Forecasting the Mirror in the Sky

Photo by MC3 Paul Kelly


By Dr. Christoph R. Englert, Dr. Fabrizio Sassi, Dr. Sarah E. McDonald, Dr. Douglas P. Drob, Clayton E. Coker, Dr. Katherine A. Zawdie, Andrew C. Nicholas, Dr. Scott A. Budzien, Dr. Andrew W. Stephan, and Dr. Daniel P. Eleuterio

Charged layers on the very top of Earth’s atmosphere have long been used to reflect radio waves, acting much like a mirror in the sky. This effect has been known and used for about a century, and today it enables skywave over-the-horizon radar and long-distance communication.

While the ionosphere makes these capabilities possible, the high-altitude weather between about 40 and 600 miles (65-1,000 kilometers) can cause significant ionospheric disturbances. These disturbances can cause signal absorption and scattering, which can seriously degrade reflected radio signals. Ionospheric disturbances are often also detrimental to signals that go through the ionosphere for satellite communication or geolocation.

During the day, several layers of charged particles that form the ionosphere are created by the intense solar ultraviolet and x-ray radiation, which ionizes neutral air molecules and atoms. The ionosphere decays at night, with the lower layers decaying quickly after sunset and the top layer lingering until sunrise. It is well known that changes in solar radiation forcing can cause significant changes in the ionosphere. However, we now know that the dynamics and composition of the neutral atmosphere, which still constitutes more than 99 percent of the air mass at these altitudes, also is a significant driver of the ionosphere. It is especially important for the bottom-side of the ionosphere, which is used as the “mirror in the sky” for radio waves. In fact, research performed in the past decade has confirmed that during the long periods without much variation in solar forcing, the meteorological forcing from below can cause ionospheric electron density changes of 50 percent or more.

Forecasting the Ionosphere

The state of the ionosphere can be measured using a variety of observational techniques. From the ground, the electron density profile of the bottom side ionosphere can be measured using vertically reflected radio signals from a collocated transmitter/receiver pair (ionosonde) or using signal paths from a transmitter to a receiver at a different location (oblique sounder). Another common method is to derive the total electron content or column density between a ground receiver and a GPS satellite by examining the delay of the radio signals going through the ionosphere. Observations also can be made with satellite instruments by observing, for example, the naturally occurring ultraviolet airglow that originates from recombination of charged particles in the ionosphere, or by observing radio signals emitted on the ground or by other satellites.

All these observational techniques have their strengths and limitations—e.g., they might only contain information on the bottom side or the integrated electron content along a given path from emitter to receiver, which complicates forming a complete picture of the global, three-dimensional ionosphere at any given time. There are, however, several models that ingest ionospheric measurements from across the globe to estimate its state. Within the Department of Defense, operational ionospheric specifications and short-term persistence forecasts are currently produced by the Global Assimilation of Ionospheric Measurements model, which runs at the 557th Weather Wing of the US Air Force.

We now know that any skillful, multiday ionospheric forecast must include reliable meteorological forcing from below and solar forcing from above. Civilian and military researchers are working on both of these aspects. Concerning the inclusion of the lower atmospheric meteorology, the community is standing on the cusp of creating a fundamentally new operational capability, by extending lower atmospheric weather prediction systems up in altitude and coupling them to a physics-based ionospheric model. Such a “ground to space” model, coupled with the ingestion of real-time data, is expected to be able to produce previously unavailable, multiday ionospheric forecasts, much like the weather reports for the lower atmosphere, on which we rely on a daily basis.

Preliminary results from a whole-atmosphere model, coupling the lower atmospheric meteorology to the upper atmosphere, where the ionosphere is formed. The color contours illustrate peak electron densities in the ionosphere. Photo courtesy of Naval Research Laboratory

A targeted basic research effort to explore the coupling of the neutral atmosphere with the ionosphere is currently under way at the US Naval Research Laboratory’s space science division. First results of this model coupling show previously observed ionospheric structures that are the signature of lower atmospheric forcing. The figure on the opposite page, for example, depicts the modeled peak electron density across the globe, assuming the local time is close to noon everywhere on Earth (i.e., the daytime ionosphere is well formed everywhere). The image clearly shows two bands to the north and south of the equator, a well-known ionospheric feature called the equatorial arcs, but it also shows a prominent longitudinal structure, peaking over the Pacific, the west coast of Africa and Southeast Asia. This three-peak structure is a consequence of atmospheric solar tides that are excited by thunderstorm activity in the tropical lower atmosphere (below 17 kilometers) and then propagate all the way to the altitudes of the ionosphere.

Similar to the conventional numerical weather prediction, ionospheric forecasting also is an initial condition problem, which means that the current condition of the atmosphere has to be well constrained by observations to provide an optimal initialization of the forecast model. Obtaining measurements of the neutral atmosphere (e.g., winds and temperatures) and the ionosphere (e.g., electron density and electric field) with sufficient global coverage and spatial resolution is difficult and expensive, especially over the oceans where satellite instruments are often the only option. In an effort to develop sufficiently sensitive and economical options for suitable space based instrumentation, the US Naval Research Laboratory is developing novel measurement techniques building on previous payloads, including the operational Special Sensor Ultraviolet Limb Imager sensor on the Defense Meteorological Satellite program satellites.

Recent instruments include the Limb-imaging Ionospheric and Thermospheric Extreme Ultraviolet Spectrograph/GPS Radio Occultation and Ultraviolet Photometer Co-located instrument (launched in February 2017) measuring neutral and ionized constituents by looking toward the horizon, through the upper atmosphere from the International Space Station; the miniaturized Winds Ions Neutrals Composition Suite sensor (next launch in 2018), which measures atmospheric properties at the satellite location; and the Michelson Interferometer for Global High-resolution Thermopheric Imaging instrument (to be launched in 2018) measuring neutral wind and temperature in the lower ionosphere (below 300 kilometers) on board the NASA Ionospheric Connection Explorer mission.

The overall goal of our research efforts is to provide high-altitude specification and forecasting, and necessary high altitude data to the national Earth System Prediction Capability vision, which is to establish a global physical Earth system analysis and prediction system to provide seamless predictions covering hours to decadal timescales including the atmosphere, ocean, land, cryosphere, and space.

About the authors:

Dr. Englert is the head of the geospace science and technology branch of the space science division at the US Naval Research Laboratory. Dr. Sassi, Dr. McDonald, Dr. Drob, Clayton Coker, Dr. Zawdie, Dr. Nicolas, Dr. Budzien, and Dr. Stephan are researchers in the geospace branch of the space science division at the US Naval Research Laboratory. Dr. Eleuterio is the program manager of the earth system prediction capability at the Office of Naval Research.

This work was supported by the Office of Naval Research, NASA, and the Department of Defense Space Test Program.