Geophys. Res. Lett., 20, 1-4, 1993
© Copyright 1993 by the American Geophysical Union
Paper Number 92GL03012
G. Le and C. T. Russell
Institute of Geophysics and Planetary Physics, University of California, Los Angeles, CA
Abstract. The observations obtained during the IMS from the IGS magnetometer chain extending from Cambridge, England, to Tromso, Norway are used to study the time rate of change of the magnetic field at subauroral latitudes at the time of interplanetary shock passages. The time rate of change of the H component maximizes in the high latitude dayside sector. For these typical interplanetary shocks, the dayside value of time rate of change can be as high as ~ 3 nT/sec at Tromso and ~ 1 nT/sec at York. The time rate of change in the dayside roughly depends on the change of square root of solar wind dynamic pressure. The largest of these time rates of change are similar to but slightly smaller than those known to cause disruptive disturbances in power distribution and communication systems. Thus, the daytime effects of sudden impulses may be equal to or greater than the nighttime effects associated with substorms as measured by their impact on terrestrial systems.
Impacts of geomagnetic disturbances on human technologies have been recognized since the mid-nineteenth century (see Lanzerotti  and Lanzerotti and Gregori  for historical reviews). Time variations of the geomagnetic field during strong geomagnetic disturbances induce an electric field, that gives rise to potential differences on the Earth's surface. The potential differences can produce large currents in long conductors which are grounded at both ends, such as power transmission lines, telecommunication cables, and pipelines. Today, power distribution networks have become much more vulnerable than ever before because they have expanded to include many grounded points that span large distances (thousands of kilometers) and link large cumulative Earth-surface potentials. To avoid harmful effects from geomagnetic disturbances in these systems, it is essential to design power distribution systems to withstand large earth potential variations and induced currents. Thus, a better understanding of the solar wind-magnetosphere interaction as well as a better ability to predict its effects will greatly help the engineering of modern technology.
Sudden changes in solar wind dynamic pressure play an important role in producing geomagnetic disturbances (see reviews by Matsushita , Akasofu and Chapman , and Nishida ). Disturbances caused by the sudden changes in solar wind dynamic pressure are global rather than localized phenomena. When the solar wind dynamic pressure suddenly increases, the dayside magnetosphere is compressed, and complex current systems are set up in the magnetosphere, ionosphere and the Earth's interior. These current systems, varying with time in complicated and as yet not very well understood ways, can cause large disturbances in the geomagnetic field. The time rate of change of the geomagnetic field controls the induced voltage that is a threat to the human technologies.
Previous work has revealed that the disturbances caused by sudden changes in solar wind dynamic pressure depend on the local time and latitude of the observation site [eg., Araki, 1977], and on the change of solar wind dynamic pressure [eg., Siscoe et al., 1968; Russell et al., 1992]. Russell et al.  concluded that the simplest response occurs at low and mid latitudes (15 to 30), where the effects of equatorial and auroral ionospheric currents are minimal. They found that the horizontal component of the field at low and mid latitudes changes 16 nT for each change of 1 nPa^1/2 in the square root of the solar wind dynamic pressure. At high latitudes, the response in the H component can be classified as a rapid preliminary impulse followed by a longer main response [Araki, 1977]. We have recently completed a study of these changes at subauroral latitudes using IGS magnetometer data [Le et al., 1992]. We find that the main response in the H-component is stronger in the higher latitudes (> 65 deg) than that in the lower latitudes (< 65 deg) in the dayside, and the magnitude of the main response increases qualitatively with increasing change of square root of solar wind dynamic pressure.
While the response of the geomagnetic field to the sudden change of solar wind dynamic pressure has been studied by many authors, and much has been done on the change of the geomagnetic field caused by the disturbance, not much has been done on the time rate of change of the geomagnetic field during the disturbance In this paper, we will study the time rate of change of the geomagnetic field caused by the sudden change of solar wind dynamic pressure at subauroral latitudes by examining the IGS magnetometer data during years of 1978 and 1979. We will include only disturbances produced by the compression of the magnetosphere and exclude substorm and reconnection associated currents by examining the data under conditions of northward interplanetary magnetic field.
Historically, magnetic storms have been characterized by the strength of the main phase minimum in the horizontal magnetic field following the onset of the storm which may be marked by a sudden increase in the H-component or which may occur gradually [Chapman and Bartels, 1940]. The size of the sudden increases in magnetic field that mark the onset of the storm and that we now attribute to interplanetary shocks are not correlated with the size of the ensuing main phase minimum [Suguria and Chapman, 1960]. In fact, these sudden impulses can occur with no following main phase decrease. Furthermore, the intensity of auroral zone currents is not necessarily correlated with the depth of the main phase maximum [Russell et al., 1974]. Since these three current systems show much independence, even though often obviously causally linked, there has arisen some confusion about the meaning of the phrase "great magnetic storm". A period with extremely strong interplanetary shocks, such as the August 4, 1972 events, may possess a strong sudden impulse without an accompanying strong main phase. This has caused some controversy [Lanzerotti, 1992; Tsurutani et al. 1992] which we will sidestep herein by referring to periods of great magnetic disturbances. Thus the August 4, 1972 event, while not necessarily a great geomagnetic storm because of its modest main phase minimum, certainly ranks as a great magnetic disturbance because of the size of its sudden impulse, about 60 nT.
Since it is not generally recognized that the disruption of ground systems by this geomagnetic disturbance was associated with the sudden impulse when the IMF was northward, we review first the August 4, 1972 event. Moreover, this discussion also places into proper context our studies of the rate of change of the magnetic field along the IGS chain during the passage of interplanetary shocks when the IMF is northward, discussed in the following section. We also emphasize the complexity of geomagnetic induction effects due to their large-scale and integrated nature. Ground-based individual-site observations of time rate of change can not tell us directly the magnitude of the induced electric field on the ground, which requires good knowledge of Earth's conductivity. Nevertheless, observations of the time rate of change of the magnetic field during great magnetic disturbances in the past can serve as a guide to make rough estimations and predictions of these effects.
There are several well documented power system disruptions and cable outages caused by the induced currents associated with great magnetic disturbances. For example, the magnetic disturbance of February 10, 1958 caused a temporary power blackout in Toronto, Canada [Brooks, 1959] and the disruption of the first transatlantic telecommunication cable [Winckler et al., 1959]. During the great magnetic disturbance in August, 1972, the electric power system in the northern part of U.S. experienced several major disturbances [Albertson and Thorson, 1974]. It also caused a complete outage of the continental telecommunication cable near Plano, Illinois [Anderson et al., 1974]. The magnetic disturbance of March 12--14, 1989 caused a long lasting power outage in Quebec, Canada [Coles, 1989], but there was no disruption in the transatlantic telecommunication cable during this storm due to improved technology [Medford et al., 1989].
The great magnetic disturbance in August, 1972 was the first large magnitude event that has been thoroughly observed by many instruments, both ground-based and on spacecraft. These data were documented in World Data Center A, Report UAG-28. Several sudden commencements (SSCs) occurred during the period August 4--9, 1972 of intense geomagnetic disturbances. The one which caused disruptions in power distribution systems occurred near 2040 UT on August 4. Low-latitude magnetograms showed step-like increases in the H component of ~ 60 nT at Honolulu (21.1 deg MLAT, 12 LT) and San Juan (29.8 deg MLAT, 18 LT) [World Data Center A, 1972] and ~ 70 nT at Kakioka (25.8 deg MLAT, 08 LT) [Solar-Geophysical Data, 1973]. Since the response in H is ~ 16 nT/nPa^1/2 at mid and low latitudes [Russell et al., 1992], we can estimate that the change of square root of solar wind dynamic pressure was ~ 4 nPa^1/2. The interplanetary magnetic field was northward during this storm [Burlaga and King, 1979] and there were no substorms reported following the SSCs. Anderson et al.  showed that the SSC at ~ 2040 UT on August 4, 1972 affected the entire area of the North American continent which was on the dayside at local time from ~ 1000 to 1600 hours. The time rate of change of the magnetic field maximized at Meanook (61.9 deg MLAT), in western Canada, which was located near local noon. The maximum value was as high as ~ 1800 nT/min (30 nT/sec). The time rate of change of the magnetic field at the major site of communication disruption (Plano, Illinois) was ~ 700 nT/min (12 nT/sec). This time rate of change induced a surface electric field of as high as ~ 7 V/km along the disrupted continental telecommunication cable. In short, the major disruption of ground systems was not at the time of major energization of the magnetotail or substorms but at the time of sudden impulse (interplanetary shock) well ahead of the main phase of the storm.
Fortunately for ground systems disturbances such as those of August 4, 1972 are rare but this rarity make them difficulty to study. Thus to understand this process we must use observations under more normal solar wind conditions when effects are smaller but occur more often. The ground stations that we will use in this study are those in the magnetometer chain operated by the Geomagnetism Unit of the Institute of Geological Sciences (IGS) during the IMS [Stuart, 1982]. The IGS magnetometer chain forms a closely spaced north-south line from Northern Scandinavia to mid latitudes. There is a local time spread of about 2 hours along this north-south line. These stations are located at subauroral latitudes covering the range from 55 deg to 70 deg geomagnetic latitudes. The magnetometers of the IGS stations digitally sample magnetic variations every 2.5 seconds in three orthogonal directions. For the solar wind monitor, we use measurements from ISEE 3 and IMP-8 at 5 min resolution. We look for sudden impulse events at times of interplanetary shock passages when the solar wind dynamic pressure suddenly increases, in the years of 1978 and 1979. We examine the response of the ground-level magnetic field only when the IMF is northward. The magnitude of changes of magnetic field has been discussed previously by Le et al. . In this paper, because of our interest in the effects on communication and power distribution systems we will discuss only the time rate of change of the magnetic field.
Figure 1 shows an example of the response of geomagnetic field and the time rate of change of the magnetic field at York (56.4 deg MLAT), Lerwick (62.2 deg MLAT) and Tromso (66.9 deg MLAT) during a sudden impulse event at 1628 UT on January 5, 1978. All the IGS stations are in the local afternoon hours during this event. The response of the magnetic field is mainly in the H-component, which includes a short preliminary decrease followed by a main increase. The time rate of change in the H-component is greater than that in the D and Z-components for all these stations. The maximum time rate of change can occur during the main increase (e.g., Lerwick) or the preliminary decrease (e.g., Tromso). On this particular event, a dayside event, it is clear that the absolute value of the time rate of change is much higher at higher latitudes.
We studied 30 events in total. For some of events, data are not available for all the stations. In Figure 2, we show the time rate of change in H as a function of local time for York (triangles), Lerwick (squares) and Tromso (circles) for all the sudden impulse events. The observed maximum time rate of change (open symbols) is very scattered. The solar wind dynamic pressure increases different amounts for each of the sudden impulse events. This accounts at least in part for some of the scatter. The solid symbols connected by lines are the median value in each 6--hour bin with 3--hour overlap. For all the 3 stations, the median time rate of change show clear local time variations. The time rate of change is maximum on the dayside and is minimum near local midnight. Near local midnight, the time rate of change is similar for all three stations. But on the dayside, the time rate of change at Tromso is much higher than those at the lower latitude stations, York and Lerwick. The dayside value is as high as 3 nT/sec at Tromso and ~ 1 nT/sec at Lerwick and York. These values are less than those reported to have caused disruptions but all our events were rather small changes in the square root of solar wind dynamic pressure.
In Figure 3 are shown the maximum time rate of change in H versus the change in square root of solar wind dynamic pressure only for those events in dayside hours (0700--1700 LT) and we do not include data in the nightside when the ionospheric density is significantly less at these latitudes. It includes only events with solar wind dynamic pressure data available from ISEE 3 or IMP-8. The solid, dashed and dotted lines correspond to median slopes for Tromso, Lerwick and York, respectively. The correlation coefficients are 0.49, 0.62 and 0.56 for Tromso, Lerwick and York, respectively. Since we have a very limited number of data points for each station, the correlation is not significant. Qualitatively speaking, however, the time rate of change in H increases with increasing change of solar wind dynamic pressure. The median slope is larger for the higher latitude station. For Tromso, the median slope is ~ 1.5 nT/sec/nPa^1/2.
The time rate of change is greater in H, the north-south component, than in D and Z for all the IGS stations. Furthermore, the time rate of change in H maximizes at the higher latitude stations of IGS chain. The observations show that this maximum in the time rate of change in H covers entire dayside (~ 06--18 LT) and the effect is global rather than small scale. Thus, a sudden change in solar wind dynamic pressure will produce an induced electric field mainly in the east-west direction on the dayside at high (auroral) latitudes. Long conductors, aligned mainly in the east-west direction and grounded at both ends, will be the most vulnerable to the induced voltages caused by the sudden increase in solar wind dynamic pressure. The telecommunication cables which were disrupted or affected in the storms of August 1972, February, 1958, and March, 1989 were roughly in the east-west direction [Anderson et al., 1974; Winckler et al., 1959; Medford et al., 1989]. The observed Earth current during magnetic storm on February 11, 1958 was also reported to be in the west-east direction [Winckler et al., 1959].
The time rate of change in H roughly increases with an increasing change in the square root of solar wind dynamic pressure in the dayside. For Tromso, the median slope is ~ 1.5 nT/sec/nPa^1/2. If we go back to the great magnetic disturbance on August 4, 1972, the estimated change of square root of solar wind dynamic pressure during the SSC at 2040 UT was ~ 4 nPa^1/2, then the time rate of change should have been ~ 6 nT/sec at latitudes similar to Tromso's according to our median slope. However, this predicted value is lower than that observed. Observations by Anderson et al.  showed that the time rate of change was ~ 1800 nT/min (30 nT/sec) at Meanook (61.9 MLAT), Canada and ~ 700 nT/min (12 nT/sec) at Plano, Illinois. These higher values are consistent with the large scatter in our observation and may be associated with the large interplanetary velocities of August, 1972 events. Even so, our predicted rate of ~ 6 nT/sec for the August 1972 geomagnetic disturbances would induced an electric field well above 1 V/km, which although smaller than the 7 V/km observed at Plano, Illinois, is large enough to produce damaging voltage excursions. During the magnetic disturbance on February 11, 1958, the induced electric field of 0.5 V/km caused disruption of the transatlantic telecommunications cable [Axe, 1968].
Anderson et al.  speculated that the large magnetic variation on August 4, 1972 was caused by the enhanced magnetopause boundary current when the magnetopause was compressed to 5 RE. According to Tsyganenko's ellipsoidal magnetosphere model [Tsyganenko, 1989], the ground response to the increase of magnetopause current decays rapidly at higher latitudes. The model predicts that the enhanced magnetopause boundary current would only cause an increase of ~ 10 nT in the H-component in the auroral zone for ~ 4 nPa^1/2 change in square root of solar wind dynamic pressure, that is much smaller than observed. Thus, we believe that it is the enhanced ionospheric current system that played the most important role in producing the large magnetic variation.
In summary, our results show that sudden impulses caused by the passage of interplanetary shocks can cause disturbances sufficiently severe to cause disruption of communication and power distribution systems even when the interplanetary magnetic field is northward. Even though the overall energization of the magnetosphere depends on the coupling of the solar wind momentum flux to the magnetosphere through reconnection with a southward interplanetary magnetic field, the global nature of the shock associated disturbances and their predominant disturbance in the north-south direction make these interplanetary shocks very effective in disrupting voltage sensitive networks oriented in the east-west direction as they often are especially in North America and across the North Atlantic. Thus, sudden impulses may have daytime effects equal to or greater than those nighttime effects associated with substorms as measured by their impact on terrestrial systems.
Acknowledgments. This research was supported by the National Science Foundation under research grants ATM90--16900 and ATM91--11913.
Akasofu, S.-I., and S. Chapman, Solar-Terrestrial Physics, pg. 901, Clarendon Press, Oxford, 1072.
Albertson, V. D., and J. M. Thorson, Jr., Power system disturbance during a K-8 geomagnetic storm: August 4, 1972, IEEE Trans Power App. Sys., PAS-93, 1025, 1974.
Anderson, C. W., L. J. Lanzerotti and C. G. Maclennan, Outage of the L-4 system and the geomagnetic disturbances of August 4, 1972, Bell Syst. Techn. J., 53, 1817, 1974.
Araki, T., Global structure of geomagnetic sudden commencements, Planet. Space Sci., 25, 373, 1977.
Axe, G. E., The effect of Earth's magnetism on submarine cables, Post Off. Electr. Eng. J., 61, 37, 1967.
Brooks, J., A reporter at large: A subtle storm, New Yorker Magazine, pg. 34, Feb. 7, 1959.
Burlaga, L. F., and J. H. King, Intense interplanetary magnetic fields observed by geocentric spacecraft during 1963--1975, J. Geophys. Res., 84, 6633, 1979.
Chapman, S. and J. Bartels, Geomagnetism, Vol. II, Oxford University, 1940.
Coles, R. L., News Release, Geophysical Survey of Canada, 6 April, 1989.
Lanzerotti, L. J., Geomagnetic induction effects in ground-based systems, Space Sci. Rev., 34, 347, 1983.
Lanzerotti, L. J., Comment on "Great magnetic storms" by Tsurutani et al., Geophys. Res. Lett., 19, 1991, 1992.
Lanzerotti, L. J., and Gregori, Telluric currents: The natural environment and interactions with man-made systems, in The Earth's Electrical Electrical Environment, pp.232, National Academy Press, Washington, 1986.
Le, G., C. T. Russell, S. M. Petrinec, and M. Ginskey, Effect of sudden solar wind dynamic pressure changes at subauroral latitudes: Change in magnetic field, J. Geophys. Res., 98, 3983-3990, 1993.
Matsushita, S., Geomagnetic disturbances and storms, Physics of Geomagnetic Phenomena, Vol. 2, pg. 793, edited by S. Matsushita and W. H. Campbell, Academic Press, New York, 1967.
Medford, L. V., L. J. Lanzerotti, J. S. Kraus, and C. G. Maclennan, Transatlantic Earth potential variations during the March 1989 magnetic storms, Geophys. Res. Lett., 16, 1145, 1989.
Nishida, A., Geomagnetic Diagnosis of the Magnetosphere, pg. 256, Springer-Verlag, New York, 1978.
Russell, C. T., R. L. McPherron, and R. K. Burton, On the cause of geomagnetic storms, J. Geophys. Res., 79, 1105, 1974
Russell, C. T., M. Ginskey, S. Petrinec, and G. Le, The effect of solar wind dynamic pressure changes on low and mid-latitude magnetic records, Geophys. Res. Lett., 19, 1227, 1992.
Siscoe, G. L., V. Formisano and A. J. Lazarus, A calibration of the magnetopause, J. Geophys. Res., 73, 4869, 1968.
Stuart, W. F., Arrays of magnetometers operated in NW Europe, The IMS Source Book, 141, American Geophysical Union, Washington, D.C., 1982.
Suguire, M., and S. Chapman, The average morphology of geomagnetic storms with sudden commencement, Abh. Akad. Wiss. Gottingen, Math.-Phys. Klasse, Sonderheft Nr. 4, Gottingen, 1960.
Tsyganenko, V., A solution of the Chapman-Ferraro problem for an ellipsoidal magnetopause, Planet. Space Sci., 37, 1037, 1989.
Tsurutani, B. T., W. D. Gonzalez, F. Tang, Y. T. Lee, M. Okuda, and D. Park, Reply to L. J. Lanzerotti: Solar wind ram pressure corrections and an estimation of the efficiency of viscous interaction, Geophys. Res. Lett., 19, 1993, 1992.
Winckler, J. R., L. Peterson, R. Hoffman, and R. Arnoldy, Auroral X-ray, cosmic rays, and related phenomena during the storm of February 10--11, 1958, J. Geophys. Res., 64, 597, 1959.
(Received: July 30, 1992; revised: October 26, 1992; acceptd: November 9, 1992)
Figure 1. An example of sudden impulse event on the dayside. Shown are the response of geomagnetic field and the time rate of change of the magnetic field at York (56.4 MLAT), Lerwick (62.2 MLAT) and Tromso (66.9 MLAT).
Figure 2. Time rate of change in the H-component as a function of local time for York (triangles), Lerwick (squares) and Tromso (circles) for all the sudden impulse events.
Figure 3. Time rate of change in the H-component versus the change in square root of solar wind dynamic pressure for sudden impulse events between local time 0700 and 1700. The lines correspond to median slopes.