The term electromagnetic pulse (sometimes abbreviated EMP) has the following meanings:
The fact that an electromagnetic pulse is produced by a nuclear explosion was known since the earliest days of nuclear weapons testing, but the magnitude of the EMP and the significance of its effects were realized very slowly.
In July 1962, a 1.44 megaton United States nuclear test in space, 400 km. above the mid-Pacific Ocean, called the Starfish Prime test, demonstrated to nuclear scientists that the magnitude and effects of a high altitude nuclear explosion were much larger than had been previously calculated. Starfish Prime also made those effects known to the public by causing electrical damage in Hawaii, more than 800 miles away from the detonation point, knocking out about 300 streetlights, setting off numerous burglar alarms and damaging a telephone company microwave link.
The EMP damage of the Starfish Prime test was quickly repaired because of the ruggedness (compared to today) of the electrical and electronic infrastructure of Hawaii in 1962. Realization of the potential impacts of EMP became more apparent to some scientists and engineers during the 1970s as more sensitive solid-state electronics began to come into widespread use.
The larger scientific community became aware of the significance of the EMP problem after a series of three articles were published about nuclear electromagnetic pulse in 1981 by William J. Broad in the weekly publication Science.  
The relatively small magnitude of the Starfish Prime EMP in Hawaii (about 5,600 volts/meter) and the relatively small amount of damage done (for example, only 1 to 3 percent of streetlights extinguished) led some scientists to believe, in the early days of EMP research, that the problem might not be as significant as was later realized. Newer calculations showed that if the Starfish Prime warhead had been detonated over the northern continental United States, the magnitude of the EMP would have been much larger (22,000 to 30,000 volts/meter) because of the greater strength of the earth's magnetic field over the United States, as well as the different orientation of the earth's magnetic field at high latitudes. These new calculations, combined with the accelerating reliance on EMP-sensitive microelectronics, heightened awareness that the EMP threat could be a very significant problem.
In 1962, the Soviet Union also performed a series of three EMP-producing nuclear tests in space over Kazakhstan called "The K Project". Although these weapons were much smaller (300 kilotons) than the Starfish Prime test, since those tests were done over a populated large land mass (and also at a location where the earth's magnetic field was greater), the damage caused by the resulting EMP was reportedly much greater than in the Starfish Prime nuclear test. The geomagnetic storm-like E3 pulse even induced an electrical current surge in a long underground power line that caused a fire in the power plant in the city of Karagandy. After the collapse of the Soviet Union, the level of this damage was communicated informally to scientists in the United States. Formal documentation of some of the EMP damage in Kazakhstan exists but is still sparse in the open scientific literature.
The case of a nuclear electromagnetic pulse differs from other kinds of electromagnetic pulse (EMP) in being a complex electromagnetic multi-pulse. The complex multi-pulse is usually described in terms of 3 components, and these 3 components have been defined as such by the international standards commission called the International Electrotechnical Commission (IEC).
The 3 components of nuclear EMP, as defined by the IEC, are called E1, E2 and E3.
The E1 pulse is the very fast component of nuclear EMP. The E1 component has an intense electric field that can quickly induce very high voltages in electrical conductors. E1 is the component that can destroy computers and communications equipment and is too fast for ordinary lightning protectors.
The E2 component of the pulse has many similarities to the electromagnetic pulses produced by lightning. Because of the similarities to lightning-caused pulses and the widespread use of lightning protection technology, the E2 pulse is generally considered to be the easiest to protect against.
The E3 component of the pulse is a very slow pulse, lasting tens to hundreds of seconds, that is caused by the nuclear detonation heaving the earth's magnetic field out of the way, followed by the restoration of the magnetic field to its natural place. The E3 component has similarities to a geomagnetic storm caused by a very severe solar flare.  Like a geomagnetic storm, E3 can produce geomagnetically induced currents in long electrical conductors, which can then damage components such as power line transformers.
The strongest part of the pulse lasts for only a fraction of a second, but any unprotected electrical equipment — and anything connected to electrical cables, which act as giant lightning rods or antennas — will be affected by the pulse. Older, vacuum tube (valve) based equipment is much less vulnerable to EMP; Soviet Cold War–era military aircraft often had avionics based on vacuum tubes. There are a number of websites that explore methods for protecting equipment in the home or business from the effects of an EMP attack.
Many nuclear detonations have taken place using bombs dropped by aircraft. The B-29 aircraft that delivered the atomic weapons at Hiroshima and Nagasaki did not lose power due to damage to their electrical or electronic systems. This is simply because electrons (ejected from the air by gamma rays) are stopped quickly in normal air for bursts below 10 km, so they do not get a chance to be significantly deflected by the Earth's magnetic field (the deflection causes the powerful EMP seen in high altitude bursts), but it does point out the limited use of smaller burst altitudes for widespread EMP.
If the aircraft carrying the Hiroshima and Nagasaki bombs had been within the intense nuclear radiation zone when the bombs exploded over those cities, then they would have suffered effects from the charge separation (radial) EMP. But this only occurs within the severe blast radius for detonations below about 10 km altitude.
During nuclear tests in 1962, EMP disruptions were suffered aboard KC-135 photographic aircraft flying 300 km from the 410 kt Bluegill and 410 kt Kingfish detonations (48 and 95 km burst altitude, respectively)  but the vital aircraft electronics were far less sophisticated than today and did not down the aircraft.
Several major factors control the effectiveness of a nuclear EMP weapon. These are:
Beyond a certain altitude a nuclear weapon will not produce any EMP, as the gamma rays will have had sufficient distance to disperse. In deep space or on worlds with no magnetic field (the moon or Mars for example) there will be little or no EMP. This has implications for certain kinds of nuclear rocket engines. See Project Orion.
A high-altitude nuclear detonation produces an immediate flux of gamma rays from the nuclear reactions within the device. These photons in turn produce high energy free electrons by Compton scattering at altitudes between (roughly) 20 and 40 km. These electrons are then trapped in the Earth's magnetic field, giving rise to an oscillating electric current. This current is asymmetric in general and gives rise to a rapidly rising radiated electromagnetic field called an electromagnetic pulse (EMP). Because the electrons are trapped essentially simultaneously, a very large electromagnetic source radiates coherently.
The pulse can easily span continent-sized areas, and this radiation can affect systems on land, sea, and air. The first recorded EMP incident accompanied a high-altitude nuclear test over the South Pacific and resulted in power system failures as far away as Hawaii. A large device detonated at 400–500 km (250 to 312 miles) over Kansas would affect all of the continental U.S. The signal from such an event extends to the visual horizon as seen from the burst point.
Thus, for equipment to be affected, the weapon needs to be above the visual horizon. Because of the nature of the pulse as a large, long, high powered, noisy spike, it is doubtful that there would be much protection if the explosion were seen in the sky just below the tops of hills or mountains.
The altitude indicated above is greater than that of the International Space Station and many low Earth orbit satellites. Large weapons could have a dramatic impact on satellite operations and communications; smaller weapons have less such potential.
Typical nuclear weapon yields used during Cold War planning for EMP attacks were in the range of 1 to 10 megatons. This is roughly 50 to 500 times the sizes of the weapons the United States used in Japan at Hiroshima and Nagasaki. Physicists have testified at United States Congressional hearings, however, that weapons with yields of 10 kilotons or less can produce a very large EMP. 
The EMP at a fixed distance from a nuclear weapon does not depend directly on the yield but at most only increases as the square root of the yield (see illustration above). This means that although a 10 kiloton weapon has only 0.7% of the total energy release of the 1.44 megaton Starfish Prime test, the EMP will be at least 8% as powerful. Since the E1 component of nuclear EMP depends on the prompt gamma ray output, which was only 0.1% of yield in Starfish Prime but can be 0.5% of yield in pure fission weapons of low yield, a 10 kiloton bomb can easily be 5 x 8% = 40% as powerful as the 1.44 megaton Starfish Prime at producing EMP.
The total prompt gamma ray energy in a fission explosion is 3.5% of the yield, but in a 10 kiloton detonation the high explosive around the bomb core absorbs about 85% of the prompt gamma rays, so the output is only about 0.5% of the yield in kilotons. In the thermonuclear Starfish Prime the fission yield was less than 100% to begin with, and then the thicker outer casing absorbed about 95% of the prompt gamma rays from the pusher around the fusion stage. Thermonuclear weapons are also less efficient at producing EMP because the first stage can pre-ionise the air which becomes conductive and hence rapidly shorts out the electron Compton currents generated by the final, larger yield thermonuclear stage. Hence, small pure fission weapons with thin cases are far more efficient at causing EMP than most megaton bombs.
This analysis, however, only applies to the fast E1 and E2 components of nuclear EMP. The geomagnetic storm-like E3 component of nuclear EMP is more closely proportional to the total energy yield of the weapon. 
A unique and important aspect of nuclear EMP is that all of the components of the electromagnetic pulse are generated outside of the weapon. The important E1 component is generated by interaction with the electrons in the upper atmosphere that are hit by gamma radiation from the weapon -- and the subsequent effects upon those electrons by the earth's magnetic field.
For high-altitude nuclear explosions, this means that the much of the EMP is actually generated at a large distance from the detonation (where the gamma radiation from the explosion hits the upper atmosphere). This causes the electric field from the EMP to be remarkably uniform over the large area affected.
According to the standard reference text on nuclear weapons effects published by the U.S. Department of Defense, "The peak electric field (and its amplitude) at the earth's surface from a high-altitude burst will depend upon the explosion yield, the height of the burst, the location of the observer, and the orientation with respect to the geomagnetic field. As a general rule, however, the field strength may be expected to be tens of kilovolts per meter over most of the area receiving the EMP radiation."
The same reference book also states that, ". . . over most of the area affected by the EMP the electric field strength on the ground would exceed 0.5Emax. For yields of less than a few hundred kilotons, this would not necessarily be true because the field strength at the earth's tangent could be substantially less than 0.5Emax."
In other words, the electric field strength in the entire area that is affected by the EMP will be fairly uniform for weapons with a large gamma ray output; but for much smaller weapons, the electric field may fall off at a comparatively faster rate at large distances from the detonation point.
It is the peak electric field of the EMP that determines the peak voltage induced in equipment and other electrical conductors on the ground, and most of the damage is determined by induced voltages.
(For nuclear detonations within the atmosphere, the situation is more complex. Within the range of gamma ray deposition, simple laws no longer hold as the air is ionised and there are other EMP effects such as a radial electric field due to the separation of Compton electrons from air molecules, together with other complex phenomena. For a surface burst, absorption of gamma rays by air would limit the range of gamma ray deposition to approximately 10 miles, while for a burst in the lower-density air at high altitudes, the range of deposition would be far greater.)
Non-nuclear electromagnetic pulse (NNEMP) is an electromagnetic pulse generated without use of nuclear weapons. There are a number of devices that can achieve this objective, ranging from a large low-inductance capacitor bank discharged into a single-loop antenna or a microwave generator to an explosively pumped flux compression generator. To achieve the frequency characteristics of the pulse needed for optimal coupling into the target, wave-shaping circuits and/or microwave generators are added between the pulse source and the antenna. A vacuum tube particularly suitable for microwave conversion of high energy pulses is the vircator.
NNEMP generators can be carried as a payload of bombs and cruise missiles, allowing construction of electromagnetic bombs with diminished mechanical, thermal and ionizing radiation effects and without the political consequences of deploying nuclear weapons.
NNEMP generators also include large structures built to generate EMP for testing of electronics to determine how well it survives EMP. In addition, the use of ultra-wideband radars can generate EMP in areas immediately adjacent to the radar; this phenomenon is only partly understood.
Information about the EMP simulators used by the United States during the latter part of the Cold War are now in papers under the care of the SUMMA Foundation, which is now hosted at the University of New Mexico.
The SUMMA Foundation web site includes documentation about the huge wooden Trestle simulator in New Mexico, which was the world's largest EMP simulator; and which used a specialized version of a Marx generator.
Typical modern scenarios seen in news accounts speculate about the use of nuclear weapons by rogue states or terrorists in an EMP attack. Details of such scenarios are always controversial. It is impossible to know what kind of capabilities such terrorists might acquire, especially if they are aided by state sponsors with access to advanced technology.
Some rogue states have developed an ability to deliver a light missile payload to the necessary altitude for an EMP attack. Nuclear weapons in general have a much heavier missile payload, however advanced weapons design enables larger weapon yields with lighter weight. It is difficult to know if any particular rogue state has the necessary combination of advanced missile technology and nuclear weapons technology to perform an effective nuclear EMP attack over an industrialized country.
A common scenario is the detonation of a device over the middle of the U.S. using long-range missiles that have historically been available only to major military powers. An offshore detonation at high altitude, by contrast, would present less technical difficulty and would disrupt both an entire coast and regions hundreds of miles inland (e.g. 120 mile altitude, 1,000 mile EMP radius).
The United States EMP Commission was authorized by the United States Congress in Fiscal Year 2001, and re-authorized in Fiscal Year 2006. The commission is formally known as the Commission to Assess the Threat to the United States from Electromagnetic Pulse (EMP) Attack.
The United States EMP Commission has brought together a group of notable scientists and technologists to compile several reports. In 2008, the EMP Commission released the Critical National Infrastructures Report. This report describes, in as much detail as practical, the likely consequences of a nuclear EMP on civilian infrastructures. Although this report was directed specifically toward the United States, most of the information can obviously be generalized to the civilian infrastructure of other industrialized countries.
The 2008 report was a followup to a more generalized report issued by the commission in 2004.
Non-technical writings about nuclear EMP, both in print and on the internet, quite often contain some very common misconceptions about EMP. Although these misconceptions usually contain an element of truth, they lead to a very considerable amount of confusion about the subject.
Such misconceptions include:
(1) Nuclear EMP is caused by a nuclear explosion in the atmosphere. The fact: Although it is true that a nuclear explosion in the atmosphere produces a strong electromagnetic pulse, the range of the EMP is extremely limited compared to the EMP produced by a nuclear explosion in space. (See the discussion above in the "Weapon altitude" and "Weapon distance" sections.) The statement about EMP as a result of a nuclear explosion in the atmosphere (or an air burst) is very commonly made when the writer is actually referring to a nuclear explosion above the atmosphere. The lack of public understanding of the importance of the relationship of altitude to weapon effects is a longstanding problem. The standard reference text on nuclear weapon effects published by the U.S. Department of Defense discusses this relationship extensively in the first two chapters, and provides mutually-exclusive definitions for phrases such as "air burst" and "high-altitude burst." 
(2) EMP is produced directly by special EMP weapons. The fact: This statement is only true about special types of non-nuclear EMP weapons of limited range (as discussed in the NNEMP section above). The statement is commonly made when referring to nuclear EMP. Electromagnetic pulse is a prompt secondary effect of a nuclear explosion, and nearly all of the nuclear EMP is produced outside of the weapon. (All nuclear weapons can produce EMP as a secondary effect, but the effect can be enhanced by special weapon design.)
(3) The only defense against nuclear EMP is a ballistic missile defense system. The fact: Although an anti-ballistic missile system can be a component of protection against nuclear EMP, it is seldom mentioned in the writings of scientists or the testimony of members of the United States EMP Commission. The experts on the EMP problem tend to concentrate upon making electronic equipment and electrical components resistant to EMP and upon keeping adequate spare parts on hand, and in the proper location, to enable prompt repairs to be made.
(4) EMP requires the use of very large thermonuclear weapons. The fact: This misconception continues to persist, and it contains an element of truth only because the E3 component of the pulse is roughly proportional to weapon yield. Thermonuclear weapons are very inefficient in producing the fast E1 component of EMP. (See the discussion in the "Weapon yield" section above.) Large thermonuclear weapons produce large energy yields through a multi-stage process. This multi-stage process is completed within a small fraction of a second, but it nevertheless requires a finite length of time. The first fission reaction is usually of relatively small yield, and the gamma rays that it produces pre-ionize the molecules of the upper atmosphere. This pre-ionization causes the gamma ray emission from the high-energy final fission explosion of the thermonuclear weapon (a fraction of a second later) to be relatively ineffective at producing a large E1 pulse.  (See the blue pre-ionization curve in the "Peak Electric Field at Ground Zero" graph above.)
(5) Nuclear EMP is not a problem because there are ways to protect against it. The fact: Although there are ways to protect against nuclear EMP (or to quickly begin repairs where protection is not practical), the United States EMP Commission determined that such protections are almost completely absent in the civilian infrastructure of the United States, and that even large sectors of the United States military services were no longer protected against EMP to the level that they were during the Cold War. The United States EMP Commission did not look at the civilian infrastructures of other nations.