which They Provide for Substorm Theories Gordon ROSTOKER

Transcription

1 J. Geomag. Geoelectr., 43, Suppl., , 1991 Auroral Signatures Magnetospheric Substorms and Constraints which They Provide for Substorm Theories Gordon ROSTOKER Institute Earth & Planetary Physics and Department Physics, University Alberta, Edmonton, Alberta, Canada, T6G 2J (Received September 28, 1990; Revised March 29, 1991) Magnetospheric substorms reflect most pronounced periods energy transfer from solar wind to magnetosphere and ionosphere Earth. Modem views substorm process have been substantially modified in past few years by information inferred from analysis and interpretation auroral images acquired by UV imager flown aboard Viking satellite. In this paper, I shall summarize some findings from ongoing study Viking data base dealing with nature auroral surges in evening sector and omega bands in morning sector. These auroral forms reflect growth instabilities in magnetosphere-ionosphere system and provide important information on properties geospace environment in which free energy to drive instabilities is found. Substorms were originally defined using ground based arrays detectors and any ory substorms must account for temporal and spatial properties auroral forms. Any ory that does not address se questions is incomplete. The information presented in this paper refore provides constraints which are necessary to incorporate into any ory magnetospheric substorms. 1. Introduction Much early history substorm physics involved observational studies auroral and electromagnetic phenomenology using eir detectors at single stations or in arrays which were quite inadequate to reflect global behaviour magnetosphere-ionosphere system and its response to changing solar wind environment. Inadequacy data usually is reflected in an ambiguity in interpretation those data which are acquired, and ensuing conflicts which occur in building framework for interpretation data satisfying both local and global constraints is entirely understandable. Three notable conflicts which have occurred due to paucity data are: (1) wher all substorm currents were confined to ionosphere (cf. CHAPMAN, 1935) or wher field-aligned currents linked ionospheric currents to outer magnetosphere (cf. ALFVEN,1939). (2) wher substorm current could be described by a one cell (ef. AKASOFU et al., 1965) or a two cell (cf. SUGTURA and HEFFNER,1963) equivalent current system. This turned out to be a particular phenomenological component much larger question relative roles directly driven activity and loading-unloading in substorm process (cf. ROSToKER et al., 1987). 233

2 234 G. ROSTOKER (3) Wher substorm expansive phase current "wedge" connects to region a near earth neutral line (cf. HONES,1984) or wher it threads plasma sheet boundary layer extending to a source region more distant in tail (cf. ROSTOKER and EASTMAN,1987). Improved arrays instrumentation, both in space and on ground, have helped to resolve first two controversies. The third controversy awaits a yet more comprehensive data base before it can be resolved. The launch Viking satellite in February 1986 opened a new chapter in acquisition data needed to minimize level ambiguity in interpretation global observations substorm phenomenon. The satellite carried a UV imager into an eccentric orbit with apogee near `2RE and an orbital period 262min. The camera was able to record images with aid a CCD detector at sample rate as high as one image every 20s (although one image per minute was more norm for mission). The wealth images (over 40,000 in all) acquired over period Viking mission afforded space researchers opportunity greatly reducing ambiguity interpretation ground based optical detectors and this paper is devoted to documenting some more important pieces knowledge gained from analysis Viking data set. The reader is referred to ANGER et al. (1987) for a more detailed description Viking imager. 2. Working Definition a Magnetospheric Substorm Historically, term substorm was used to describe auroral breakup and subsequent poleward expansion an azimuthally localized regime discrete auroral arcs as recorded by all-sky cameras (cf. AKASOFU,1964). Ultimately signatures this phenomenon obtained using or detectors (e.g. magnetometers, riometers) were also described in terms substorm, leading to all-encompassing name magnetospheric substorm which was applied by AKASOFU (1968) to combined effect various disturbances accompanying auroral breakup. Subsequently, satellite investigators chose to use AE index to characterize state magnetosphere at time ir in situ observations. Figure 1 shows rise and fall AE over interval UT which would typically be called a substorm by those using index (rar than data from individual ground based detectors) to define that phenomenon. However, following on studies ROSTOKER (1969) and AKASOFU (1981) amongst ors, it was becoming clear that magnetic perturbations on which AE index was based stemmed from two characteristically different current systems. In modern times se have been called directly driven system (cf. AKASOFU, 1981) and substorm current wedge (cf. BAUMJOFIANN,1983). Directly driven activity is characterized by two large scale electrojets-an eastward electrojet across dusk sector and a westward electrojet across dawn sector as shown in Fig. 2(a). CLAUER et al. (1983) have demonstrated, using linear prediction filtering techniques, that at least 40% AL disturbance can be accounted for by effects directly driven activity. They also showed that impulse response time magnetosphere as characterized by directly driven activity was N2 hr. In contrast, substorm current wedge, which is viewed as a manifestation unloading stared magnetotail energy, is dominated by an intense westward electrojet which develops poleward eastward electrojet in evening sector. Although substorm current wedge is ten viewed as a large scale current feature, it actually represents combined effect an ensemble small scale current elements which develop in rapid succession during

3 Auroral Signatures Magnetospheric Substorms and Constraints 235 Fig, 1. Plot auroral indices AU, AL and AE (=AU+ AL ) during UT day August 17, 1978, Episodes what some call substorm activity are seen over intervals and , There is no particular threshold for perturbations above which y are considered to be substarms Fig. 2. Magnetospheric substorms involve co-existence two characteristically different current systems. Directly driven activity (Panel a) involves large scale eastward and westward electrojets flowing across dusk and dawn meridians respectively. These current systems vary on time scale impulse response time magnetosphere (viz. `2hr). Expansive phase activity is thought to reflect release stored magnetotail energy and features an ensemble co-existing regimes upward field aligned current (FAC) toger with spatially localized electrojet elements connecting upward FAC with diffusely spread downward FAC. One such element is shown in Panel b, toger with its characteristic response time (which is considerably shorter than impulse response time magnetosphere characterizing directly driven activity).

4 236 G. ROSTOKER substorm expansive phase (cf. KOROTOVA, et al., 1992). Each se small scale wedge elements (cf. Fig. 2(b)) features an intense upward field-aligned current (FAC) at its western edge, with downward return current flowing diffusely into ionosphere to east and west. The western return current tends to be dominant and this connecting westward ionospheric current is part expansive phase westward electrojet. Since many small scale current wedges can coexist if y develop in rapid succession, one may have several upward FAC regions and hence several surges (which are auroral signature azimuthally localized upward FAC) may coexist as we shall discuss later in this paper. The characteristic lifetime (and hence impulse response) each substorm wedge is `10min, significantly shorter than impulse response time associated with directly driven system (KOROTOVA et al., 1992). This suggests that wedge elements have a significantly smaller scale size than directly driven system (when one considers self inductances two current systems). Finally, we note that it is wedge elements that correspond to auroral features that AKASOFU (1964) termed original substorm. Nowadays, term substorm is considered to represent combined effect directly driven process and storage-release process (that leads to current wedge formation). For event shown in Fig. 1, interval from UT involves purely driven activity, while interval UT involves significant current wedge development in addition to contribution driven system (cf. ROSTOKER (1983) for details this particular event). The reader is referred to ROSTOKER et al. (1987) for furr details regarding properties directly driven and storage release processes. 3. Auroral Features Expansive Phase Activity Based on description substorm in previous section, one gains an impression auroral oval behaviour as shown in Fig. 3. The expansion and contraction oval (accompanied by concomitant increase and decrease driven system electrojets) follows closely increase and decrease energy flow from solar wind into magnetosphere. Surge (and hence current wedge) activity can be present for any polarity interplanetary magnetic field or level energy flow into magnetosphere, however it tends to maximize during large scale expansive phase outburst which ten accompanies poleward retreat poleward border oval. Viking has taught us a great deal about behaviour oval during expansive phase activity. Some important points to note are: (1) Contrary to statements on earlier literature, surges do not travel westward. (Indeed, terms westward traveling surge is a misnomer.) The westward edge a surge form may expand slightly westward as surge form grows in spatial extent after onset. The region expansive phase disturbance also expands westward in early stages a substorm expansive phase. However it does so irregularly and through formation new surge forms. Once formed and having reached ir maximum size, surges remain more or less in place (ROSTOKER et al., 1987). Figure 4 presents a sequence Viking images showing typical behaviour a surge over its lifetime. One can see that western edge can expand westward over a short distance during early stages its formation but furr development can even involve an eastward contraction. Figure 5 shows maximum westward expansion western edge surges for a sample 67 such events examined in detail by KIDD and ROSTOKER (1991). The maximum expansion for any events was ` km, which represents a maximum about 0.5time zones at `67 N geomagnetic.

5 Auroral Signatures Magnetospheric Substorms and Constraints 237 Fig. 3. Evolution auroral oval during an episode enhanced energy input from solar wind (ten created by enhanced southward interplanetary magnetic field). Some most spectacular episodes expansive phase activity (featuring a multiplicity aurora) surge, Pi 2 pulsation and short-lived magnetic bay activity) accompany return IMF to a less southward configuration after a significant episode enhanced southward IMF. The expansive phase auroral surges ten exhibit significant multiplicity along poleward edge oval as it contracts polewi rd following reduction energy flow into magnetosphere. (2) Historically, substorm is pictured as having one surge (AKASOFU, 1964) and one associated current wedge. This, in fact is not case. KISABETI I and ROSTOKER (1974) demonstrated that current wedge develops through appearance a sequence filamentary westward jets, each appearing poleward previous one. Each electrojet element is likely to have a surge at its western edge, suggesting that large events have a multiplicity surges. Figure 6 shows evolution a large magnetospheric substorm expansive phase, with actual onset being detected by Viking UV images. The poleward border disturbed region expands poleward over almost 10 degrees latitude as expansive phase proceeds. This event occurred over an interval northward turning IMF, and multiple surge activity along poleward edge disturbed oval later in event is a real example what is shown in cartoon version substorm auroral activity presented in Fig. 3. Figure 7 shows three images taken by Viking `3min apart. The surge multiplicity in middle frame shows unambiguously that substorm physics cannot be understood in context production a single surge form. This particular event is also interesting as lifetime easternmost surge is ƒ2min. This suggests that growth times for substorm surges can be order a minute (or less, perhaps). Figure 8 shows statistics surge multiplicity for a large subset Viking imager data. While single

6 238 Fig. G. ROSTOKER 4. Characteristic activity. MLT behaviour The images meridian 21.9 surge expands manifestation Viking significant were imager range in a typical on (shown briefly taken on all sense that 3, frames). westward dimming surge May before it auroral demonstrates which grows The During its This little decays develops life-time contracting form). and surge in an its `15min, back to surge is typical during with way episode east westward surges substorm edge western (which western near edge may just detected propagation be by over a a longitudes. Fig. 6. Evolution a magnetospheric substorm showing longitudinal confinement initial expansive phase onset (top left and top right), initial expansion disturbed region, which can sometimes be eastward (bottom left) and multiplicity surge activity at height expansive phase (bottom right). The magnetic local times bracketing initial brightening are noted in bottom right panel, with same meridians appearing on or panels for sake reference. This event took place on April 1, 1986 and times frames are 1847:54 UT (top left), 1850:55 UT (top right), 1853:36 UT (bottom left) and 1913:42 UT (bottom right).

7 Auroral Signatures Magnetospheric Substorms and Constraints 239 Fig. 7. Four consecutive Viking images showing rapid evolution surge multiplicity. The images are taken on September 27, 1986 from 1158: :57 UT in order top left, top right, bottom left and bottom right. Clearly simultaneous occurrence three surge forms in top right panel is an observational constraint which any substorm ory must address. The rapid disappearance most easterly surge form less than 2minutes after its appearance is a commentary on lifetime surges. Fig. 10. Viking image taken at 0112:10 UT on May 3, 1986 at time an expansive phase onset in evening sector. The brightening equatorward portion oval in evening sector accompanies expansive phase onset, but substorm westward electrojet is at high latitude close to Poste-de-la-Baleine (marked by white dot on east coast Hudson Bay). Omega bands arc clearly evident near equatorward edge morning sector aurora] oval. These omega bands (accompanied by Ps 6 magnetic variations) appear to be a near Earth Central Plasma Sheet phenomenon.

8 240 G. ROSTOKER Fig. 5. Plot occurrence frequency surges indicating length time y were tracked and distance that ir western edge expanded westward, which is at most a half a time zone at typical aurora! oval latitudes (after KIDD and ROSTOKER, 1991). Fig. 8. Surge multiplicity for a large subset Viking UV imager data. While highest probability is that a single surge will be seen, single surge events clearly constitute considerably less than half cases in which at least one surge is seen (after KIDD and ROSTOKER, 1991). surges represent most probable situation for substorm activity, if one surge is observed, n more than one surge will co-exist for 50% time. Figure 9, however, provides us with surprising result that any hour long segment evening sector features surge activity ƒ6% time, While occurrence frequency might be somewhat larger since an absence a surge in (say) MLT sector might coincide with a surge in some or local time sector, surge multiplicity implies that a surge in one MLT sector could be coexistent with a surge in anor MLT sector. For criteria chosen for surge selection by KIDD and ROSTOKER (1991) (from which Fig. 9 is abstracted), surges appear to be a rar

9 Auroral Signatures Magnetospheric Substorms and Constraints 241 Fig. 9. Percentage time that surges are seen in each hour long local time sector evening sector auroral oval. It would appear that surges are a surprisingly uncommon feature aurora! oval. It bears mentioning that se data are taken at sunspot minimum with average Ap for days involved being 12 (after KIDD and ROSTOKER, 1991). uncommon phenomenon. This implies that it is driven system which is major component substorm activity, in accord with judgement AKASOFU (19$1). (3) Every so ten, one obtains a set data which serves to answer several outstanding questions at same time. The sequence images obtained by Viking on orbit 386 taken on May 3, 1986 provided just such a data set which is now being explored in detail as part Coordinated Data Analysis Workshop CRAW 9. Figure 10 shows one images this sequence, taken at time onset a clear episode substorm expansive phase activity. What this one image taught us was: (i) Aurora! omega bands, visible across morning sector from North Atlantic through Scandinavia into U.S.S.R., occur in equatorward part auroral oval which is site topside ionosphere cps and Region 2 Birkeland currents. (ii) If auroral structures with longitudinal periodicity can be attributed to a Kelvin- Helmholtz instability along a velocity shear zone, n one may have more than one velocity shear zone capable producing such structures at any given local time. This is evident from periodic structures near 72 N and 62 N visible across Magnetic Local Time range (iii) Not all auroral brightenings can be associated with expansive phase arcs in process breaking up and, at least in UV spectrum, auroral brightenings deep in cps can appear more spectacular than breakup arcs mselves. In evening sector Fig. 9 extensive area in red near equatorward edge oval had intensified at same time as arcs located at same Magnetic Local Time near poleward edge oval. In this case, two ground magnetometer stations (Ottawa and Poste-de-la-Baleine) were located near equatorward and poleward edges oval respectively, as indicated by white dots on figure. The magnetometer records clearly indicated that substorm westward electrojet was located much closer to high latitude station a Poste-de-la-Baleine than to lower latitude station Ottawa notwithstanding fact that intense UV emissions near equatorward edge oval were much closer to Ottawa than to Post-de-la-Baleine. From this, we can conclude that expansive phase westward electrojet did not lie in region most intense UV emissions, but rar in less spectacular region discrete arcs near poleward border oval.

10 242 G. ROSTOKER 4. Concluding Comments In this paper, I have tried to clarify nature substorm activity as a preface to presenting some interesting new facts we have learned about auroral signatures expansive phase activity over past few years. One cannot properly attempt to understand and model substorm phenomenon without attempting to provide explanations both driven process and storage-release process. Anyone who attempts to model one aspect substorm while ignoring constraints imposed by existence or component activity does so at ir peril. For example, surges we have described in this paper originate at or in near vicinity Harang discontinuity (i.e. region meridional gradient in north-south component ionospheric electric field within which a polarity transition in that electric field is found). However, it is across this same Harang discontinuity that Region 1 currents associated with directly driven system flow prior to and during substorm expansive phase activity. This strongly suggests that surge activity (which has, in past, been intimately associated with unloading magnetotail energy) represents a perturbation directly driven system and, more specifically, a spatially localized perturbation large scale Region 1 currents. If one is to understand physical mechanism in outer magnetosphere which leads to surge formation, one is going to have to explain why individual surge forms do not move west-ward after formation to any great extent. It is true that western edge a surge can expand rapidly westward for a brief period after formation, however this represents an increase in scale size surge form rar than a westward motion moment luminosity distribution. While a Kelvin-Helmholtz instability in region velocity shear which defines Harang discontinuity seems attractive (cf. ROSTOKER, 1987), near stationary surge forms demand that momentum densities on eir side reversal in direction plasma E ~B drift velocity be approximately equal in magnitude. This would appear to be a rar restrictive condition. However, it is possible that momentum densities are unequal, but inequality is so small as to render phase velocity waves resulting from Kelvin-Helmholtz instability to be small enough that, when mapped into ionosphere, motion resultant surge forms is not detectable for most cases. The multiplicity surge forms is a crucial clue to understanding substorm expansive phase process. In extreme cases, surges can be observed to co-exist over several time zones in evening sector. This indicates that, whatever mechanism is for substorm expansive phase activity, it must allow for simultaneous existence up to several longitudinally localized regions upward field-aligned current. Some surges can appear to east pre-existing surge activity (cf. PYTTE et al., 1976; ROSTOKER et al., 1987) which means that magnetotail regime which is source region for surge currents must be conducive to surge formation at any time during an episode expansive phase activity. This leads one to question wher major topological reconfigurations magnetotail occur at onset expansive phase activity when activated region can recover to extent being able to initiate fresh surge activity back towards midnight before expansive phase episode has come to an end. Canada. This research was supported by Natural Sciences and Engineering Research Council

11 Auroral Signatures Magnetospheric Substorms and Constraints 243 REFERENCES AKASOFU, S.-I., The development aurora) substorm, Planet, Space Sci., 12, , AKASOFU, S.-I., Polar and Magnetospheric substorms, D. Reidel Publ. Co., Dordrecht, Holland, AKASOFU, S.-I., Energy coupling between solar wind and magnetosphere, Space Sci. Rev., 28, , AKASOFU, S.-I., S. CHAPMAN, and C.-I. MENG, The polar electrojet, J. Atmos. Terr, Phys., 30, 227, ALFVEN, H., A ory magnetic storms and aurorae, Kungl. Sv. Vetensk.-Akad. Handlingar III, 18, No.3, Stockholm, ANGER, C.D., S.K. BABEY, A.L. BROADFOOT, R.G. BROWN, L.L. COGGER, R. GATTIHGER, J.W. HASLETT, R. A. KING, D.J. MCEWEN, J.S. MURPHREE, E.H. RICHARDSON, B.R. SANDEL, K. SMITH, and A. VALLANCE JONES, An ultraviolet auroral imager for Viking spacecraft, Geophys. Res. Lett.,14, , BAUMJOHANN, W., Ionospheric and field-aligned current systems in auroral zone: a concise review, Advances in Space Research, 2, 55-62, CHAPMAN, S., The electric current systems magnetic storms, Terrest. Magnetism and Atmos. Elec., 40, 349, CLAUER, C.R., R.L. MCPHERRON, and C. SEARLS, Solar wind control low-latitude asymmetric magnetic disturbance field, J. Geophys. Res., 88, , HONES, E.W., Jr., Plasma sheet behaviour during substorms, in Magnetic Reconnection in Space and Laboratory Plasmas, edited by E.W. Hones, Jr., p.178, Amer. Geophysical, Geophys. Union, Monograph 30, Washington, D.C., KIDD, S.R, and G. ROSTOKER, Distribution auroral surges in evening sector, J. Geophys. Res., 96, , KISABETH, J.L. and G. ROSTOKER, The expansive phase magnetospheric substorms 1. Development auroral electrojets and aworal arc configuration during a substorm, J. Geophys. Res., 79, , KOROTOVA, G., G. ROSTOKER, and M. CONNORS, Evolution component current systems expansive phase a magnetospheric substorm, manuscript in preparation, PYTTE, T., R.L. MCPHERRON, and S. KOKUBUN, The ground signatures expansion phase during multiple onset substorms, Planet. Space Sci., 24, , ROSTOKER, G., Classification polar magnetic disturbances, J Geophys. Res., 74, , ROSTOKER, G., Triggering expansive phase intensifications magnetospheric substorms by northward turnings interplanetary magnetic field, J. Geophys. Res., 88, , ROSTOKER, G., The Kelvin-Helmholtz instability and its role in generation electric currents associated with Ps 6 and westward travelling surges, in Magnetotail Physics, edited by A.T.Y. Lui, pp , Johns Hopkins Univ. Press, Baltimore, ROSTOKER, G. and T.E. EASTMAN, A boundary layer model for magnetospheric substorms, J Geopliys. Res., 92, , ROSTOKER, G., A. VALLANCE JONES, R.L. GATTINGER, C.D. ANGER and J.S. MURPHREE, The development substorm expansive phase: "eye" substorm, Geophys. Res. Lett., 14, , ROSTOKER, G., S.-I. AKASOFU, W. BAUMJOHANN, Y. KAMIDE, and R.L. MCPIIERRON, The roles direct input energy from solar wind and unloading stored magnetotail energy in driving magnetospheric substorms, Space Sci. Rev., 46, , SUGIURA, M. and J.P. HEPPNER, The earth's magnetic field, in Introduction to Space Science, edited by W. N. Hess, p.45, Gordon and Breach, New York, 1965.