JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 117, A10223, doi:10.1029/2012JA018046, 2012
Cluster multipoint study of the acceleration potential pattern and electrodynamics of an auroral surge and its associated horn arc G. T. Marklund,1 S. Sadeghi,1 Bin Li,1 O. Amm,2 J. A. Cumnock,1,3 Y. Zhang,4 H. Nilsson,5 A. Masson,6 T. Karlsson,1 P.-A. Lindqvist,1 A. Fazakerley,7 E. Lucek,8 and J. Pickett9 Received 20 June 2012; revised 29 August 2012; accepted 10 September 2012; published 19 October 2012.
[1] Cluster results are presented from the acceleration region of an auroral surge and connected horn arc, observed during an extended time period of substorm activity. The Cluster spacecraft crossed different magnetic local time (MLT) sectors of the surge and horn, with lag times of 2–10 min. Acceleration potential patterns are derived for the horn arc and for the double arc (surge and horn) at the surge front and deeper into the surge. The parallel potential drop of the horn arc ranged between 4 and 7 kV. At the surge front, two weakly coupled U-potentials with parallel potential drops of 8 (7) kV and 7 (5) kV were derived for the surge and horn, respectively, from the C3 (C4) data. A similar, more coupled pattern was derived for the region deeper into the surge. We also address how the field-aligned currents of the surge and horn system close in the ionosphere. The Cluster data allow almost simultaneous estimates of the latitudinal current closure at various MLT sectors. Significant net upward currents are derived for the horn and surge, whereas the currents at the surge front were found to be balanced. The net upward horn current is proposed to be fed by the zonal divergence of the westward Pedersen current in the horn, consistent with the acceleration potential decrease in the westward horn direction. The net upward surge current is proposed to be fed by the divergence of a westward electrojet and by localized downward currents adjacent to the surge. Citation: Marklund, G. T., et al. (2012), Cluster multipoint study of the acceleration potential pattern and electrodynamics of an auroral surge and its associated horn arc, J. Geophys. Res., 117, A10223, doi:10.1029/2012JA018046.
1. Introduction [2] The auroral emissions are caused by high-energy beams of electrons hitting the upper atmosphere, after being accelerated by quasi-static electric fields, with a component parallel to the geomagnetic field, and formed at altitudes around one Earth radius, or by wave electric fields. The
1 Space & Plasma Physics, School of Electrical Engineering, KTH, Stockholm, Sweden. 2 Finnish Meteorological Institute, Helsinki, Finland. 3 Center for Space Sciences, University of Texas at Dallas, Richardson, Texas, USA. 4 Johns Hopkins University Applied Physics Laboratory, Laurel, Maryland, USA. 5 Swedish Institute of Space Physics, Kiruna, Sweden. 6 ESA/ESTEC, Noordwijk, Netherlands. 7 MSSL, University College, Holmbury St. Mary, UK. 8 Space and Atmospheric Physics Group, Blacket Laboratory, Imperial College London, London, UK. 9 Department of Physics and Astronomy, University of Iowa, Iowa City, Iowa, USA.
Corresponding author: G. T. Marklund, Space & Plasma Physics, School of Electrical Engineering, KTH, Stockholm SE-10044, Sweden. (
[email protected]) ©2012. American Geophysical Union. All Rights Reserved. 0148-0227/12/2012JA018046
quasi-static parallel electric fields and potentials are formed mainly as a result of charge neutrality and current continuity requirements in the high-altitude, dilute plasma above the aurora. This region is commonly referred to as the auroral acceleration region (AAR), and is typically located between 4000 and 12000 km above the polar atmosphere [Paschmann et al., 2003]. The term quasi-static is used to indicate that the structures are stable on a time scale long compared to the transit time of the charged particles. Structures of this kind form and accelerate charged particles, producing aurora and plasma outflow, not only around Earth but around other solar system planets, such as Jupiter and Saturn, illustrating that the formation and maintenance of such structures are fundamental and ubiquitous processes in space plasmas. [3] That aurora is produced by electric fields, with a component parallel to the Earth’s magnetic field (referred to as parallel electric fields), accelerating particles toward the Earth’s atmosphere was first suggested by Alfvén [1958]. Since then it has been confirmed experimentally by numerous spacecraft and rocket measurements. The parallel electric fields occur together with converging or diverging electric fields, perpendicular to the magnetic field, in U-shaped potential structures [Carlqvist and Boström, 1970; Mozer et al., 1980], or with monopolar electric fields in S-shaped potential structures [Mizera et al., 1982; Marklund et al.,
A10223
1 of 15
A10223
MARKLUND ET AL.: MULTI-POINT STUDY OF AURORAL SURGE
A10223
Figure 1. Different sectors of generic aurora [after Fujii et al., 1994]. The part of the generic aurora (surge horn, surge front and surge) explored by Cluster in this study is indicated by the rectangle. The positive and negative values correspond to net upward and downward R1 FAC in the pre- and post-midnight local time sector, respectively.
1997]. Negatively charged potential structures form in the upward current region. These are associated with upward pointing electric fields, accelerating electrons downward, producing intense displays of aurora, and energetic ion beams moving upward, away from Earth. Downward parallel electric fields associated with positive potential structures may develop in the adjacent downward current region [Marklund et al., 1994, 1997, 2001], accelerating electrons away from, and ions toward Earth, to energies ranging up to a maximum of a few thousand electron volts. This acceleration region is closer to Earth, typically located between 1000 km and 4000 km [Marklund et al., 1997]. [4] A number of mechanisms have been proposed for maintaining the parallel electric fields, such as strong double layers [Block, 1972], weak double layers [Temerin et al., 1982], Alfvén waves [Song and Lysak, 2001], magnetic mirror supported fields [Alfvén and Fälthammar, 1963; Knight, 1973; Chiu and Schultz, 1978] and anomalous resistivity [Hudson and Mozer, 1978]. Experimental evidence of parallel electric fields has been presented from both sounding rockets [McIlwain, 1960; Mozer and Bruston, 1967] and satellite missions, such as Polar [Mozer and Kletzing, 1998] and FAST [Ergun et al., 2000, 2002; Andersson et al., 2002], including the first experimental verification of strong double layers. [5] There are many unresolved issues concerning the parallel electric fields and potentials apart from those related to the above listed mechanisms, such as their altitude distribution and stability in space and time. This is not possible to determine experimentally from single-satellite observations but require simultaneous measurements by multiple spacecraft at different altitudes of the acceleration region. This became an opportunity in late 2008 when the Cluster perigee was lowered such that direct crossings of the AAR became a reality. First results from event studies addressing the altitude distribution and stability of the parallel potential structures, and how these interact with Alfvénic processes, such as within the polar cap boundary region of Inverted-V aurora
were presented in two papers by Marklund et al. [2011a, 2011b]. [6] How different arcs, such as the Inverted-V arcs produced by quasi-static potential structures, interact and connect to each other when being part of a large-scale auroral system extended over many MLT hours, represents another interesting topic that can be addressed by Cluster data. On this topic, one much debated issue is how the field-aligned currents at different MLT of the surge close in the ionosphere, which has been addressed by e.g., Hoffman et al. [1994] and Fujii et al. [1994], who used imager data provided by the Dynamics Explorer 1 satellite to identify substorm events and simultaneous in situ data of fields and particles measured by the Dynamics Explorer 2 satellite crossing different sectors of the aurora associated with substorms. The DE-2 data were binned into a number of different sectors of the generic aurora according to Figure 1. In this picture we have added the surge front, which was probed by the Cluster 3 and 4 spacecraft in this study. The percentage values given for each sector are the average net currents normalized to the R1 current intensities, derived from the magnetometer data based on several DE-2 satellite crossings within each sector. The values are thus average values, based on a large number of different substorms, having a standard deviation of +/10%. The present study benefits from using almost simultaneous Cluster multispacecraft data from several of the above sectors of the generic aurora (such as the surge horn and front, and surge), enabling close in time estimates of the net currents for one given event which is done here for the first time. [7] The Cluster data presented here were obtained by the EFW electric field instrument [Gustafsson et al., 1997], the FGM fluxgate magnetometer [Balogh et al., 1997], the PEACE electron instrument [Johnstone et al., 1997] and the CIS ion instrument [Réme et al., 1997]. In this study data are presented from Cluster 1, 2, and 3. Given the similar orbits and short lag time between the C3 and C4 spacecraft, the data from C3 and C4 are very similar in terms of electron and magnetic field data,
2 of 15
A10223
MARKLUND ET AL.: MULTI-POINT STUDY OF AURORAL SURGE
A10223
Figure 2. The southern hemisphere auroral oval, as viewed by the DMSP F16 satellite between 21:23 and 21:43 UT. F16 crossed the western horn roughly ten min after the Cluster surge and horn crossings. The red dashed line indicates the equatorward boundary of the aurora. The direction of the F16 spacecraft is indicated by the white arrow. An expanded view of the surge and horn region is shown to the right, together with the projected spacecraft trajectories.
but since the ion and electric field data are of lower quality as compared to C3, we decided not to present the C4 data here.
2. Event 31 May 2009 Description [8] The data presented here are from an equatorward crossing by the Cluster quartet above the southern auroral oval in the 21–22 MLT sector, between 21:05 and 21:35 UT on 31 May 2009. The Cluster quartet traversed the upper AAR of a surge and associated horn. The four spacecraft were spread in the east-west direction enabling almost simultaneous measurements at different MLT sectors of the surge and horn. The auroral distribution was monitored by the DMSP F16 imager shortly after the Cluster crossings (see Figure 2, left). The F16 image is composed from successive scans taken along the orbit. As indicated in Figure 2 (left), the F16 spacecraft entered the dawnside oval around 21:26 UT and exited the eveningside oval at around 21:40 UT. An expanded view of the surge, surge front and surge horn is shown to the right where the ionospheric projections of the Cluster and DMSP F16 spacecraft have been overlaid on the image. It can be seen that C2 crossed above the horn, C3 and C4 above the surge front, and C1 above the surge. The DMSP F16 satellite crossed the western part of the horn, about ten minutes after the Cluster crossings. Note that C3 and C4 have almost identical trajectories, separated in time by roughly two minutes. The geomagnetic activity indices AE and AO between 11 and 24 UT on 31 May 2009 are shown in Figure 3. The AE index for the time of the Cluster crossings is seen to be surprisingly low, given the active auroral distribution monitored by the DMSP F16 spacecraft, most likely due to a poor coverage of AE stations within the substorm MLT sector. Since the image is a good indicator of the prevailing substorm activity, a search was made to check
whether significant geomagnetic activity was monitored by any magnetometer station close to the footprint of the Cluster spacecraft for this time period. From field line mapping of the position of the Cluster quartet, the closest magnetometer station was found to be Mawson station in eastern Antarctica. The Mawson magnetogram presented in Figure 4 shows moderate but clear substorm activity between 1900 UT and up to at least 24:00 UT. This long period of activity agrees well with the presence of large-scale auroral activity over several hours MLT and up to 10 degrees of latitude width, during several hours UT, as monitored by the DMSP F17 and F13 satellites at 19:59 UT and 20:40 UT, respectively (images not shown here), as well as by DMSP F16 at 21:33 UT, shown in Figure 2. As can be seen in Figure 4, several activations took place during this time period, one around 20:30, another shortly after 21:00, and the strongest at 22:00 UT. The second of these is likely to be related to the surge and horn formation seen in Figure 2.
Figure 3. AE and AO indices for 31 May 2009 between 11 and 24 UT. The Cluster event between 21:05 and 21:35 UT is a time period which, judging from the AE index, was very quiet.
3 of 15
A10223
MARKLUND ET AL.: MULTI-POINT STUDY OF AURORAL SURGE
A10223
Figure 4. Mawson station magnetogram between 16:00 and 24:00 UT on 31 May 2009. This station was closest to the footprint of the Cluster spacecraft for the time period of interest here. In contrast to the low AE index, which clearly was not sensitive to the large-scale high-latitude activity for this event, significant geomagnetic activity occurred in the region explored by the Cluster spacecraft, for a rather extended time period between 18:30 and 00:00 UT. [9] Data obtained above the horn arc within the 19–20 MLT sectors are available from crossings by three DMSP spacecraft, F17, F13, and F16, some ten minutes before and after the Cluster crossings. The electron, ion and magnetic field data are presented in Figure 5. The panels show from top to bottom, the
electron and ion energy flux, the average electron and ion energy, time-energy spectrograms of electrons and ions, and the eastward (red line) and northward (blue line) magnetic field components measured by the DMSP F17, F13, and F16 spacecraft. The R1/R2 system was traversed by F17 around
Figure 5. Energy flux (ions and electrons), average energy (ions and electrons), and time-energy spectrograms of electrons and ions, eastward and northward magnetic field components as measured by the F17, F13, and F16 spacecraft from crossings above the southern oval within the 19 to 20 MLT sector. The R1/R2 crossings by F17, F13, and F16 took place around 20:46 UT, 21:01 UT, and 21:40 UT, respectively. Note that the R1/R2 current sheets were fairly stable in width and magnitude over this 54 min time period during which the Cluster crossings took place. Also the electron distributions, associated with the double arc structure in R1, are seen to be fairly stable during this time period. 4 of 15
A10223
MARKLUND ET AL.: MULTI-POINT STUDY OF AURORAL SURGE
A10223
Figure 6a. Overview of the electric and magnetic field data obtained by C1–C4 between 21:05 and 21:35 UT on 31 May 2009. From the top are shown: the equatorward electric field component measured by C1, C4, C3, and C2 (first through fourth plots); the eastward magnetic field component (fifth plot); the field aligned current (sixth plot); the electric potential (seventh plot); the negative of the spacecraft potential (eighth plot); and the geocentric distance in RE (ninth plot). 20:46 UT at 19 MLT, by F13 around 21:00 UT at 19.3 MLT, and by F16 around 21:40 UT at 20 MLT. Note that the R1/R2 current sheets are relatively stable both in width and magnitude over this 50 min time period, during which the Cluster crossings took place (21:10 UT–21:33 UT). Two arc structures are clearly visible in R1 by all three spacecraft and the average and peak energies of the precipitating electrons remain stable at values of about 2 keV and 3–4 keV, respectively. The western
part of the horn thus appears to have been fairly stable for almost one hour, during which the Cluster crossings took place.
3. Cluster Observations [10] Figure 6a presents an overview of the electric and magnetic field data obtained by the Cluster spacecraft between 21:05 and 21:35 UT. The panels show, from top to
5 of 15
A10223
MARKLUND ET AL.: MULTI-POINT STUDY OF AURORAL SURGE
A10223
Figure 6b. E? vectors in light blue color and B? vectors in black, red, and green colors for C1, C2, and C3, respectively, plotted along the ionospheric projections of the C2, C3/C4, and C1 orbits on top of a schematic of the DMSP F16 image obtained around 21:33 UT, shown at the bottom right. C1 passed above the central surge, C2 above the surge horn, and C3/C4 above the surge front. Intense converging electric fields are seen as the spacecraft cross the arc structures.
bottom, the normal component of the electric field, En, (with respect to the large-scale current sheet), being roughly the equatorward electric field component on C1–C4; the tangential magnetic field component, Bt, being roughly the eastward magnetic field component, after subtraction of the background magnetic field data (using a polynomial fit to the measured magnetic field); the field-aligned current (FAC), calculated from the gradient of the eastward magnetic field component, the electric potential, derived by integrating the perpendicular (to B) electric field along the satellite orbit; the negative of the spacecraft potential, indicative of relative plasma density variations; and the geocentric distance of the Cluster spacecraft in units of RE. The large-scale R1 and R2 FACs (Figure 6a, sixth panel) are identified by the negative and positive gradients of the eastward magnetic field component (Figure 6a, fifth panel). Within the R1 FAC, electric potential wells are seen (Figure 6a, seventh panel) as well as irregular and low, negative values of the –spacecraft potential, indicative of plasma density cavities. The estimates of the potential drops below the spacecraft were done essentially by eye inspection, by fitting a curve to the ambient potential in the
polar cap and sub-auroral regions, respectively, given that the variation of the fitted ambient potential, over the perpendicular scale size of the structures, is small. The estimates of the potential wells using this approach will be more accurate, the smaller the structures are. For the large-scale potential wells in this study, the relative uncertainties are estimated to range between 20% and 30%. Figure 6b shows vector plots of the perpendicular electric field (light green) and magnetic field in different colors along the C1 (black), C2 (red), and C3 (green) trajectories. Figure 7 summarizes the C1 observations for the same time interval as in Figure 6a. The C1 orbit is seen from Figure 6b to have passed close to the surge. (1) Time-energy spectrograms of upgoing electrons; (2) downgoing electrons; (3) upgoing ions; (4) the equatorward electric field component; (5) the eastward magnetic field component, after subtraction of a polynomial fit to the measured magnetic field; (6) the field-aligned current (FAC), calculated as described above; (7) the electric potential, from integrating the perpendicular (to B) electric field along the satellite orbit; (8), the negative of the spacecraft potential; and (9) the plasma density, calibrated using Whisper data, showing the Auroral Density Cavity
6 of 15
A10223
MARKLUND ET AL.: MULTI-POINT STUDY OF AURORAL SURGE
A10223
Figure 7. C1 data obtained between 21:05 and 21:35 UT on 31 May 2009. The panels show (from the top) time-energy spectrograms of upgoing electrons, downgoing electrons, upgoing ions, the equatorward electric field component, the eastward magnetic field component, the field-aligned current, the electric potential along the C1 trajectory, the negative of the spacecraft potential, and the plasma density. For more information on the data presentation, see the text.
(ADC). Invariant latitude (ilat), Magnetic Local Time (MLT) and geocentric distance expressed in units of RE are shown at the bottom. C1 crossed a narrow region of downward current, which is interpreted as a R0 current [Fujii et al., 1994; Gjerloev and Hoffmann, 2002], before entering the poleward boundary of the surge. The poleward boundary of the aurora
close to the surge front can be seen from Figure 6b to be roughly S-shaped. We associate the poleward and equatorward parts of R1 with the surge and the eastern horn, respectively. C1 crossed the upward FAC region, between 21:10:30 UT and 21:19 UT, in the upper part of the acceleration region of the surge and horn, as indicated by the high peak energies of
7 of 15
A10223
MARKLUND ET AL.: MULTI-POINT STUDY OF AURORAL SURGE
the upgoing ions and relatively low energies of the downgoing electrons. The local FAC density of the upward current varies between 0.02 and 0.1 mA/m2 corresponding to 1 and 5 mA/m2 mapped to the ionosphere. The peak energies of the downgoing electrons and upgoing ions are used as estimates of the parallel potential drops, DFll, above and below the spacecraft orbit. The acceleration potentials above C1 are estimated to be 1 kV and 1.5 kV, and those below C1 are estimated to be 6 kV and 4 kV, for the surge and horn, respectively. Another estimate of the parallel potential drop below C1 is from the depth of the electric potential well, associated with the large-scale converging electric field. The potential wells are about 6 kV and 4 kV for the surge and horn, respectively, consistent with the peak energies of the upgoing ions. The plasma density (ninth panel) shows a broad cavity which is colocated with the potential well, deep over the surge, with a density of a few times 102 cm3, and less deep over the horn, where the density is about 101 cm3. In terms of the total acceleration potential, 6 of the 7 kV in the surge and 4 of the 5.5 kV in the horn, were located below C1; hence C1 crossed the upper part of the AAR of the surge and horn. [11] Figure 8 summarizes the particle and field data obtained by C2 for the same time interval. The C2 orbit is seen to pass above the horn arc, one MLT hour to the west of its connection to the surge. The format is the same as in Figure 7, except that for C2 there are no ion data available, so panels 2 and 3 here show instead time energy spectrograms of the perpendicular (to B) electrons, and of the downgoing electrons. The crossing above the horn arc takes place between 21:18 UT and 21:23:30 UT. The acceleration potential above C2 is quite weak, as indicated by the peak energies of the downgoing electrons of about 0.5 keV or less. The electric field data reveal a large-scale converging electric field structure, which is also evident in the vector plot in Figure 6b, and which is associated with a parallel potential drop of about 4 kV below C2. In the central parts of the arc, the electric field is weaker and directed mainly westward. The constant slope in the eastward magnetic field component implies a fairly constant upward FAC of about 0.03 mA/m2, corresponding to 1.3 mA/m2 in the ionosphere. The density cavity, with a density of about 5 102 cm3, is roughly collocated with the potential well. A major part of the acceleration potential (4 of 4.5 kV) was concentrated below the C2 orbit, similar as for C1, hence C2 crossed the upper acceleration region of the horn. Note that multiple low plasma density regions appear poleward of the horn in the polar cap, and that the eastward magnetic field is flat in this region, implying that there are no FACs associated with these low density polar cap plasma regions, which have densities as low as a few times 102 cm3. Such regions of low plasma densities are a common feature of the polar cap plasma, and may be related to the outflow of cold ionospheric plasma. [12] The C3 data are summarized in Figure 9 using the same format as for C1 in Figure 7. Before entering into the upward current region, C3 crossed a narrow downward FAC, or R0 current, similar to what was observed by C1. Within R1, the electron and ion data show two adjacent Inverted-V distributions with a clear gap between the two. The energies of the upgoing ion distributions peak at about 7 and 6 keV for the two structures, respectively. The energies of the downgoing electrons are about 1 keV. The associated electric field
A10223
is highly irregular but converging, as can also be seen in the vector plot in Figure 6b, and associated with a broad and relatively deep potential well, with values of 8 kV and 6 kV, corresponding to the surge and horn arcs, respectively, and being roughly consistent with the acceleration potentials inferred from the ion data. The surge and horn identified from the particle data occur in regions of negative slopes in the eastward magnetic field. The upward FAC density for the surge and horn arcs at the surge front, have local values of 0.06 and 0.02 mA/m2, corresponding to 3 and 1 mA/m2, respectively, in the ionosphere. For the particle gap between the surge and horn arc, the FAC density is seen to be close to zero (see sixth plot). The plasma density in ninth panel shows a broad cavity, roughly collocated with the potential well (seventh plot) but being deepest in the gap between the two arcs. This gap is interpreted as polar cap plasma between the surge front and the horn where the plasma density is as low as 102 cm3. Similar to C1 and C2, the major part of the acceleration potential was concentrated below the C3 orbit, hence C3 also crossed in the upper AAR of the surge front and horn arcs. Given the similar orbits and short lag time for the C3 and C4 spacecraft, the data from the two spacecraft are very similar, especially in terms of the electron and magnetic field data, but since the quality of the ion and electric field data is much lower on C4 than on C3 we have decided not to present the C4 data.
4. Synthesis and Discussion [13] The Cluster data presented in Figures 7–9, are used to infer the acceleration potential patterns of the surge and horn arc system from the Cluster C1–C4 crossings at different MLT sectors. Of particular interest are the variations in space and time of these patterns on the temporal scales given by the spacecraft separations, which ranged between 2 and 10 min. Figure 10 shows the derived acceleration potential patterns for the horn arc (C2), the surge and horn arc system close to the surge front (C3/C4) and to the central surge (C1). The horn arc is a classical Inverted-V arc, with an acceleration potential of about 4.5 kV, 4 kV of which is concentrated to altitudes below, and 0.5 kV above, 2.5 RE. The upward FAC density, scaled to the ionospheric level is 1.3 mA/m2 and evenly distributed over the horn width of about 200 km, also scaled to the ionosphere. The C3/C4 pattern shown in the middle has been inferred from the C3 data, but is roughly representative also of the C4 data. The main difference between the C3 and C4 patterns is that the potential wells are slightly deeper on C3 than on C4. The C3 (C4) pattern shows two adjacent U-shaped potential structures, the acceleration potential being 8 (7) kV for the surge and 7 (5) kV for the horn, most of which were concentrated below the altitudes of the C3 and C4 spacecraft. Note that most of the potential contours of the surge and horn are unconnected and separated by the gap in the particle data and the density cavity. As discussed above, the particle gap is interpreted as a region of low density polar cap plasma. C3 and C4 first crossed the surge front then exited for about 1.5 min into the polar cap bay, after which the two spacecraft crossed the horn, as illustrated by the image in Figure 10. For the potential pattern inferred from the C1 data, closer to the central surge, the potential contours of the two structures are more strongly connected. The inferred potential pattern for C3/C4 has a
8 of 15
A10223
MARKLUND ET AL.: MULTI-POINT STUDY OF AURORAL SURGE
A10223
Figure 8. C2 data obtained between 21:05 and 21:35 UT on 31 May 2009. The panels show (from the top) time-energy spectrograms of upgoing, perpendicular, and downgoing electrons, the equatorward electric field component, the eastward magnetic field component, the field-aligned current, the electric potential along the C2 trajectory, the negative of the spacecraft potential, and the plasma density. For more information on the data presentation, see the text. throat-like appearance in contrast to the C1 pattern. The existence of the throat relies mainly on the C3 ion and electron data and not on the electric potential, which shows no clear gap. The reason for relying more on the particle data for the narrow region between the surge front and the horn, is that the electric field measurements within regions of extremely low plasma densities, such as encountered in the gap, are considered less reliable. At the bottom of the acceleration potential patterns
are gray shaded areas, representing plasma density cavities or regions of low-density polar cap plasma, the darker the lower the plasma densities, together with rough estimates of these densities in units of cm3. As can be seen, the densities in the auroral density cavities range between 102 and 101 cm3, and the polar cap plasma cavities have densities as low as a few times 102 cm3. In contrast to the ADC which is associated with an upward FAC, the low density polar cap
9 of 15
A10223
MARKLUND ET AL.: MULTI-POINT STUDY OF AURORAL SURGE
A10223
Figure 9. C3 data obtained between 21:05 and 21:35 UT on 31 May 2009, presented in the same format as in Figure 7. For more information on the data presentation, see the text.
plasma regions are not associated with any FAC. Similar regions of low density PC plasma regions were observed by C2, poleward of the horn. [14] The Cluster and DMSP F16 data presented and discussed here are also used to address another issue regarding the surge electrodynamics, namely how the R1 and R2 FACs at different MLT sectors of the surge system close in the ionosphere. Whereas efforts have been made earlier on this topic (see papers by Fujii et al. [1994], Hoffman et al. [1994], and
Gjerloev and Hoffmann [2002]), the short lag time between the Cluster spacecraft (2–10 min), allow relatively close in time estimates of the degree of latitudinal closure of the surge system at various longitudinal sectors. Figure 11 summarizes the calculations of the net currents derived from the magnetometer data obtained by the DMSP F16 spacecraft and by the four Cluster spacecraft. The subtraction of the background magnetic field was done by subtracting a polynomial fit to the measured magnetic field, averaged over a scale exceeding
10 of 15
A10223
MARKLUND ET AL.: MULTI-POINT STUDY OF AURORAL SURGE
A10223
Figure 10. Acceleration potential pattern of the horn arc inferred from the C2 data, of the double-arc (surge and horn) system, near the surge front, as inferred from the C3 and C4 data, and deeper within the surge as inferred from the C1 data. For more details, see the text.
those of the R1, R2 FAC widths. By trying various fits, and different averaging time intervals, the accuracy in the net current calculations are estimated to be about +/10%, Note that a net current of 25% of the R1 current is estimated for the horn from the F16 data, close to the 23% derived from the C2 data. The net upward currents at the surge front, estimated from the C4 and C3 data, are seen to be very small, 0% and 5%, respectively, whereas the net current in the surge is 19%. Regarding the net current estimate for the surge front, some caution needs to be applied. Given the blob-like appearance of the bright aurora in the surge front region (see Figure 10), and the intense and highly variable electric fields observed there by the C3 and C4 spacecraft (see Figure 12) that hints at significant fine structure in the aurora and electrodynamics [Marklund et al., 1998], the assumption of straight current sheets used for the FAC analysis in this paper might not necessarily apply in this region. In another event study by Amm and Fujii [2008] they found that the surge front was the most important region for the zonal current closure, i.e., by the westward substorm electrojet diverting into upward FAC, while its importance for the meridional current closure between R0, R1 and R2 was small. In Figure 12 we have inserted the net current estimates
and schematic meridional current loops and zonal currents, and connected the R0, R1 and R2 boundaries from the C1, C2, and C3 data on top of the previously shown vector plot in Figure 6b, to get an overview of the event. It is of interest to compare the net current estimates with those presented by Fujii et al. [1994], based on DE-2 and DE-1 data, presented in Figure 1. For the horn, our estimates of the net currents are 23% and 25%, roughly twice the average value presented by Fujii et al. [1994] of 13%, but within the range of the individual values for the net horn current in their study. For the surge and middle surge, the Fujii et al. [1994] results are roughly the same as our results, around 20%. [15] Equatorward of the horn, an intense southwestward electric field was observed. This field was driving a roughly poleward directed current across the horn, being the sum of a Pedersen and a Hall current, rather than a Pedersen current. This implies a fairly weak zonal current equatorward of the horn, given that the zonal components of the Pedersen and Hall currents are oppositely directed and thus canceling each other. Within the horn was observed a westward directed electric field. The absence of a southward electric field component inside the arc is proposed to be due to an oppositely directed polarization electric field which, in
11 of 15
A10223
MARKLUND ET AL.: MULTI-POINT STUDY OF AURORAL SURGE
A10223
Figure 11. The eastward magnetic field component derived from magnetometer data obtained by DMSP F16 across the western horn, by the C2 spacecraft across the horn, by the C4 and C3 spacecraft over the surge and horn near the surge front, and by the C1 spacecraft over the surge. The yellow boxes show the derived net currents in % of the R1 FAC.
addition to the field-aligned currents, was generated in the ionosphere for current continuity maintenance across the arc [Marklund, 1984]. A zonal divergence of the westward Pedersen current inside the horn, for example caused by a gradually decreasing conductance in the horn in the westward direction, is proposed to have fed the net upward horn current. Support for this scenario is that the total acceleration potential decreases from 7 kV in the eastern horn at 22 MLT
(C3, Figure 9), 4.5 kV in the central horn at 21 MLT (C2, Figure 8), to 4 /3 kV in the western horn at 20/19.3 MLT (DMSP F16/F13, Figure 5). Results similar to these were presented by Marghitu et al. [2009]. [16] The equatorward electric field measured by C1 in the region poleward and eastward of the surge, was driving a westward Hall current (WEJ), which is likely to have been diverted into the net upward surge current. Also the downward
12 of 15
A10223
MARKLUND ET AL.: MULTI-POINT STUDY OF AURORAL SURGE
A10223
Figure 12. Vector plots of E? and B? along the C2, C3 and C1 orbits, where the boundaries of the R2, R1, and R0 currents have been combined (dotted yellow lines). The net upward currents derived from C2, C4/C3, C1, are also shown here. The westward electric field measured by C2 within the horn arc drives a westward Pedersen current, JP, the zonal divergence of which is likely to feed the net upward current (23% of the R1 upward FAC) observed in the horn arc. The equatorward electric field measured by C1 poleward of the surge, will drive a westward Hall current, JH, the divergence of which will feed the net upward surge current (19%).
R0 current, contributed partly for feeding the upward surge current. Although the net currents derived from the C3 and C4 data at the surge front were found to be zero or very small, we can, as discussed above, not exclude the possibility that the surge front area as a whole was associated with a net upward current. For that case, two likely candidates for feeding this net current are: (1) a localized downward current at the western edge of the surge front, as observed by the Freja satellite [Marklund et al., 1998], and (2) a possible remnant WEJ flowing between the surge and the surge front, diverting into upward FAC.
5. Summary and Conclusions [17] Observations are presented from Cluster satellite crossings through the upper auroral acceleration region of an auroral surge and connected surge horn arc, as viewed by the DMSP F16 imager in the southern hemisphere during a period of moderate but significant geomagnetic activity associated
with large-scale auroral activity spread over several MLT hours, for many UT hours prior to and after this event. The four spacecraft covered different local time sectors of the surge and horn arc system, with lag times ranging between 2 min and 10 min between the spacecraft. The Cluster data have been used to derive (1) acceleration potential patterns of the horn arc and of the double arc system both at the surge front and deeper inside the surge, and to provide (2) close-in time estimates of how the field-aligned currents at different MLT sectors of the surge system close in the ionosphere. The main findings are summarized below. [18] The horn is a classical Inverted-V structure with a parallel potential drop ranging between 3 kV and 7 kV along its extent. The upward FAC density was about 1.3 mA/m2 uniformly distributed over the 200 km arc width, both values mapped to the ionosphere. The plasma density cavity, with a density of about 0.05 cm3, was collocated with the electric potential well. A major part of the parallel potential drop
13 of 15
A10223
MARKLUND ET AL.: MULTI-POINT STUDY OF AURORAL SURGE
was concentrated below 2.5 RE altitude in the AAR of the horn arc. [19] The patterns derived from the C3 (C4) data for the surge and horn double arc near the surge front, are two U-shaped potential structures with parallel potential drops of 8 (7) kV and 7 (5) kV, respectively, mainly concentrated below 2.6 RE altitude. The corresponding upward FAC densities were 3 and 1 mA/m2, mapped to the ionosphere. The two potential structures were separated by a region of low density polar cap plasma, with a density of about 102 cm3, even lower than observed in the ADCs. In contrast to the ADCs, these low density PC plasma regions were not associated with any FACs. This gap, confirmed by the drop in the ion and electron energies and fluxes, gives the potential pattern a throat-like appearance with a weak coupling between the two U-shaped structures. [20] The pattern derived for the double arc system (surge and horn) close to their connection point and to the central surge, inferred from the C1 data, also consists of two U-potential structures, with parallel potential drops of 7 kV and 5 kV for the surge and horn, respectively, and mainly concentrated below 2.6 RE altitude. The FAC densities mapped to the ionosphere varied between 0.8 and 8 mA/m2, the higher value obtained at the poleward surge boundary. The peak energies of the upgoing ions were consistent with the depth of the electric potential wells. In contrast to the C3/C4 patterns, no low density polar cap plasma separates the two structures at C1, which are also more coupled than the C3/C4 patterns. [21] A net upward current of 23% of the R1 current was estimated from the C2 data across the horn. Within the horn, only the westward component was maintained of the intense southwestward electric field at the equatorward edge of the horn. The residual southward component that mapped to the ionosphere was presumably suppressed inside the horn by an oppositely directed polarization electric field [Marklund, 1984]. The zonal divergence of this westward Pedersen current is proposed to have fed the intense net upward horn current. Support for this scenario is the acceleration potential decrease from the eastern to the western horn observed by the C3 and C2 spacecraft around 21:20 UT and by the DMSP F13 and F16 satellites 20 min before and after this, respectively. [22] The R1 and R2 currents derived from the C3 and C4 magnetometer data near the surge front were roughly balanced, or equivalently, the net FAC was very small (5% and 0%, respectively). However, this result should be taken with some caution, as the assumption of straight current sheets used to estimate the FAC in this study may not be fully applicable in the surge front vicinity. If the surge front should have been associated with a nonzero net upward FAC, the divergence of the WEJ, possibly flowing between the surge and the surge front, or a localized downward FAC at the western edge of the surge front [Marklund et al., 1998], are possible candidates for having fed this net upward FAC. [23] The net upward current calculated from the C1 magnetometer data was 19% of the R1 current, comparable to the 23% derived by Fujii et al. [1994]. The intense equatorward electric field poleward of the surge was driving a WEJ, the divergence of which, as well as localized downward FACs, are likely to have fed the net upward surge FAC. [24] The above findings are only made possible with the multispacecraft measurements of the Cluster fleet. Placing
A10223
the various spacecraft in strategic orbits with respect to each other and with respect to the magnetic field as they transited the auroral acceleration region allowed us to determine the acceleration potential pattern and associated electrodynamics of various large-scale auroral structures, something that has not been possible to this scale with past missions. [25] Acknowledgments. The authors are grateful to a large number of people who have contributed to the Cluster mission. The Cluster project was supported by the European Space Agency and NASA. Project support has also been provided by a grant from NASA Goddard Space Flight Center to the University of Iowa. This study has been supported by the Swedish National Space Board and by Space and Plasma Physics, School of Electrical Engineering, KTH, Stockholm. [26] Robert Lysak thanks the reviewers for their assistance in evaluating this paper.
References Alfvén, H. (1958), On the theory of magnetic storms and auroae, Tellus, 10, 104–116, doi:10.1111/j.2153-3490.1958.tb01991.x. Alfvén, H., and C.-G. Fälthammar (1963), Cosmical Electrodynamics: Fundamental Principles, Int. Ser. Monogr. Phys., pp. 162–167, Clarendon Press, Oxford, U. K. Amm, O., and R. Fujii (2008), Separation of Cowling channel and local closure currents in the vicinity of a substorm breakup spiral, J. Geophys. Res., 113, A06304, doi:10.1029/2008JA013021. Andersson, L., et al. (2002), Characteristics of parallel electric fields in the downward current region, Phys. Plasmas, 9, 3600–3609, doi:10.1063/ 1.1490134. Balogh, A., et al. (1997), The Cluster magnetic investigation, Space Sci. Rev., 79(1–2), 65–91, doi:10.1023/A:1004970907748. Block, L. (1972), Potential double layers in the ionosphere, Cosmic Electr., 3, 349–376. Carlqvist, P., and R. Boström (1970), Space-charge regions above the aurora, J. Geophys. Res., 75, 7140–7146, doi:10.1029/JA075i034p07140. Chiu, Y. T., and M. Schultz (1978), Self-consistent particle and parallel electrostatic field distributions in magnetospheric-ionospheric auroral region, J. Geophys. Res., 83, 629–642, doi:10.1029/JA083iA02p00629. Ergun, R. E., et al. (2000), Parallel electric fields in discrete arcs, Geophys. Res. Lett., 27, 4053–4056, doi:10.1029/2000GL003819. Ergun, R. E., et al. (2002), Parallel electric fields in the upward current region of the aurora: Indirect and direct observations, Phys. Plasmas, 9, 3685, doi:10.1063/1.1499120. Fujii, R., R. A. Hoffman, P. C. Andersen, J. D. Craven, M. Sugiura, L. A. Frank, and N. Maynard (1994), Electrodynamic parameters in the nighttime sector during auroral substorms, J. Geophys. Res., 99, 6093–6112, doi:10.1029/93JA02210. Gjerloev, J. W., and R. A. Hoffmann (2002), Currents in auroral substorms, J. Geophys. Res., 107(A8), 1163, doi:10.1029/2001JA000194. Gustafsson, G., et al. (1997), The Electric Field and Wave Experiment for the Cluster mission, Space Sci. Rev., 79(1–2), 137–156, doi:10.1023/ A:1004975108657. Hoffman, R. A., R. Fujii, and M. Sugiura (1994), Characteristics of the field-aligned current system in the nighttime sector during auroral substorms, J. Geophys. Res., 99, 21,303–21,325, doi:10.1029/94JA01659. Hudson, M., and F. S. Mozer (1978), Electrostatic shocks, double layers, and anomalous resistivity in the magnetosphere, Geophys. Res. Lett., 5, 131–134, doi:10.1029/GL005i002p00131. Johnstone, A., et al. (1997), PEACE: A plasma electron and current experiment, Space Sci. Rev., 79(1–2), 351–398, doi:10.1023/A:1004938001388. Knight, S. (1973), Parallel electric fields, Planet. Space Sci., 21, 741–750, doi:10.1016/0032-0633(73)90093-7. Marghitu, O., T. Karlsson, B. Klecker, G. Haerendel, and J. McFadden (2009), Auroral arc and oval electro-dynamics in the Harang region, J. Geophys. Res., 114, A03214, doi:10.1029/2008JA013630. Marklund, G. T. (1984), Auroral arc classification scheme based on the observed arc-associated electric field pattern, Planet. Space Sci., 32, 193–211, doi:10.1016/0032-0633(84)90154-5. Marklund, G. T., et al. (1994), On intense diverging electric fields associated with black aurora, Geophys. Res. Lett., 21, 1859–1862, doi:10.1029/ 94GL00194. Marklund, G. T., T. Karlsson, and J. Clemmons (1997), On low-altitude particle acceleration and intense electric fields and their relationship to black aurora, J. Geophys. Res., 102, 17,509–17,522, doi:10.1029/ 97JA00334.
14 of 15
A10223
MARKLUND ET AL.: MULTI-POINT STUDY OF AURORAL SURGE
Marklund, G. T., et al. (1998), Observations of the electric field fine structure associated with the westward traveling surge and large-scale auroral spirals, J. Geophys. Res., 103, 4125–4144, doi:10.1029/97JA00558. Marklund, G. T., et al. (2001), Temporal evolution of the electric field accelerating electrons away from the auroral ionosphere, Nature, 414, 724–727, doi:10.1038/414724a. Marklund, G. T., S. Sadeghi, T. Karlsson, P.-A. Lindqvist, H. Nilsson, C. Forsyth, A. Fazakerley, E. A. Lucek, and J. Pickett (2011a), Altitude distribution of the auroral acceleration potential determined from Cluster satellite data at different heights, Phys. Rev. Lett., 106, 055002, doi:10.1103/PhysRevLett.106.055002. Marklund, G. T., et al. (2011b), Evolution in space and time of the quasi‐static acceleration potential of inverted‐V aurora and its interaction with Alfvénic boundary processes, J. Geophys. Res., 116, A00K13, doi:10.1029/2011JA016537. McIlwain, C. E. (1960), Direct measurements of particles producing visible aurora, J. Geophys. Res., 65, 2727–2747, doi:10.1029/JZ065i009p02727. Mizera, P., D. Gorney, and J. Fennell (1982), Experimental verification of an S-shaped potential structure, J. Geophys. Res., 87, 1535–1539, doi:10.1029/JA087iA03p01535.
A10223
Mozer, F. S., and P. Bruston (1967), Electric field measurements in the auroral ionosphere, J. Geophys. Res., 72(3), 1109–1117, doi:10.1029/ JZ072i003p01109. Mozer, F. S., and C. A. Kletzing (1998), Direct observation of large, quasistatic, parallel electric fields in the auroral acceleration region, Geophys. Res. Lett., 25, 1629–1632, doi:10.1029/98GL00849. Mozer, F. S., et al. (1980), Satellite measurements and theories of lowaltitude auroral particle acceleration, Space Sci. Rev., 27, 155–213, doi:10.1007/BF00212238. Paschmann, G., S. Haaland, and R. Treumann (eds.) (2003), Auroral Plasma Physics, Space Sci. Ser. ISSI, vol. 15, Kluwer Acad., Boston, Mass. Réme, H., et al. (1997), The Cluster ion spectrometry (CIS) experiment, Space Sci. Rev., 79(1–2), 303–350, doi:10.1023/A:1004929816409. Song, Y., and R. L. Lysak (2001), Towards a new paradigm: From a quasisteady description to a dynamical description of the magnetosphere, Space Sci. Rev., 95, 273–292, doi:10.1023/A:1005288420253. Temerin, M., et al. (1982), Observations of double layers and solitary waves in the auroral plasma, Phys. Rev. Lett., 48, 1175–1179, doi:10.1103/ PhysRevLett.48.1175.
15 of 15
