The antenna array will be used as an imaging radar receiver for transmissions from the HF Vertical Incidence Pulsed Ionospheric Radar (VIPIR) in Cayey, Puerto Rico, for studies of phenomena such as atmospheric gravity waves, plasma irregularities, and space weather monitoring (Grubb et al., 2008). These observations would complement past observations made by the Cayey HF and Arecibo 430-MHz radars, such as those shown in Figure 1, with mesoscale radar images of the ionosphere surrounding and in-between the Cayey and Arecibo instruments. In the same way, radar imaging can be used to supplement airglow imagers, providing a different view of the same geographic regions and also operating through gaps in airglow data during the day and on cloudy nights. The array might also be used for space weather studies in a passive imaging mode as an interferometric riometer (McKay et al., 2015), and scintillation measurements could provide additional data on ionospheric irregularities and structure (Booker et al., 1987; Booker and Tao, 1987).

Plasma turbulence is part of the small-scale physical foundation of space weather. If we can better understand the fundamental processes that drive plasma turbulence, we will be able to better understand space weather as well. Ionospheric plasma turbulence may be divided into two categories: natural and artificially stimulated. Natural turbulence involves microscopic processes directly related to space weather. Stimulated turbulence involves the same microscopic processes but is controllable through the use of high-power transmitting, or heating, facilities (Grach et al., 2016; Streltsov et al., 2018; Yampolski et al., 2019), and this capability can be used to better understand natural turbulence. The radar and radio imaging array discussed here is designed to allow in-depth studies of both kinds of turbulence.

High-power, high-frequency ionospheric research transmitters capable of creating artificial ionospheric turbulence are operated at the High-Frequency Active Auroral Research Program (HAARP) observatory in Gakona, Alaska, at the European Incoherent Scatter Scientific Association (EISCAT) observatory in Tromsø, Norway, at the Sura observatory in Vasilsursk, Russia, and, until 2020, and, likely again in the near future, at Arecibo Observatory in Arecibo, Puerto Rico (Streltsov et al., 2018; Ferguson et al., 2022). Artificial ionospheric turbulence produces stimulated ionospheric radio emissions (also known as stimulated electromagnetic emissions or SEE), which share many characteristics with natural radio emissions (Thidé et al., 1982; Leyser, 2001; LaBelle and Treumann, 2002; LaBelle, 2012; Sergeev et al., 2013; Eliasson et al., 2021). Many spectral features have been identified, and some have been associated with specific plasma processes. Much attention has been paid to the heights of the emission regions, however, little work has been done in regard to the horizontal locations of the emissions, or to their geometrical or geographic relation to the geomagnetic field and to each other. High-frequency radio waves reflect or refract not only at different heights, but at different horizontal offsets (Rietveld et al., 1993). Different spectral features, due to differing plasma processes, might be expected to occur not only at different altitudes, but with different spatial distributions, for example, at different horizontal offsets relative to the geomagnetic zenith, i.e., roughly in the north-south direction (Isham et al., 2005; Tereshchenko et al., 2006). Figure 2 and Figure 3 show that simple HF radio interferometric methods are sensitive enough to detect changes in the locations of SEE source regions. The array will extend that technique to multiple antennas with multiple baselines, which can then be used for imaging of the emission region.

Radio imaging of artificially-stimulated radio emissions would be transformative. Images of the structure of stimulated radio emissions from the radio/plasma interaction region will help in determining the exact geometries of the source regions of the many radio spectral features with respect to the geomagnetic field. In addition, radio images at each polarization (O and X) for each emission feature could be constructed (Carozzi et al., 2001). Turbulence at high latitudes is known to be dependent on the longitudinal aspect angle with respect to the geomagnetic field, and this could be studied through radio imaging using the array (Isham et al., 2005). A variety of observations of stimulated radio emissions from Arecibo have been done using previous Arecibo HF transmitters (Thidé et al., 1989; Thidé et al., 1995; Mahmoudian et al., 2019). The array discussed here will be capable of much higher time and frequency resolution than could previously be measured.

The antenna elements of the array will be active crossed electric dipoles, and so will be capable of making a separate image in each polarization, which for ionospheric emissions would be the two circular polarizations, O and X. Very little work has been done on the polarization of stimulated ionospheric emissions, but the available results indicate that the different spectral features likely have distinct polarization properties (see the left panel of Figure 4). Polarization data might also be applied to the study of possible orbital angular momentum (OAM) modes in the received emissions (see the right panel Figure 4). The ability to routinely measure the polarization images of the stimulated radio signals will almost certainly lead to new and unexpected discoveries.

The imaging and polarization capabilities of the array promise to significantly advance our understanding of these little-studied aspects of stimulated radio emissions. At the longer wavelengths, where imaging is not feasible, the phase coherence of the radio receivers of the array allows radio direction-finding to be used. Direction-finding of stimulated radio emissions from the ionosphere (Isham et al., 2005; Tereshchenko et al., 2006) (see Figures 2, 3) has not yet been done at mid or low latitudes.

Radio imaging and direction-finding measurements can be complemented through the placement of the relocatable receivers of the array on long baselines to the east or west, for example, in eastern Puerto Rico or on nearby islands. This can provide a variety of oblique perspective angles from which to observe the radio emissions. Variations in the properties of the radio emissions as a function of observing angle would provide additional clues useful in the unraveling of turbulent processes, as the radiation pattern of the emissions is determined by ionospheric currents driven by those processes.

The second harmonic and subharmonics of stimulated radio emissions have been observed on limited occasions in the past (Karashtin et al., 1986; Derblom et al., 1989; Blagoveshchenskaya et al., 1998). The Arecibo HF transmitter most recently operated at two frequencies, 5.1 and 8.175 MHz; the receivers of the array will allow observations of radio emission harmonics of those frequencies up to the fourth or fifth at 5.1 MHz and third at 8.175 MHz, and also of subharmonics. Measurements of harmonic and subharmonic emissions may shed light on the emission mechanisms by either verifying or constraining emission theories and models.

An interesting and related topic is the search for the two-plasmon decay instability (Kruer, 1990; DuBois, 2000; Russell and DuBois, 2001; DuBois et al., 2011). To search for two-plasmon decay, the HF transmission frequency should ideally be chosen to be greater than the peak F-region frequency, and the receiving instrument tuned to one-half and three-halves of the transmitted value. Oblique radio emissions, as discussed above, are also thought to be a possible diagnostic for two-plasmon decay (Meyer and Zhu, 1993). While there has been extensive theoretical work on the mechanisms of a variety of SEE features, including the broad upshifted maximum (BUM), the downshifted maximum (DM), and the narrow continuum (NC) (Cheung et al., 1998; Mjølhus, 1998; Eliasson et al., 2012; Sergeev et al., 2013; Grach et al., 2016; Streltsov et al., 2018), there is much room for improvement in our understanding of the emission processes. In particular, to date there is no predictive theory of electromagnetic radio emissions such as exists for the electrostatic plasma waves that typically accompany them (Cheung et al., 2001; DuBois et al., 2001). Measurements made with the imaging radio array discussed here will assist in improving existing models.

Because of the refractive properties of the ionosphere at high frequencies, the radio imaging techniques discussed here give apparent direction of arrival of the signals being imaged. To obtain the true directions, HF wave propagation must be taken into account. HF propagation codes, also known as ray tracing codes, along with observations and models of the profile and structure of the ionosphere, may be used as needed to correct images and arrival angles for distortions caused by the ionosphere (Henderson et al., 2022). Density profile information for ray tracing calculations may be obtained from the instrument cluster operating on Culebra Island, Puerto Rico (PI Dr. Pedrina Terra), from the Digisonde and VIPIR HF radars operating in Cayey, Puerto Rico (PI Dr. Terence Bullett), and from instruments to be installed by collaborators alongside the array in Aguadilla, Puerto Rico. Ionospheric models may be used to help fill in data gaps. In addition, as previously mentioned, it is likely that the Arecibo HF transmitter will be restored to operation (Ferguson et al., 2022). If so, the Arecibo optical imagers and lidars could continue in operation. With relatively little additional work, an ionosonde could be installed at the Arecibo site or nearby, and the Arecibo 430-MHz radar could be restored and restarted.

There is a long-standing discussion about the physics of high-frequency radio wave scattering from the ionospheric plasma and the optimal way to process measurements of the scattered waves. One of the measurement methods used is known as the dynamic sounding, or Dynasonde, technique. A fundamental strategy of the Dynasonde technique is the use of individual echoes, detected without averaging the data from multiple transmitted pulses, as is frequently done by other methods (Wright and Pitteway, 1979; Wright et al., 1980; Sedgemore et al., 1998; Reinisch et al., 2005; Bertoni et al., 2006; Zabotin et al., 2006; Reinisch et al., 2009; Paznukhov et al., 2012). During Dynasonde data processing, it has been noted that an echo detected on a given pulse may frequently be seen again on the following pulse, suggesting that the sources of the individual echoes are deterministic structures and not random. However, dynamic sounding relies on direction-finding, without the angular resolution that radio imaging techniques may provide, so that it is not possible to say whether a given Dynasonde echo is from a defined and well-located structure or due to a geometric average of the scattering from one or more extended structures. One way to study this process would be to compare radar aperture synthesis images made using the radio array with the locations of particular echoes measured by the Dynasonde technique.

Aguadilla is located north of the rugged hills of central Puerto Rico, where many thunderclouds form and pass overhead, frequently undergoing electrical activity while doing so. The radio array discussed here can be used as a lightning interferometer capable of locating the positions of tens of thousands of discrete radio emission sources from a single flash of lightning (Stock et al., 2014; Hare et al., 2018), and to study medium and high-frequency electromagnetic radio emissions due to collisions between sprite streamers (Garnung et al., 2021). These capabilities would dramatically complement the types of thunderstorm measurements that were made using the previous Arecibo 430-MHz and 46.8-MHz radars (Holden et al., 1986; Chilson et al., 1993). Measurements of lightning emissions by the radio array may also be used to compute rapid ionograms using the ionospheric reflections of the emissions (Obenberger et al., 2018).

The array may be used for radar and radio observations of meteors, which would complement past observations of meteors made by the Arecibo Observatory 430-MHz and 46.8-MHz radars (Holden et al., 1986; Chilson et al., 1993; Isham et al., 2000; Obenberger et al., 2020).

The multiple, phase-coherent measurement channels of the array could be used to develop the full-polarization capabilities of vector antennas, both singly and in small arrays, of interest in radio communications and radar and radio remote sensing, and in the study of radio or electromagnetic orbital angular momentum (OAM), also known as radio phase modes. Orbital angular momentum has been studied extensively in the optical frequency range and has relatively recently come into the area of active research in radio frequencies (Thidé et al., 2007; Mohammadi et al., 2010a; Mohammadi et al., 2010b). Careful measurement of the phase differences in the received waves at widely separated locations, as would be possible using the distributed polarization measurement capabilities of the array, can also yield clues to possible OAM content (Mohammadi et al., 2010a; Mohammadi et al., 2010b). The previous Arecibo high-power HF transmitter had a three-point transmitting geometry that could be phased to transmit OAM modes ±1, although not pure modes due to the limited number of transmitting elements. This would likely also be possible, and even improved, in a restored Arecibo high-power HF transmitter (Ferguson et al., 2022) (see the right panel Figure 4).

In addition to vector antennas, the flexibility of the array allows exploration of other antenna designs and radio techniques, both at HF and higher frequencies. For optimal compatibility and performance with the Cayey and Arecibo transmitters, the primary receiver frequencies are optimized between 2 and 25 MHz. However, an alternate input allows the receivers of the array to operate down to 0.1 MHz and unaliased up to 60 MHz, and aliased up to 400 MHz. Thus, with some additional preparation and using suitable antennas, signals at other frequencies may be received. For example, at frequencies above 50 MHz, passive radar techniques might be developed using existing FM and digital television transmitters (Meyer and Sahr, 2004; Mir and Sahr, 2007; Vertatschitsch, 2013).

The multiple channels of the receivers will allow studies of the natural and man-made radio background as a function of polarization. Vertical polarization is expected to contain more noise than the horizontal polarizations, but this has not been quantified in Puerto Rico. This is important for the effective use of vector antennas, and for measurements of radio angular momentum, as discussed above, and will provide opportunities for student publications in professional journals. Graduate student projects could include the development of software to remove coupling effects between antennas and their surroundings, including nearby antennas and other nearby objects.

An intriguing project for both university and high school students is the use of the array as a radio telescope. The first radio astronomical observations were made by Karl Jansky at 20 MHz (Jansky, 1933), a frequency at which the array would commonly operate. Such a frequency in Puerto Rico may sometimes be impeded by the ionosphere during the daytime, but would be possible to use routinely at night.

As noted above, using the alternative signal inputs of the receivers, radio frequencies of up to 400 MHz may be received, opening possibilities for many interesting projects for students of all levels in building antennas, filters, amplifiers and attenuators, firmware, and software, for special radar and radio projects. The differences between signals received on different antennas may be easily displayed, and the software-defined radio firmware of the receiving systems may be programmed to create a software radio which can detect and play radio stations received by the equipment, to aid in exploring the properties of electromagnetic waves and to learn about radio and communications technology.