Thermospheric and Ionospheric Disturbances

The focus of this work, under NASA and NSF support is the response of the thermosphere (that part of the atmosphere above about 90 km) and ionosphere to geomagnetic activity that manifests itself in various forms. The most familiar is bright auroral displays (northern lights) over extensive high latitude regions (within the auroral oval). Another is intense heating within the auroral oval by the particles responsible for bright aurora (electrons and to a lesser extent protons and hydrogen atoms) and by Joule heating arising from energy transfer from ionospheric currents to neutrals (mostly molecular nitrogen [N₂] and molecular oxygen [O₂]) in the lower regions of the thermosphere. The heating leads to upwelling of molecular-rich air into the upper thermosphere that is then transported out of the heated region. Strong winds are generated during extended periods of high activity (in excess of a hour) that on occasions blow the disturbed air to mid-latitudes over larger geographical regions.

The presence of this disturbed air (referred to as a thermospheric disturbance) can be seen from satellites that carry instruments capable of recording atmospheric emission at far ultraviolet (FUV) wavelengths (the FUV spectral region extends from about 100 to 200 nanometers). In particular, the disturbance can be seen on the dayside provided the observed spectral region contains significant emission from atomic oxygen [O] (one or both of the bright atomic oxygen features at 130.4 nm and 135.6 nm [designated as OI 130.4 nm and OI 135.6 nm]). A characteristic feature of satellite FUV observations is the absence of emission from below the thermosphere due to absorption by molecular oxygen. At longer wavelengths, emission from the thermosphere is overwhelmed by contributions from lower altitudes arising from Rayleigh scattering of sunlight.

It was noted above that the disturbed air arises from the upwelling of molecular-rich air that is then transported elsewhere. The FUV emission properties of this air are different than surrounding undisturbed air. In particular, there is a reduction in emission from O due to a decrease in the relative abundance of this species compared to the molecular species (N₂ and O₂). A striking example of this is seen in Figure 1 that shows data from the FUV imager on the DE-1 satellite (References to the experiment are Frank et al. [Space Sci. Instrum., 5, 369, 1981] and Frank and Craven [Rev. Geophys., 26, 249, 1988].) Two images are shown to contrast the dayglow (the bright extended portions of the images) between an undisturbed time (upper image recorded on Day 267 [September 24], 1981) and a strongly disturbed time (lower image recorded on Day 295 [October 22], 1981). The emission recorded within these images is over a broadband FUV spectral region dominated by OI 130.4 nm. If one were to show contours of constant dayglow brightness, they would be expected to follow contours of constant solar zenith angle (SZA) in the central portions of these images provided the atmosphere is in an undisturbed state. The contours displayed are for constant SZA to examine how well this holds for the two images. The variations in brightness in the central portion of the upper image are seen to be controlled primarily by solar zenith angle. There is a significant breakdown in this relationship in the lower image due to the presence of an extended region of reduced O relative to N₂ and O₂ (of these two latter species, the concentration of O₂ is minor compared to that of N₂ within the emitting region).

A technique has been reported by Strickland et al. [J. Geophys. Res., 104, 4251, 1999] that converts the dayglow seen in Figure 1 into a measure of the abundance of O relative to N₂. Designated as O/N₂, it is the ratio of the O to N₂ column densities from within the emitting region referenced to a fixed value of the N₂ column density. Figure 2 shows an image of O/N₂ derived from the data in the lower image of Figure 1. Seen is an extended region of reduced O/N₂ over portions of North America and further to the west. Reductions greater than a factor of two are seen to occur compared to adjacent undisturbed regions. It should be noted that the disturbance seen in Figure 2 is one of the most severe recorded by the DE-1 experiment but many others have been seen, especially early in the mission when solar activity was high.

A recent paper by Strickland et al. [J. Geophys. Res. 106, 21,049, 2001, 2000] has examined ionospheric data coincident with DE-1 based O/N₂ to determine the impact on the ionosphere of thermospheric disturbances characterized in terms of reduced O/N₂. The ionospheric data are in the form of N_max, the maximum of the electron density profile as recorded by ground-based ionosondes. Figure 3, taken from the paper, shows DE-1 data and O/N₂ during a disturbed period that resulted in a thermospheric disturbance over Russia. Data are shown for images recorded 24 minutes apart demonstrating the reproducibility of the derived O/N₂ as would be expected over this time span. The numbers in the O/N₂ images refer to ionosonde sites providing the ionospheric data considered in the study. Figure 4 shows the ionosonde data on the day of interest (day 287 [October 21], 1981) and on the following day as well. The time of interest within Figure 4 is when the data in Figure 3 were recorded (near 0600 universal time). A clear correlation is seen between reduced O/N₂ and reduced Nmax and is discussed in detail in the above referenced paper. The reported work demonstrates an important use of satellite FUV data, namely in being able to infer regions experiencing composition-related reductions in the ionosphere (referred to as negative ionospheric storm effects). One would like to know when and where such regions occur since they impact ground-to-ground radio communications, ground-based radar signals used to track space objects, and geolocation of transmitters using multiple space-based receivers.

Future FUV satellite measurements should play an important role in recording thermospheric and ionospheric disturbances, both in terms of their temporal and spatial characteristics. Examples of FUV instruments that should provide measurements superior to those in the past are GUVI on the NASA TIMED satellite (to be launched in 2000) and SSUSI on a series of DMSP satellites, with the first to also be launched in 2000. Physicists at CPI are continuing to investigate the relationship between FUV-based O/N₂ and the ionosphere including first-principles ionospheric modeling. These individuals have also taken the lead in the development of key remote sensing algorithms for GUVI and SSUSI (auroral and dayglow) that should offer excellent opportunities to extend the above work after the launch of these instruments.

Figure 1. DE-1 FUV images during undisturbed (upper image) and disturbed times. The contours refer to constant solar zenith angle. Dayglow dominated by emission from OI 130.4 nm is responsible for the bright portions of the images. Emission at high latitudes from the auroral oval can be seen in the lower image. Reductions in O relative to N₂ are responsible for the structure in the dayglow seen in the lower image.
Figure 2. O/N₂ derived from the dayglow portion of the lower image in Figure 1. A strong geomagnetic disturbance occurred several hours prior to the time shown above the image and led to the large thermospheric disturbance displayed in this image in the form of reduced O/N₂.
Figure 3. DE-1 FUV images recorded on Day 287 [October 21], 1981 for the times shown above the images along with corresponding images of O/N₂ derived from the dayglow portions of the data images. A large thermospheric disturbance is seen over Russia. The numbers identify ionosonde sites whose data are shown in Figure 4.
Figure 4. Ionosonde data (in the form of daily and monthly median values of Nmax) from the 13 sites identified in Figure 3. Significant negative ionospheric storm effects are seen at the sites located within the thermospherically disturbed region shown in Figure 3.