M. J. Buonsanto, Massachusetts Institute of Technology, Haystack Observatory, Westford, MA 01886; T. J. Fuller-Rowell, CIRES, University of Colorado, and NOAA Space Environment Center, Boulder, CO 80303
Disturbances on the Sun can produce dramatic effects in the space environment surrounding the Earth. Energetic particle effects become more intense and pose a hazard to astronauts and damage spacecraft electronics; satellite lifetimes are shortened by increased atmospheric drag, and communications and navigation are disrupted by the changing plasma environment.
"Space weather" has become the modern idiom for these effects, and periods of high activity are called geomagnetic storms. During a storm the ionosphere can be severely altered. A typical episode may reveal either a large decrease (negative phase) or increase (positive phase) in the normal daily peak ion density (NmF2) or total electron content (TEC). These changes in ion density are sometimes called ionospheric storms, and often persist for more than a day after a period of high geomagnetic activity.
In this article we have placed the emphasis on the response of the middle and low latitude ionosphere to geomagnetic storms, although we will show that many of the characteristic "ionospheric phases" are reactions to changes in the thermospheric wind and composition. The thermosphere is the electrically neutral component of the upper atmosphere. The ionosphere comprises less than 1% of the mass of the upper atmosphere, yet it has a significant influence on advanced communication and navigation systems; both have important economic consequences. As society begins to rely on more complex technologies, these systems become more vulnerable to environmental effects. Predictions of storm effects in the Earth's upper atmosphere are far from reliable at present, having been hindered by a paucity of observations and gaps in our understanding of the physics of these complex phenomena.
Many approaches to mitigating the effects of space weather require understanding the thermosphere and ionosphere, specifying or predicting its current state, and improving forecasts of space weather changes. These tasks present a challenge, and will depend on sophisticated global models of the Earth's environment based on our physical understanding that are updated constantly with real-time solar and geophysical data.
Through the Coupling, Energetics, and Dynamics of Atmospheric Regions (CEDAR) Storm Study, scientists from the United States and other countries who are interested in the F region and thermosphere are joining together to study the coupled upper atmospheric regions at different heights and on a global scale. They are studying particular events that occurred in March 1990, June 1991, November 1993, and May 1995, for which especially good data coverage exists.
Geomagnetic storms occur when high-speed plasma injected into the solar wind from coronal mass ejections or coronal holes impinges on Earth's geomagnetic field. If the arriving solar wind plasma has a southward magnetic field, energy is coupled efficiently into Earth's magnetosphere and upper atmosphere. The magnitude of the ensuing geomagnetic storm has come to be defined by the strength of the low-latitude magnetic index, Dst, which is a measure of the magnetospheric ring current. Although the ring current is not the main driver of the upper atmosphere, the sources of both are related; Dst can reflect the level of magnetospheric energy input to the upper atmosphere. So a geomagnetic storm, from the perspective of the upper atmosphere, is a period when there is a large increase in this magnetospheric source for several hours to days.
When the storm occurs, auroral particle precipitation increases and expands to lower latitudes than normal, and the magnetospheric electric field mapped to the atmosphere intensifies and expands in concert with the auroral oval. The auroral particles heat and ionize the gas and increase the conductivity of the atmosphere, and the electric field combines with the increased conductivity to enhance Joule heating, which is the dominant atmospheric energy source during a storm.
Joule heating can increase from tens of gigawatts during quiet times to hundreds of gigawatts during severe geomagnetic disturbances. The combined input can dump thousands of terajoules of energy into the upper atmosphere during a storm and can raise the temperature of the gas by hundreds of degrees Kelvin. Thermal expansion of the atmosphere raises neutral density and can have significant effects on satellite drag. Ionospheric ions drift in response to the electric field, and, by colliding with the atmosphere, can drive winds in excess of 1 km/s at high latitudes. Many interesting science problems center around these high-latitude processes, and the combined energy source is the driver of the global storm-time changes in the upper atmosphere.
Much of the interest in understanding the response of the upper atmosphere to geomagnetic storms stems from the need to predict ionospheric variability for operational needs. The ionospheric response often appears chaotic, so it is difficult to capture in an empirical model. It is now clear that many of the ionospheric changes at midlatitude can be understood as a response to thermospheric wind and composition perturbations. Global thermospheric storm winds and composition changes are driven by energy injection at high latitudes. Wind surges propagate from high latitudes around the globe and transport plasma along magnetic field lines to regions of altered chemical composition, changing ion recombination rates. Many of the increases in NmF2 and TEC are thought to result from these traveling atmospheric disturbances.
The divergent nature of the wind causes upwelling of molecular rich thermospheric gas from lower altitudes. These regions of enhanced molecular species at F region altitudes, or composition "bulges," can be transported by the preexisting background wind fields and by the storm winds [Fuller-Rowell et al., 1996a] ; the changed composition again feeds back to the ionosphere. The regions of upwelling, in which N2 and O2 increase, cause the ionosphere to decay faster and create negative phases of ionospheric storms. Further equatorward, downwelling (decreases of N2 and O2) causes the ionosphere to decay more slowly and is responsible for some of the positive phases of ionospheric storms. The generation of these composition features and their subsequent transport by the wind field can explain the seasonal, local time, and regional dependence of the ionospheric response to a storm. Figure 1 shows a composition bulge developing in a simulation of a storm in December 1982. It illustrates the strong relationship between the regions of increased molecular species N2 and O2 on the left side and depletions in the ionosphere on the right side.
Fig. 1. Illustration of the high correlation between regions of enhanced molecular species in the neutral atmosphere (on the left-hand side) and decreases in the ratio of the storm to quiet NmF2 (on the right-hand side). The regions of enhanced molecular species are shown as increases of mean molecular mass (yellow/orange), and the ionospheric depletions are shown by a decrease in the ratio (blue). A snapshot of the southern hemisphere at 0600 UT from 10°S to the South Pole is depicted. The figure is taken from Fuller-Rowell et al. [1996b] .
The dynamical changes are complicated because they are driven by the highly variable magnetospheric energy sources. The wind surges propagate and interact around the globe and often appear as a random mixture of waves. Exactly where a composition bulge will be created, like the example in Figure 1, is difficult to determine; composition changes are created by persistent divergence of the wind field in areas of significant energy injection (mainly Joule heating). Accurate knowledge of the spatial and temporal distribution of the magnetospheric sources is required to predict where and when these composition changes will manifest themselves. In spite of these difficulties, systematic features are revealed, such as seasonal and local-time dependencies.
Ionospheric changes are frequently characterized by the response of the peak F2 layer electron density (NmF2) obtained from ionosondes. Figure 2 [Rodger et al., 1989] illustrates the seasonal and local-time response of NmF2 at the magnetic midlatitude site of the Argentine Islands. Data from many storms have been used to illustrate the UT-averaged variation of the storm/quiet ratio of NmF2. On the log scale, -0.5 units represents a reduction in NmF2 by 40%, or a loss of nearly half the ionosphere at the peak. Each month shows a similar local-time variation, with a minimum in the morning hours around 0600 LT and a maximum in the evening hours around 1800 LT. This local-time "ac" variation is superimposed on a "dc" shift of the mean level that varies with season. It is most positive in winter (May–July) and most negative in summer (October–February). The figure shows the prevalence of positive storms in winter and of negative storms in summer. This averaged picture is helpful in understanding the processes, but it should be noted that individual storms show large deviations from the average behavior. The complexity of the response is the main reason why no comprehensive model of ionospheric storms has emerged.
Fig. 2. Average seasonal and local time variations in ln (N/No), where N/No is the observed storm/quiet ratio of peak F2 layer electron density, at Argentine Is. (65°S) for 1971–1981, taken from Rodger et al. [1989] . The zero level (N/No) = 1 is shown for each month, and the shading denotes times of increased electron density where (N/No)&#;1.
Numerical simulations such as those illustrated in Figure 1 can help us understand the averaged picture and also help us explain why storms do not always fit this classical picture. Model simulations show that the local-time effect is caused by the background diurnal wind field moving composition bulges in latitude. On the dayside the prevailing winds tend to flow poleward, pushing the composition bulge away from midlatitudes. On the nightside the winds are more equatorward, so they transport the bulge back towards the equator. The result is an "ac" variation as seen in Figure 2. The seasonal variation is then a natural extension of this theory and can easily be explained by the prevailing circulation from summer to winter transporting the composition bulge to summer midlatitudes; this is when the main negative ionospheric phases are seen (Figure 2).
A particular storm will develop wind surges and composition bulges in regions defined by the particular history of the magnetospheric sources, which will be unique. Maximum thermospheric composition changes tend to occur in the longitude sector that passes through the nightside during the driven phase of a storm. During the December 1982 storm the Australian sector and the Indian Ocean encountered the large composition changes (Figure 1). Ionosonde observations over Australia during this storm recorded ionospheric depletions of up to 50%. This regional response gives rise to the longitude or UT dependence that has made thermospheric and ionospheric storms so difficult to decipher.
An interesting positive ionospheric feature, called the "dusk effect," has been observed in eastern North America. It is seen as a large short-lived enhancement in NmF2 and TEC, usually in the evening hours following a geomagnetic storm sudden commencement. Eastern North America is at a high geomagnetic latitude, compared to its geographic latitude, due to the offset of the geomagnetic pole to this longitude sector, so it has been natural to try to relate this phenomenon to the proximity of the auroral zones. Figure 3 shows an example of the dusk effect, as observed by the Millstone Hill (42.6°N, 288.5°E) incoherent scatter radar. The steerable antenna was pointed sequentially at a 45° elevation angle to the south, west, north, and east. The observation to the south probed the F region ionosphere at a geodetic latitude of 39.7°, while the one to the north probed the ionosphere at 45.5°, a difference in latitude of only 5.8°. In the figure, Ne to the south of Millstone Hill (at 350 km altitude) was more than 4 times larger than Ne to the north at the peak of this event. The extremely large temporal and spatial gradients apparent in the figure pose a severe test for operational models needed by modern technological systems.
Fig. 3. Electron density at an altitude of 350 km observed on May 26, 1990, by the Millstone Hill steerable antenna pointing at 45° elevation angle to the south, west, north, and east. An exceptionally large north-south gradient in electron density is apparent during the evening dusk effect.
Ionospheric changes at low latitude are particularly sensitive to electrodynamic phenomena. Variations in equatorial plasma drifts affect the evolution of the equatorial ionization anomaly (maxima in Ne 15° either side of the geomagnetic equator), and development of equatorial Ne irregularities. Disturbance effects have been determined for Jicamarca (11.9°S, 76.8°W, magnetic dip 2°N) by Fejer and Scherliess [1995] , but longitudinal variations of disturbance effects are only just now being addressed. Low-latitude vertical drifts during storms at Jicamarca depend on local time as well as the time elapsed since the onset of the storm. The drifts are affected by prompt penetration of magnetospheric electric fields, as well as by lingering dynamo electric fields from the disturbed neutral winds and storm-related changes in ionospheric conductivity.
Electron density is extremely variable in the equatorial zone between sunset and midnight due to the presence of irregularities with scale sizes ranging from less than 1 m to greater than 200 km. The Ne irregularities cause scintillations in satellite signals, spread F in ionograms, and strong coherent backscatter radar returns, which may appear as plumes extending above 800 km in altitude. The plumes are related to large-scale Ne depletions called "bubbles" that can cause Ne to vary by 2–3 orders of magnitude across the bubble boundaries. The unstable conditions after sunset that may result in irregularities are a steepening gradient in Ne below the F2 peak and the post-sunset enhancement in the upward E × B vertical drift. Gravity waves likely provide a seed mechanism for the onset of the irregularities.
How a geomagnetic storm affects equatorial irregularities is unclear. Some researchers believe that storm-related traveling atmospheric disturbances provide a seed mechanism for some of the very high altitude depletions that are observed as plumes. Other scientists find geomagnetic activity to be anticorrelated with the occurrence of equatorial irregularities. Storm electric fields may suppress the postsunset enhancement in the vertical E × B plasma drift and end up stabilizing the plasma rather than creating irregularities.
If the magnetospheric drivers were known perfectly, could the atmospheric response be modeled accurately? Can the main features of the ionospheric response be explained simply by transport effects of the winds and neutral composition changes? We do not know.
What causes the dusk effect remains unknown (Figure 3). Several different mechanisms have been proposed, but it appears that more than one may occur simultaneously. Enhanced equatorward neutral winds from high-latitude heating pushes plasma that is constrained to move along geomagnetic field lines. Equatorward winds lift the ionization upward to a region of lower neutral air density, reducing the O+ recombination rate that depends on the concentrations of N2 and O2 molecules. If this occurs during sunlit hours when photoionization is taking place, Ne increases. The uplifting could also be caused by electric fields, which may also transport higher-density, lower-latitude plasma to higher latitudes where it contributes to the dusk effect.
Decreases in N2 and O2 densities, and thus in the O+ recombination rate, are seen in global circulation model simulations in the dusk sector because of neutral winds set up by ion drag blowing in the direction of positive gradients in N2 and O2 mass mixing ratios. Traveling atmospheric disturbances, set up by localized heating at high latitudes, may cause the dusk effect by compressing ionization as the tilted wave fronts pass by. Learning the exact mechanisms will probably depend on simulating the dusk effect with global models such as the National Center for Atmospheric Research Thermosphere Ionosphere Electrodynamics General Circulation Model (NCAR-TIEGCM) [Richmond et al., 1992] or the Coupled Thermosphere Ionosphere Model [Fuller-Rowell et al., 1996a] , that self-consistently simulate ionospheric and thermospheric variations.
Another outstanding issue is the role played by vibrationally excited molecular nitrogen (N2*). The O+ recombination rate increases markedly if N2* is present in appreciable quantities. Concentrations of N2* rise with increasing electron temperature (Te) and reach their highest levels at summer solar maximum. They may also increase, when Te and concentrations of N2 rise during a major storm. There is still considerable uncertainty in calculations of the effects of N2* on the O+ recombination rate because of uncertainties in the calculations of the populations of the various excited states, and in the vibrational dependence of the O+ + N2 reaction rate. Comparison of NmF2 at Millstone Hill during a major storm with the Field Line Interhemispheric Plasma (FLIP) model [Torr et al., 1990] indicates that the increase of N2* decreases NmF2 by a factor of 2 or more.
Many other issues remain unresolved, such as what is the role of plasmaspheric loss and refilling of the midlatitude ionosphere during a storm? What is the effect of penetration and dynamo electric fields and what causes the fascinating electrodynamic changes observed at low latitudes? And, will we ever be able to predict the onset of irregularities at low latitudes that cause scintillations in radio signals? There are also many questions relating to the impact of storms in the middle and lower atmosphere to be addressed in the future.
We thank J. C. Foster and D. S. Evans for helpful comments. Work at MIT was supported by National Science Foundation Grant ATM-9523673; work at the University of Colorado and NOAA was supported by NASA grant NAGW-3530.
Fejer, B. G., and L. Scherliess, Time dependent response of equatorial ionospheric electric fields to magnetospheric disturbances, Geophys. Res. Lett., 22, 851, 1995.
Fuller-Rowell, T. J., M. V. Codrescu, H. Rishbeth, R. J. Moffett and S. Quegan, On the seasonal response of the thermosphere and ionosphere to geomagnetic storms, J. Geophys. Res., 101, 2343, 1996a.
Fuller-Rowell, T. J., M. V. Codrescu, R. G. Roble, and A. D. Richmond, How Do the Thermosphere and Ionosphere React to a Geomagnetic Storm?, Chapman Conference on Magnetic Storms, AGU Monogr. Ser., in press, 1996b.
Hargreaves, J. K., The Solar-Terrestrial Environment, 420 pp., Cambridge University Press, New York, 1992.
Kelley, M. C., The Earth's Ionosphere: Plasma Physics and Electrodynamics, 487 pp., Academic Press, San Diego, Calif., 1989.
Prölss, G. W., Ionospheric F-region storms, in Handbook of Atmospheric Electrodynamics, vol. II, edited by H. Volland, pp. 195–248, CRC Press, Boca Raton, Fla., 1995.
Richmond, A. D., E. C. Ridley, and R. G. Roble, A thermosphere/ionosphere general circulation model with coupled electrodynamics, Geophys. Res. Lett., 19, 601, 1992.
Rodger, A.S., G.L. Wrenn, and H. Rishbeth, Geomagnetic storms in the Antarctic F-region, II, Physical interpretation, J. Atmos. Terr. Phys., 51, 851,1989.
Torr, M. R., D. G. Torr, P. G. Richards, and S. P. Yung, Mid- and low- latitude model of thermospheric emissions, 1, O+(2P) 7320 Å, and N2(2P) 3371 Å, J. Geophys. Res., 95, 21,147, 1990.
Back to Science for Everyone
