Just as an opera soprano needs to sing at just the right frequency to break a glass, so a heater must target frequencies that match the natural plasma resonances in the ionosphere. Primary targets include the plasma frequency, a function of electron density; multiples of the cyclotron frequency of electrons spiraling around the magnetic field; and hybrid resonances that combine those fundamental frequencies.

Measurements of optical emissions excited by heated electrons yielded HAARP’s first unexpected result. Spotting such emissions at all was a feat, inasmuch as 20 years of attempts to do so at the EISCAT (European Incoherent Scatter) heater in Norway’s Arctic had been unsuccessful; in fact, HAARP scientists had been warned that looking for optical emissions would be a waste of time. Nevertheless, images recording the red 630.0-nm oxygen line revealed a faint blob turning on and off in sync with the heater; that could only mean HAARP had heated the electrons and excited the oxygen. The airglow showed an unexpected enhancement well away from the beam center, at the magnetic zenith—that is, the direction of the magnetic field. The obvious next step was to point the beam toward the magnetic zenith, which at HAARP is about 15° south-southwest of vertical. When the experiment was finally performed in 2002, as the HAARP array swung the beam through the magnetic zenith, the blob lit up and was 10 times as bright as airglow in any other location. A variation of that magnetic zenith effect had been previously observed at EISCAT, but neither the EISCAT nor HAARP version of the effect had been predicted and neither is fully understood.

Thanks to its frequency agility, the HAARP antenna can heat the ionosphere at specific altitudes where the transmission frequency simultaneously matches two resonances. In 2004, experiments exploiting that possibility produced green-line oxygen emissions at 557.7 nm. (The figure shows an airglow with red and green oxygen emission.) Those lines come from an excited state with an energy 4 eV above the ground-state energy; evidently, by “surfing” plasma waves, the electrons accelerated to energies well beyond the thermal energy. Another HAARP experiment in the same series heated an ephemeral ionospheric layer produced by an aurora; the resulting green spots were as bright as the aurora itself. Those extremely bright spots have since been reproduced but are not yet explained.

In 2007 HAARP expanded to its full design capability of 12 × 15 antennas and 3.6 MW of total power. During its first postexpansion science campaign, in February 2008, my colleagues and I obtained optical images with strange, unpredicted rings around the airglow spot. We hypothesized that if the plasma in the center of the beam were slightly enhanced in density relative to the background ionosphere, the density gradient could divert rays away from the center of the beam toward the location where the ring was observed. Careful examination of echoes from radio waves bounced off the ionosphere turned up evidence for a density-enhancing artificial plasma layer just below the natural ionosphere. Moreover, simulations of RF waves propagating through the observed layer put additional power right where the rings were seen.

We had not expected such artificial ionization to be possible, but we followed up with new experiments designed to optimize ionization production. In March 2009, just over 10 years after we were told that looking for airglow was futile, I stepped outside with a couple of coworkers during an ionization experiment and marveled at the light—visible with unaided eyes—from an artificial ionospheric plasma produced and sustained by radio waves transmitted from the ground.

In addition to generating unexpected phenomena, HAARP scientists used and further developed a diagnostic technique pioneered at EISCAT: stimulated electromagnetic emissions. The effect arises when plasma waves stimulated by the heater regenerate radio waves that are received on the ground as a complex spectrum of narrow peaks and broad bumps on either side of the transmission frequency. Some of those depend not only on electron density but also on ion mass, magnetic field strength, or other parameters. Thus the stimulated emissions provide a potentially powerful tool for analyzing conditions in the heated volume.

Admittedly, research at HAARP has not directly contributed to new corrections for ionospheric effects on navigation or communications systems. Instead, the many surprises encountered in HAARP experiments have made abundantly clear the need for quantitative predictive theory and modeling in the field of high-power RF-wave propagation. The complex equations describing plasma waves imply a whole zoo of wave modes that could potentially be excited by a transmitter.

But no one can predict with certainty whether a particular wave mode will absorb half the transmitted energy or only one part in a million. For example, observed artificial plasma production accounts for only about 5% of the energy available from the beam; some of the remaining 95% undoubtedly excites other modes that might mislead researchers into wrongly identifying the cause of the ionization. Stimulated electromagnetic emissions hold the greatest promise for helping scientists determine which wave modes are active during actual experiments.

An interesting and still unexplored aspect of artificial ionization is the complex interplay between the plasma created by radio waves and the bending or reflecting of radio waves by that plasma. As food for thought, have a look at the video that accompanies the online version of this Quick Study. You’ll see a wide range of spots, turbulence, and sharp gradients—despite the smoothly varying beam. If we are ever to develop practical applications of heating technology, we’ll need to find mathematical solutions describing the evidently complex feedback process.

In August 2015 the HAARP facility was transferred from the US Air Force to the University of Alaska so that HAARP scientists could continue their investigations of fundamental plasma physics in an academic environment.