Solar Variations and the Upper Atmosphere
Extending from the top of the middle atmosphere to some hundreds of kilometers into space is the Earth's upper atmosphere (Figure 1.2) and its embedded ionosphere. This tenuous layer of neutral and charged particles shields the human habitat from high energy solar radiation and particles, enables part of the extensive communication network on which society increasingly relies, and is the medium in which thousands of spacecraft now orbit. Unlike the relatively placid lower atmosphere, the upper atmosphere is a region of extreme spatial and temporal variability, constantly agitated by solar radiative and auroral forcings. Driving the processes that at any instant define the physical state of the upper atmosphere and ionosphere is the solar radiation at wavelengths less than about 180 nm. Many of the region's continually changing physical phenomena derive directly or indirectly from changes in this radiation and from the impact of energetic particles channeled into the upper atmosphere at high latitudes via the Earth's magnetic field.
While solar variability exerts a dominating influence on the Earth's upper atmosphere, any direct effect on the biosphere appears to be more subtle than that exerted by solar forcing of the middle and lower atmospheres. The fact that the highly variable upper atmosphere is coupled to the middle atmosphere through chemical, radiative, and dynamical
mechanisms, and to the troposphere through the global electric circuit, cannot be ignored. Understanding how the upper atmosphere varies naturally, and how it may be affected by human activities, is necessary from a societal and economic perspective because of the critical role played by the upper atmosphere in communications, navigation, national defense, and a wide assortment of space related endeavors, including the presence of humans in space. Furthermore, current modeling studies indicate that the upper atmosphere may itself be sensitive to global change caused by human activities.
Solar EUV and UV Radiation
The Sun's ultraviolet radiation at wavelengths less than about 180 nm varies considerably more than does the UV radiation that is absorbed in the middle atmosphere and the visible radiation that penetrates to the Earth's surface (see Figure 1.1). Solar cycle changes of 100 percent are typical in solar radiation at wavelengths from 10 to 100 nm; the soft X-rays (1 to 10 nm) that penetrate to the lowest layers of the upper atmosphere vary by an order of magnitude. This highly variable energy from the Sun is deposited entirely in the terrestrial upper atmosphere via absorption of the primary constituents, O2, N2, and O. Without heating from the absorption of solar extreme ultraviolet (EUV) and UV radiation, the thermosphere and the ionosphere would not exist at all. This heating, which varies with solar activity, is responsible for the increase of temperature with height above about 100 km (see Figure 1.2) and for driving most of the bulk motions of the gases within the entire region. Large variability in the basic properties of both the thermosphere and ionosphere is the direct result of the variability in the solar EUV and UV input (as illustrated in Figure 1.2 by the change in the temperature profile from minimum to maximum solar activity).
Measurements of Solar EUV Spectral Irradiance
Current knowledge of the magnitude and variability of the solar EUV energy deposited in the upper atmosphere is based almost entirely on a brief four-year period of measurements made by the Atmosphere Explorer
(AE-E) spacecraft about 15 years ago. These measurements revealed a considerable increase in the solar EUV flux during the ascending phase of solar cycle 21. Some emissions at wavelengths shorter than 30 nm increased by factors of 10 to 100 between solar minimum conditions in 1976 and maximum activity in 1980. These emissions emanate from the highest, hottest layers of the Sun's atmosphere (the solar corona). Radiation at wavelengths between 30 and 120 nm, formed lower in the solar atmosphere (the chromosphere), varied somewhat less, by factors of two to three from the minimum to the maximum of activity in solar cycle 21. At still longer UV wavelengths, solar cycle variability decreases from a factor of two near 100 nm to about 10 percent near 200 nm. In addition to the overall change in solar radiation between solar minimum and maximum, the AE-E data showed shorter term fluctuations on a monthly, daily, and even an hourly basis, with the coronal emissions being much more variable than the chromospheric emissions.
Essentially all interpretive studies of upper atmosphere phenomena now use scenarios of solar variability derived from the AE-E data base. However, AE-E did not monitor the highly variable soft X-rays, nor do the AE-E data agree with earlier rocket measurements about either the magnitude or the variability of the EUV irradiance (Lean, 1988). Concerns about the validity and limitations of the AE-E data base continue to be raised. AE-E's absolute irradiance calibration was derived from two Air Force Geophysics Laboratory rocket measurements, one during 1974 (which preceded the AE-E data) and another in 1979 (Heroux and Hinteregger, 1978; L. Heroux, private communication, 1981). Possible changes in the sensitivity of the AE-E instruments throughout the mission are unknown, since no provision was made for in-flight calibration. A comparison of the 1979 rocket measurement used for the AE-E calibration with a recent rocket measurement (Woods and Rottman, 1990) indicates significant inconsistencies in that only the strongest emission lines were enhanced in the 1979 spectrum, for which solar activity levels were higher. This contradicts current understanding of the origin of the EUV irradiance variations, which predicts that solar activity causes an increase in the EUV radiation at all wavelengths. The discrepancy is most likely the result of instrumental effects (see Lean, 1990 for details).
AE-E ceased operation at the end of 1980. In the ensuing decade only a few isolated measurements of the solar EUV spectral irradiance were
made -- from the San Marco satellite for nine months in 1988 (Schmidtke et al., 1992) and from a few rockets (Woods and Rottman, 1990). Some additional measurements of the EUV radiation integrated over very wide spectral bands have also been made from rockets (Feng et al., 1989; Ogawa et al., 1990). None of these observations has succeeded in clarifying the true amplitude and variability of the solar EUV radiation. Like the AE-E observations, they are compromised by an assortment of instrumental effects that make it extremely difficult to extract true solar spectral variability.
Since its launch in mid-1991, the Yohkoh satellite (Petersen et al., 1993) has monitored the soft X-ray flux from the Sun for most of the descending portion of solar cycle 22, providing uniquely valuable data about a highly uncertain region of the solar spectrum. Continuing these observations into the upcoming solar minimum and subsequent activity maximum will contribute to improved understanding of solar forcing of lower thermospheric NO concentrations, and the possible transfer of chemical energy between the upper and middle atmospheres.
Irradiance Variability Parameterizations
The absence of continuous, reliable observations of the solar EUV spectral irradiance has forced reliance on empirical variability models based on solar activity surrogates to estimate EUV spectral irradiances for use in upper atmosphere research and in operational applications. The AE-E solar irradiance data have been used to construct parameterizations of the solar EUV flux variations as a function of primarily the solar 10.7 cm radio emission that can be measured from the ground (Hinteregger et al., 1981; Tobiska, 1991). The measured solar EUV flux values cover the period from 1976 to 1980, but the solar flux models have been used to represent the solar EUV and UV fluxes for other periods, generating values that are typically inconsistent with earlier data (Figure 4.1). Furthermore, different empirical models developed from ostensibly the same data base can predict quite different EUV spectral irradiances (Lean, 1990).
Aeronomic studies of thermospheric and ionospheric properties indicate, not surprisingly, that existing solar EUV irradiance variability models are inadequate for many geophysical applications. While considerable
progress has been made during the past decade in developing sophisticated models of thermospheric and ionospheric aeronomic processes and global dynamics, there has been little improvement in the (currently inadequate) empirical parameterizations of solar EUV and UV radiation that these models use. This has led to the application of various correction factors to the original data to achieve agreement between model predictions and aeronomic observations. For example, the solar EUV flux below 20 nm has been doubled to force agreement with observed photoelectron spectra (Richards and Torr, 1988; Winningham et al., 1989) and the flux below 5 nm has also been scaled upwards to account for a measurement of thermospheric NO at maximum levels of solar activity (Siskind et al., 1990).
Improved solar irradiance variability models are needed not just for aeronomic research but, increasingly, for operational applications such as forecasting ionospheric conditions (Balan et al., 1994) and for predicting thermospheric density for satellite orbital and point density determinations. The lifetime and utility of an Earth-orbiting object depend on the density structure of the upper atmosphere, which is controlled by solar EUV radiation and consequently varies over time scales from hours to decades (White et al., 1994). Desired accuracies of 5 percent for thermospheric densities for operational purposes require a similar accuracy in knowledge of the solar EUV and UV spectral irradiance. The need for this knowledge is demonstrated in Figure 4.2, where the orbital decay rate of the SMM satellite can be seen to track changes in solar activity (as indicated by the 10.7 cm radio flux). SMM's reentry into the Earth's atmosphere is thought to have been accelerated by the progressively increasing solar activity in 1989, the ascending phase of solar cycle 22. High uncertainty surrounded the launch of the Hubble Space Telescope because of insufficient knowledge of the atmospheric drag that it would experience when launched at a time near maximum solar activity (Withbroe, 1989).
Auroral Particle And Electric Field Inputs
The Earth's thermosphere and ionosphere system responds not only to changes in the solar EUV and UV radiative input but also to particulate inputs of solar energy and momentum at high latitudes associated with auroral processes. Although these inputs are dominant at high latitudes
(Figure 3.3), they have a great variability that affects the basic structure and dynamics of the entire upper atmosphere system. At times, such as in brief periods during intense geomagnetic storms, energized electrons bombard the upper atmosphere, colliding with atmospheric constituents and transferring their energy, resulting in visual displays of auroral phenomena. The energy deposited at high latitudes in the aurora can increase by as much as two orders of magnitude relative to geomagnetically quiet conditions, locally exceeding the energy deposited from solar EUV radiation. Auroral energy inputs are known to have a significant effect on aeronomic processes and dynamics of the ionosphere, thermosphere, and mesosphere and perhaps indirectly (via couplings to the stratosphere) on the troposphere, even though the physical couplings are not understood.
In addition to knowledge of radiative energy inputs, global dynamic models of the thermosphere and ionosphere system require knowledge of the global distributions of auroral particle precipitation, electric fields, and currents. During the past decade, spacecraft such as the Atmospheric Explorer and Dynamics Explorer, as well as various ground based programs, have provided a good first order understanding of the energy inputs to the thermosphere and ionosphere. Some information on global particle inputs has been derived from satellite images of UV and visible auroral airglow. However, many unresolved questions remain about the variability of the fundamental energy inputs and the global distribution of electric fields and currents.
Many questions also remain about the impact of auroral processes on global change. For example, how are the atmospheric chemical species such as NO that are produced by auroral processes transported globally? How might they be transported to the lower atmosphere, where they may influence global atmospheric properties? How do the enhanced currents and fields produced during geomagnetic storms influence properties of the troposphere and at the ground by coupling into the Earth's global electric circuit?
Global Currents and Electric Field Couplings
The Earth and its atmosphere are almost permanently electrified. The Earth's surface has a net negative charge, and an equal positive charge is distributed throughout the atmosphere. The atmospheric region above about 60 km is generally considered the upper conducting boundary of the global electric circuit, which includes electrical interactions between all atmospheric regions. This upper conducting boundary is formed by solar ionization of atmospheric constituents.
Global Circuit Processes
Three main generators operate in the Earth's global electric circuit (Figure 4.3): (1) thunderstorms, (2) the ionospheric wind dynamo, and (3) the solar wind/magnetosphere dynamo. These processes are reviewed in NAS (1986) and are only briefly described here.
Thunderstorms are electrical generators whose global activity provides a current output that maintains a vertical potential difference of about 300 kV between the ground and ionosphere, with a total current flow between the two of about 103 A. The variability of thunderstorm activity results in diurnal, seasonal, and interannual variations in the potential differences and currents in the circuit. The electrical processes associated with thunderclouds are many and complex. In general, a conduction current flows from the top of the cloud toward the ionosphere and into the global circuit. Beneath the thundercloud a number of complex currents flow. The total Maxwell current, defined as the sum of all of these current systems plus the displacement current, has been shown to vary slowly over the storm's history, suggesting that this electrical quantity is coupled to the meteorological structure of the storm. Recent aircraft measurements show that Maxwell current output from thunderstorms is related to the lightning flash rate.
The ionospheric wind dynamo is produced in the region of the atmosphere where neutral winds can have the effect of moving an electrical conductor (the weakly ionized plasma) through the Earth's geomagnetic field. This produces an electromotive force that generates potential differences of 5 to 10 kV with a total current of 105 A extending over thousands of kilometers and flowing primarily on the dayside of the Earth.
This current system is highly variable because of the changing tides and other disturbances propagating into the dynamo region from the middle atmosphere. Solar and auroral variability alter thermospheric winds and ionospheric electrical conductivities and thus influence the currents and fields in the dynamo region.
The flow of solar wind around and partly into the magnetosphere produces the solar wind/magnetosphere dynamo, which sets up plasma motion in the magnetosphere as well as producing electric fields and currents. This interaction is highly variable, but typically generates horizontal dawn-to-dusk potential drops of 40 to 150 kV across magnetic conjugate polar caps. The solar wind/magnetosphere dynamo is associated with a current flow of about 106 A between the magnetosphere and ionosphere. This generator depends on the properties of the solar wind flowing past the Earth's magnetosphere as well as on any internal current
flows within the magnetosphere. The magnetosphere and the Earth's near-space environment are more fully discussed in Chapter 5.
There is considerable variability in tropospheric electrical parameters that influence the properties of the global circuit. These parameters include global cloudiness as it affects the electrical conductivity; turbulence in the planetary boundary layer; aerosols; pollution; radioactive ion production near the Earth's surface; fog; and surface processes in grasslands, forests, deserts, and ocean spray. Many of these processes have been studied in isolation, but their combined impact on the global electric circuit has not been properly evaluated.
Important processes in the middle atmosphere also have implications for variability in the global circuit. Aerosols produced by volcanoes can affect the electrical conductivities and electric fields. Rocket measurements of electric fields in the mesosphere indicate strong departures from Ohm's law, suggesting the presence of an as yet unidentified generator operating at mesospheric heights. Electrical conductivity enhancements in polar regions associated with energetic particle precipitation, solar proton events, and Forbush decreases in cosmic ray fluxes following solar eruptions all influence the properties of the global circuit in ways that are not well understood at present.
Electrical Couplings Between the Upper and Lower Atmospheres
Large horizontal electric fields (100 to 1000 km) generated within the ionosphere project downward in the direction of decreasing electrical conductivity, effectively down to the ground. Small horizontal electric fields (1 to 10 km) are rapidly damped as they map downward into the atmosphere from ionospheric heights. Since the electrical conductivity of the Earth's surface is large, horizontal electric fields cannot be maintained there, and a vertical electric field variation results to accommodate horizontal variations of ionospheric potential. Calculations show that the solar wind/magnetosphere generator can increase or decrease by up to 20 percent the air-Earth current and ground electric field at high latitudes during geomagnetic quiet times, with larger perturbations during geomagnetic storms. The magnitude of the ground variations also depends on the alignment of the potential pattern over the ground, being much enhanced by mountainous terrain in the polar region.
While electric and magnetic field coupling between the upper and lower atmospheres is well established physically, the impact on processes in the troposphere and biosphere and on other processes important for global change is unknown. Furthermore, human activities are slowly changing the atmospheric structure, atmospheric and ionospheric composition, aerosol loading, land surface properties, and other tropospheric variables, all of which will probably have some influence on the properties of the global electric circuit (Price and Rind, 1994).
Solar Forcing and Global Change Within the Upper Atmosphere
That the Earth's upper atmosphere is forced directly by variable solar energy inputs on all time scales is well established. Global mean temperature, global wind circulation, and constituent particle densities change continuously, in response to changing solar activity throughout the 11-year cycle (e.g., Evans, 1982), with the extent of the changes generally increasing with altitude. From the minimum to the maximum of the Sun's activity cycle, upper atmosphere temperatures increase by many hundreds of degrees (Figure 1.2), a direct consequence of solar EUV and UV heating. Greenhouse gases such as carbon dioxide and methane contribute to the radiative balance of the Earth's upper atmosphere as well as the lower atmosphere. Carbon dioxide cooling is the dominant cooling mechanism in the atmospheric region between 70 and 200 km; infrared cooling by CO2 is largely responsible for the temperature minimum near 80 km shown in Figure 1.2.
Most studies of the climate change anticipated from the anthropogenic loading of greenhouse gases focus on the effects on the troposphere and middle atmosphere (Rind et al., 1990; Hansen et al., 1993). Recognition that trace gases released into the Earth's atmosphere from human activity could perturb the climate of the Earth provided much of the motivation for the USGCRP. As discussed in Chapter 2, these studies suggest that the troposphere will warm and the middle atmosphere will cool as trace gas concentrations increase into the twenty-first century. In the upper atmosphere, as in the middle and lower atmospheres, increases in anthropogenic
gases are expected to affect the energy balance between solar heating and infrared cooling.
Recent studies have shown that trace greenhouse gases could effect considerable change in the structure of the Earth's upper atmosphere (Roble and Dickinson, 1989; Rishbeth, 1990). The global mean temperature of the upper atmosphere has been projected to cool by 10 K at altitudes near 70 km, and by 50 K around 150 km, in response to doublings of CO2 and CH4 concentrations (Figure 4.4). These changes will be superimposed on upper atmosphere temperature variations of many hundreds of degrees generated by changes in solar energy inputs throughout the 11-year activity cycle (Figure 1.2). Concomitant redistributions of major and minor constituents should occur throughout the entire atmospheric region. In the thermosphere, the atmospheric density at a given altitude has been projected to decrease by as much as 40 percent. The atmospheric scale heights that govern thermospheric and ionospheric properties should also be reduced, and the peak height of the ionospheric F2 region may be lowered by 20 km.
As a result of changes in the basic thermal and compositional structure of the atmosphere, increases in CO2 may also damp the response of the thermosphere and ionosphere to solar and auroral variability. These changes also could affect the propagation of atmospheric tides, gravity waves, and planetary waves into the thermosphere. It is not clear how changes in the basic atmospheric structure and dynamics will affect the ionospheric wind dynamo and the coupling of the ionosphere and magnetosphere with the solar wind, but changes in thermospheric circulation might result in a changed electrodynamic structure of the upper atmosphere, and through dynamo action, alter magnetosphere/ionosphere coupling processes, as well as the entire terrestrial electric field system.
Couplings of the Upper Atmosphere to the Lower Atmosphere
The most direct coupling between the upper and lower atmospheres is electrical. As discussed previously, large horizontal electric fields map almost unattenuated to the Earth's surface, where they perturb the vertical electric field maintained by global thunderstorm activity. In addition, the
geomagnetic quiet-time dynamo current and highly variable auroral current systems induce magnetic currents in the Earth's crust and alter the ground electric field. These perturbations have demonstrable impacts on human infrastructure, such as power networks and oil pipelines (Allen et al., 1989), but as yet unknown impacts on biological and atmospheric physical processes.
Also suspected are chemical and dynamical couplings between the upper and lower atmospheres. Nitric oxide is produced by solar soft X-ray and EUV radiation and auroral particle dissociation of molecular nitrogen into excited and ground states of atomic nitrogen that react with molecular oxygen. NO concentrations change significantly as a result of larger variations in solar soft X-ray fluxes (Siskind et al., 1990). Nitric oxide is important for understanding possible radiative and dynamical couplings of solar variability effects in the thermosphere with the middle atmosphere (Garcia et al., 1984; Siskind, 1994). For example, if transported
downward to the region of 40 km, NO may have an important influence on the ozone density of the stratosphere (Huang and Brasseur, 1993). While the catalytic destruction of ozone in this region is relatively well understood (see Chapter 3), the transport of thermospheric-generated nitric oxide into the region is not. Because the photochemical destruction lifetime is on the order of one day throughout the sunlit mesosphere, NO must move downward during the polar night to reach the upper stratosphere. Two-dimensional modeling studies, as well as spacecraft observations of middle atmospheric ozone abundances, suggest that this may indeed be a viable coupling mechanism (Garcia et al., 1984; Solomon and Garcia, 1984). Changes in nitrate content of antarctic snow associated with solar activity may reflect these processes (Dreschhoff and Zeller, 1990).
In addition to solar-related processes, the lowest layers of the upper atmosphere are influenced by turbulent breaking of atmospheric gravity waves and tides and by sporadic, intermittent compositional exchanges of atomic oxygen and nitric oxide. These complex turbulent exchange processes influence long-lived species such as carbon dioxide, carbon monoxide, water vapor, and atomic and molecular hydrogen. Although theoretical studies have indicated that chemical, dynamical, and radiative interactions are a viable coupling mechanism, the whole question of thermospheric/lower atmospheric exchange is not well understood, primarily because our atmosphere between about 50 and 200 km is virtually unexplored on a global basis.