, and is similar to extremely intense static electricity that can overload circuitry for every electronic device that is within line of sight of the burst. A secondary energy component then arrives, which has characteristics that are similar to a lightning strike. By itself, this second energy component might not be an issue for some critical infrastructure equipment, if anti-lightning protective measures are already in place. However, the rise time of the first component is so rapid and intense that it can destroy many protective measures, allowing the second component to further disrupt the electronic equipment. The third energy component is a longer-lasting magnetic signal, from about one microsecond to one full second in duration. This geomagnetic signal causes an effect that is damaging primarily to long-lines electronic equipment. A localized magnetic effect builds up throughout the length of the transmission lines and then quickly collapses, producing a magnetohydrodynamic (MHD) "heave," or "late-time," power surge that overloads equipment connected to the power and telecommunications infrastructure. This late- time effect adds to the initial HEMP effect, and systems connected to long-lines power and communications systems may be further disrupted by the combined effects. Smaller isolated systems do not collect so much of this third energy component, and are usually disrupted only by the first energy component of HEMP.