M. J. Buonsanto

Ionospheric effects of the March 1990 Magnetic Storm: Comparison of theory and measurement

This paper presents a comparison of the measured and modeled ionospheric response to magnetic storms at Millstone Hill and Arecibo during March 16-23, 1990. Magnetic activity was low until midday UT on day 18 when Kp reached 6, days 19 and 20 were quiet, but a large storm occurred around midnight UT on day 20 (Kp=7) and it was moderately disturbed (Kp=4) for the remainder of the study period. At Millstone Hill, the daytime peak electron density (NmF2) showed only a modest 30% decrease in response to the first storm and recovered to prestorm values before the onset of the second storm. The model reproduces the daytime peak electron density well for this period. However, the severe storm on March 20 caused a factor of 4 depletion in electron density, while the model densities were not greatly affected. The inclusion of vibrationally excited nitrogen (N2*) in the model was unable to account for the observed large electron density depletions afterward March 20. The storm did not appear to affect the overall magnitude of the electron density at Arecibo very much, but did cause unusual wavelike structure in the peak density and peak height following the storm. The model reproduces the daytime NmF2 very well for Arecibo, but after sunset the model densities decay too rapidly. This study indicates that successful modeling of severe ionospheric storms will require better definition of the storm time inputs, especially of the neutral atmosphere.

Comparison of models and data at Millstone Hillduring the 5–11 June 1991 storm

Abstract We compare measurements of the ionospheric F region at Millstone Hillduring the severe geomagnetic disturbances of 5–11 June 1991 with results from the IZMIRANand FLIP time-dependent mathematical models of the Earths ionosphere and plasmasphere. Somecomparisons are also made with the Millstone Hill semi-empirical model which was previouslyused to model this storm. New rate coefficients from recent laboratory measurements of the O + +N 2 and O + +O 2 loss rates are included in theIZMIRAN and Millstone Hill models. The laboratory measurements show that vibrationallyexcited N 2 and O 2 (N 2 ( v ) and O 2 ( v )) are both important at high temperatures such as found in the thermosphere during disturbedconditions at summer solar maximum. Increases in the O + +N 2 loss ratedue to N 2 ( v ) result in a factor ∼2 reduction in the daytime F 2 peak electron density. On some days inclusion of N 2 ( v ) improves theagreement between the models and the data, and on other days it worsens it. In the present workwe show for the first time the significant effect that the increase in the O + recombination rate due to O 2 ( v ) may have on the calculated NmF 2 . There are considerable uncertainties in the model calculations during the unusual,extremely disturbed conditions found during the daytime on 6 June. The results illustratedifficulties involved and the current state of the art in modelling severe disturbances, and thusprovide a benchmark against which future progress can be gauged.

TIME DEPENDENT THERMOSPHERIC NEUTRAL RESPONSE TO THE 2-11 NOVEMBER 1993 STORM PERIOD

Abstract Many satellite and ground-based observations from 2–11 November 1993 werecombined in the Assimilative Mapping of Ionospheric Electrodynamics (AMIE) procedure toderive realistic time dependent global distributions of the auroral precipitation and ionosphericconvection. These were then used as inputs to the Thermosphere–Ionosphere–ElectrodynamicsGeneral Circulation Model (TIEGCM) to simulate the thermospheric and ionospheric responseduring the storm period. The November 1993 storm was an unusually strong storm associatedwith a recurring high speed stream of solar plasma velocity in the declining phase of the solarcycle. Significant gravity waves with phase speeds of about 700 m/s caused by Joule heating werepresent in the upper thermosphere as perturbations to the neutral temperature and wind fields,especially on 4 November. The observed gravity waves in the meridional wind and in the height ofthe electron density peak at several southern hemisphere stations were generally reproduced bythe model using the AMIE high latitude inputs. Both model and observed equatorward windswere enhanced during the peak of the storm at Millstone Hill and at Australian ionosondestations. The observed neutral temperature at Millstone Hill increased about 400 K during thenight on 4 November, returning to normal on 9 November, while the model increased 300 K thefirst night at that location but was still elevated on 11 November. Enhanced westward windsduring the storm were evident in the UARS WIND Imaging Interferometer (WINDII) data. Theenhanced westward winds in the model were largest around 40–45° magnetic latitude at night,and also tended to be largest in the longitudes containing the magnetic poles. The peak westwardwind enhancements at 0 LT reached about 250 m/s at 300 km, and about 100 m/s at 125 km thefirst day of the storm at 40° magnetic latitude. At 20° magnetic latitude, the maximum westwardwind enhancements at 125 km at 0 LT appeared 2–4 days after the major part of the storm,indicating very long time constants in the lower thermosphere. The model showed global averageneutral temperature enhancements of 188 K after the peak of the storm that decayed with time,and which correlated with variations 8 h earlier in the Dst index and in the electric potential dropinput from AMIE. The global average temperature enhancement of 188 K corresponded to apotential drop increase of only about 105 kV. The results showed that the TIEGCM usingrealistic AMIE auroral forcings were able to reproduce many of the observed time dependentfeatures of this long-lived geomagnetic storm. The overall global average exospheric temperaturevariation correlated well with the time variation of the cross-tail potential drop and the Dst indexduring the storm period. However, the enhanced westward winds at mid-latitudes were stronglyrelated to the corrected Joule heating defined by the time dependent AMIE inputs.

Observed and calculated F2 peak heights and derived meridional winds at mid-latitudes over a full solar cycle

The peak height of the F2 layer, hmF2, has been calculated using the ‘servo’ model of Rishbeth et al. [(1978), J. atmos. terr. Phys. 40, 767], combined with the hedin et al. [(1988), J. geophys. Res. 93, 9959] neutral wind model. The results are compared with observed values at noon and midnight derived from ionosonde measurements at two mid-latitude stations, Boulder and Wallops Island, over a full solar cycle. The reduced height of the F2 layer, zmF2, is also computed for the same period using the observed hmF2 values and the MSIS-86 model. Day-night, seasonal, and solar cycle variations in zmF2 are attributed to neutral composition changes and winds. Anomalously low values of hmF2 and zmF2 during summer both at solar minimum and during the solar cycle maximum in magnetic activity may be associated with increases in the molecular to atomic ion concentration ratio. Under these circumstances the F2 peak may lie significantly below the O+ peak height calculated by the servo model. Neutral meridional winds at Wallops Island are derived from the servo model using the observed hmF2 values and the calculated O+ ‘balance height’. It is shown that if the anomalously low hmF2 values are used, unrealistically large poleward winds are derived, which are inconsistent with both theory and observations made using other techniques. For most conditions the F2 peak is clearly an O+ peak, and daily mean winds at hmF2 derived from the servo model are consistent with the hedin et al. (1988) wind model. Unexpectedly, the results do not show an abrupt transition in the thermospheric circulation at the equinoxes. Diurnal curves of the servo model winds reveal a larger day-night difference at solar minimum than at solar maximum.

Comparison of incoherent scatter observations of electron density, and electron and ion temperature at Millstone Hill with the International Reference Ionosphere

Abstract Millstone Hill incoherent scatter (IS) observations of electron density (Ne, electron temperature (Te) and ion temperature (Ti) are compared with the International Reference Ionosphere (IRI-86) for both noon and midnight, for summer, equinox and winter, at both solar maximum (1979–1980) and solar minimum (1985–1986). The largest difference inNe is found in the topside, where values of Ne given by IRI-86 are generally larger than those obtained from IS measurements, by a factor which increases with increasing height, and which has a mean value near two at 600 km. Apart from the bottom of the profile, which is tied to the CIRA neutral temperature, the IRI-86 Te model has no solar cycle variation. However, the IS measurements during the summer reveal larger Te at solar maximum than at solar minimum. At other seasons higher Te at solar maximum occurs only during the daytime at the greater heights. Nighttime Te is shown by the IS radar to be generally larger in winter than in summer, an effect not included in the IRI. This is apparently due to photoelectron heating during winter from the sunlit ionosphere conjugate to Millstone Hill. The day-night difference in Ti given by IRI-86 above 600km is not as large in the IS measurements.

Comparison of techniques for derivation of neutral meridional winds from ionospheric data

We carry out a detailed comparison between winds derived from F2 peak heights and winds obtained from incoherent scatter radar (ISR) line-of-sight velocity measurements. A total of 34 incoherent scatter radar experiments at Millstone Hill spanning all seasons and levels of solar activity are included in this study. For two experiments we compare results from five different wind-derivation techniques. According to work by Titheridge [1993, 1995a, b], neutral winds derived from the servo model are inaccurate during the sunrise and morning period because of a shift in the zero-wind F peak downward from the balance height. To investigate this effect, we determine a correction factor cfac to be applied to the servo model c parameter as a function of time of day for summer, equinox, and winter at both solar maximum and solar minimum. Our results confirm that a sunrise decrease in cfac is necessary to bring about best agreement between the servo model winds and the winds derived from the ISR ion velocity data at Millstone Hill. However, the effect is not large, so that a constant cfac for each season/solar activity level usually introduces less error than other factors which may result in differences between the servo model and ISR winds. These factors include measurement errors in hmF2 and the ISR line-of-sight ion velocities, spatial variations in the wind field above the station, and the assumption that hmF2 is the peak in the O+ altitude profile.

Comparison of measured and modeled solar EUV flux and its effect on the E-F1 region ionosphere

The response of the E-F1 region ionosphere to different solar EUV flux models is investigated theoretically using two different photochemical schemes, and the results are compared with incoherent scatter radar electron density measurements taken at Millstone Hill. The latest EUV flux model (Tobiska, 1991), which incorporates more recent measurements, has generally more flux at short wavelengths compared to the Hinteregger et al. (1981) flux model based on AE-E satellite data. This results in better agreement with the measurements in the E-F1 region and above. The Tobiska flux model, however, gives a smaller E region peak density, due to the influence of low Lyman s flux in the November 10, 1988, rocket measurements of Woods and Rottman (1990). The photochemical scheme of Buonsanto (1990) has been improved and now gives results similar to the more comprehensive scheme of (Solomon et al., 1988; Solomon and Abreu, 1989; S. C. Solomon and R. G. Roble, Simulation of the global thermospheric airglow, 1, Methodology, submitted to Journal of Geophysical Research, 1992), provided that the ratios of photoelectron impact ionization to photoionization (pe/pi) given by this latter model are included. The pe/pi ratios calculated by this model and by the models of Lilensten et al. (1989) and Richards and Torr (1988) differ significantly, and work is needed to resolve these differences. In general, the photochemical model results underestimate the data, especially in winter. This result agrees with that of the earlier paper by Buonsanto and could be resolved by decreasing MSIS-86 N2 and O2 densities in winter if additional ions were produced in the E region either by photoionization or by photoelectron impact ionization. The photoionization and photoabsorption cross sections of Conway (1988) give results in somewhat better agreement with observations than the cross sections of Torr et al. (1979). For the zenith angles considered (daytime conditions), the Chapman function method for calculating photoabsorption gives results in satisfactory agreement with a more rigorous calculation method using a formula from Rees (1989).

Estimation of the O+, O collision frequency from coincident radar and Fabry‐Perot observations at Millstone Hill

The formula for the O+, O momentum transfer collision frequency has been uncertain due to a discrepancy between results of theoretical calculations and some of the joint radar/optical studies. The former suggest a multiplicative factor equal to 1.2–1.3 times the formula derived by Dalgarno [1964] and Banks [1966], while the latter suggest a multiplicative factor F = 1.7 [Salah, 1993]. We present results of a new analysis of data from 30 nights of coincident incoherent scatter radar (ISR) and Fabry-Perot interferometer (FPI) experiments conducted at Millstone Hill between 1988 and 1992. The O+, O collision frequency is estimated from FPI measurements of the horizontal neutral wind in the magnetic meridian, ISR measurements of the ion drift velocity parallel to the Earths magnetic field and other data at the calculated height of peak 630 nm emission, and the mass spectrometer and incoherent scatter 86 model. A complete error analysis is carried out for each derived value of F. This allows us to carry out Monte Carlo simulations which confirm that random errors lead to an increase in the mean value of F and which provide us with an unbiased result, F = 1.15 ± 0.2. However, this result was obtained from an analysis which neglected vertical neutral winds, about which we have little information. The most likely effect of these winds would be an increase in the value of F, so that our best estimate from this study is F = 1.4 ± 0.3, which is consistent with theoretical calculations.

Ionospheric electron densities calculated using different EUV flux models and cross sections: Comparison with radar data

The recent availability of the new EUVAC (Richards et al., 1994) and EUV94X (Tobiska, 1993b, 1994) solar flux models and new wavelength bin averaged photoionization and photoabsorption cross section sets led us to investigate how these new flux models and cross sections compare with each other and how well electron densities (Ne) calculated using them compare with actual measurements collected by the incoherent scatter radar at Millstone Hill (42.6°N, 288.5°E). In this study we use the Millstone Hill semiempirical ionospheric model, which has been developed from the photochemical model of Buonsanto et al. (1992). For the F2 region, this model uses determinations of the motion term in the Ne continuity equation obtained from nine-position radar data. We also include two simulations from the field line interhemispheric plasma (FLIP) model. All the model results underestimate the measured Ne in the E region, except that the EUV94X model produces reasonable agreement with the data at the E region peak because of a large Lyman β (1026 A) flux, but gives an unrealistically deep E-F1 valley. The ionospheric models predict that the O2+ density is larger than the NO+ density in the E region, while numerous rocket measurements show a larger NO+ density. Thus the discrepancy between the ionospheric models and the radar data in the E region is most likely due to an incomplete understanding of the NO+ chemistry. In the F2 region, the photoionization rate given by EUV94X is significantly larger than that given by the EUVAC and earlier models. This is due to larger EUV fluxes in EUV94X compared to EUVAC over the entire 300-1050 A wavelength range, apart from some individual spectral lines. In the case of EUVAC, this is partly compensated for by larger photoelectron impact ionization due to the larger EUV fluxes below 250 A. The differences between ionospheric model results for the different cross-section sets are generally much smaller than the differences with the data.

Ring current heating of the thermal electrons at solar maximum

To quantify the energy input to the thermal electrons due to Coulomb collisional degradation of hot ions in the inner magnetosphere, the heating rate is calculated from the results of a time-dependent kinetic ring current model. The large June 4–7, 1991, storm during the last solar maximum, when the hot O+ content is maximal, is chosen for this study. Modeled electron heat fluxes into the topside ionosphere reach 1011 eV cm−2 s−1 on the dusk side, a large value that will certainly have an impact on the density, temperature, and composition of the upper ionosphere and thermosphere. Comparable maximum values of heat inputs to the inner magnetospheric thermal plasma are expected to arise again during storms of the present solar maximum. The calculated heating rates from the ring current simulations are compared directly with detailed ionospheric modeling results for the Millstone Hill field line (L = 3). It is seen that heating from the ring current is more than adequate to account for the nightside topside heat input necessary to obtain the observed electron temperatures during this storm, even taking into account the limitations of the comparison. The reason for this is the abundance of O+ in the ring current at energies of a few tens of keV deep in the inner magnetosphere.