21 Dawn‐Dusk Asymmetries of Ionospheric Outflow Kun Li4, Elena A. Kronberg1,2, Mats André3, Patrick W. Daly1, Yong Wei4, and Stein Haaland1,5 ABSTRACT Ion outflow from high‐latitude region of Earth’s ionosphere is an important source of the ions in the magnetosphere. As a part of ionosphere‐magnetosphere coupling, it is also an important driver of dynamics in the magnetosphere. The dawn‐dusk asymmetries in the ion outflow are sometimes non-negligible for the outflow processes and plasma circulation in geospace. The causes of the asymmetries are diverse in terms of the morphologies, efficiencies, and responses to the solar‐wind conditions. In this review, we focus on the mechanisms of dawn‐dusk asymmetries in energization of ion outflow and in the ion transportation in the magnetotail. Asymmetric energization processes are: (1) asymmetry in auroral precipitation associated with electron heating, ion beams and conics; (2) asymmetry in Poynting flux associated with ion heating; (3) asymmetry in cusp spatial distributions that causes various heating on the dayside; (4) high‐altitude ionosphere‐magnetosphere‐coupled convection with IMF controlled dawn‐dusk asymmetry which is essential to centrifugal acceleration for transportation of polar cap ion outflow. 21.1. INTRODUCTION The presence of oxygen ions in the magnetosphere indicates that ionospheric particles flow into the magnetosphere and fill the near-Earth environment. Recent studies suggest that the ionospheric outflow is the main contributor to the plasma in the magnetosphere [Moore and Horwitz, 2007]. One of the sources of ion outflow from the ionosphere can be found in the cusp region, associated with various energy inputs [Lockwood et al., 1985]. Under average solar‐wind condition, with a spiral configuration of interplanetary magnetic field (IMF) and small Bz, ionospheric outflow can provide substantial Max Planck Institute for Solar System Research, Göttingen, Germany 2 Ludwig Maximilian University of Munich, Munich, Germany 3 Swedish Institute of Space Physics, Uppsala, Sweden 4 Institute of Geology and Geophysics, Chinese Academy of Sciences, Beijing, China 5 Birkeland Center for Space Science, University of Bergen, Bergen, Norway 1 amounts of ions into the magnetosphere through so‐ called auroral outflow. Low‐energy ions, with energies up to a few tens of electron volts, are found to dominate the ionospheric outflow through the polar wind from the polar cap. The characteristics of high‐latitude ionospheric outflow and its energization are summarized by Yau and Andre . Outflow can also be found at low latitudes, where the plasma plumes detach from the ionosphere and escape sunward [André and Cully, 2012]. Ions outflowing from the ionosphere can be found throughout the magnetosphere, since it is highly modulated by magnetospheric convection and as a function of geomagnetic activities [Li et al., 2013]. This convection is considered to be a main driver of ion transport and one of the mechanisms energizing the ion outflow. Various mechanisms of ion outflow have been revealed and summarized in Moore and Horwitz [2007, and references therein]. The ionosphere is formed by solar ultraviolet irradiation. To enable ion outflow, the ions gain energy through mechanisms, such as (1) Joule heating, (2) ion acceleration, (3) auroral heating by various waves and auroral precipitation, and (4) parallel electric fields in the Dawn-Dusk Asymmetries in Planetary Plasma Environments, Geophysical Monograph 230, First Edition. Edited by Stein Haaland, Andrei Runov, and Colin Forsyth. © 2017 American Geophysical Union. Published 2017 by John Wiley & Sons, Inc. 273 274 DAWN-DUSK ASYMMETRIES IN PLANETARY PLASMA ENVIRONMENTS auroral zone. Joule heating results from collisions between neutral particles at low altitudes and ions picked up by convecting magnetic fields [Korosmezey et al., 1992], and also due to electric current (mainly the Pedersen current) in the ionosphere passing through plasma with finite conductance. Ion acceleration parallel to the magnetic field can occur due to charge separation giving rise to the polar wind [Banks and Holzer, 1968; Moore et al., 1997] or in the auroral regions due to auroral precipitation, in which ion temperature is strongly correlated to ion outflow [Seo et al., 1997]. Further heating in the auroral zone can occur due to electromagnetic ion cyclotron (EMIC) waves, broadband extra low frequency (BBELF) waves or electrostatic ion cyclotron (EIC) waves (see, e.g., André ), which are in ion gyrofrequency range and have enough power to enable the outflow. Some of the waves in this category are Alfven waves traveling along the magnetic field from the nightside plasma sheet, with higher efficiency to heat the ionosphere [Keiling et al., 2003]. The presence of parallel electric fields in the auroral zone results in ion beams [Mozer et al., 1977]. Furthermore, a DC electric field is observed in the cusp, identified by upward ion beams and ion conics and associated with ion acceleration [Pfaff et al., 1998]. The outflow of ions essentially always exists in regions poleward of the plasmasphere. It can be observed when the spacecraft travels through the polar region. In terms of energy and height characteristics, the outflow is identified as (1) polar wind in the polar cap region, composed of H , He and O , with total energy less than a few tens of electron volts; (2) ion bulk upflow at auroral latitudes dominated by O. The ion bulk upflow turns into ion outflow by ion acceleration at those altitudes it can reach, although normally the bulk upflow does not have sufficient velocity to escape; (3) ion conics associated with the auroral oval, energized by perpendicular electric field oscillations at a frequency near to ion gyrofrequency; (4) ion beams stemming from ion conics in a diverging magnetic field and transverse heating by waves in the upward direction; (5) upwelling ions in the dayside cusp region, accelerated from a few eV, in both parallel and transverse directions. Upwelling ions are characterized by asymmetric upward and downward flux distributions. Chappell et al.  suggested the ionospheric outflow alone could supply the plasma in the magnetosphere. It is also generally accepted that the ionospheric outflow significantly impacts the dynamics in the magnetosphere. For example, (1) presence of low‐energy ions in a plasma plume changes ion densities at the reconnection site on the dayside magnetopause, and consequently decreases the reconnection rates [Walsh et al., 2014]; (2) outflow of oxygen ions, which causes pressure imbalance in the inner magnetosphere, is able to induce the sawtooth oscillations at the nightside geosynchronous orbit, as simulated by Brambles et al. ; (3) oxygen outflow is the main carrier of the currents in the ring current during storm times [Hamilton et al., 1988]. The outflows from dawnside and duskside of the ionosphere are differ greatly. Yau et al.  found the outflow of both H and O , with pitch angles between 100° and 160° and energy from 0.01 to 1 keV, is larger on the duskside. Howarth and Yau  studied the trajectories of polar‐wind ions with a single‐particle approach, with input from Akebono measurements. They found polar‐wind supplies more ions to the duskside of the plasma sheet when the IMF is duskward. When the IMF is dawnward, the distribution of polar‐wind ions in the plasma sheet is more even. They suggest there is a dependence on IMF By in the polar‐wind deposition on the plasma sheet. In contrast, a study by Li et al.  showed a persistent dawn‐dusk asymmetry exists for low‐energy ions from the ionosphere transported to the plasma sheet, without corresponding changes to the IMF direction. By checking their origin in the ionosphere, no significant dawn‐ dusk asymmetry is found. These low‐energy ions are mainly H with both thermal and kinetic energy lower than 70 eV. Liao et al.  reported O from the cusp in the Northern Hemisphere tends to supply more to the dawnside of the tail lobes when IMF By is positive. The asymmetry is reversed when IMF By is strongly negative. However, the dawn‐dusk asymmetry has attracted less attention than day‐night and hemispheric asymmetries, which can largely be explained by differences in solar irradiation and magnetic field compression. The dawn‐ dusk asymmetry shows more complicated features, and needs more investigations to explain. Even under a condition with symmetric solar‐wind parameters, the dawn‐ dusk asymmetry may exist. As an important aspect of magnetosphere‐ionosphere coupling, dawn‐dusk asymmetry in the ionospheric outflow is created by asymmetries in ionospheric heatings, which are caused by various energy inputs from the magnetosphere. Dawn‐dusk asymmetries in the ionospheric outflow and heating are studied by both observations and simulations. Rather than listing all the mechanisms involved in the ion outflow, we only summarize the asymmetries in energization mechanisms and outflow transports. Their dependence on solar‐wind conditions is also analyzed. In summary, we try to determine the relative importance of the sources of asymmetries, and outline problems to be solved. 21.2. ASYMMETRIC ENERGIZATION IN THE IONOSPHERE The ionosphere is created by the dayside upper atmosphere absorbing the ultraviolet (UV) component of solar irradiation. In the ionosphere, the electrons are removed Dawn‐Dusk Asymmetries of Ionospheric Outflow 275 from ionized atoms, become photoelectrons, and gain high enough energies to escape the Earth’s gravity. A simple dawn‐dusk asymmetric energy input can be considered as the asymmetric heat absorbed from the Sun. The corotation of the ionosphere with the Earth can turn the day‐night asymmetric heating into a dawn‐dusk asymmetric one. As the ionospheric plasma with higher temperature on the dayside turns to the dusk sector, the ionospheric plasma cooled down on nightside turns to the dawn sector. This gives rise to higher temperature in the ionosphere, and larger scale height of the ionosphere on the dusk side. One could thus expect that there is more ion outflow on the dusk side of the ionosphere. However, at high latitudes, where the main source of ion outflow is located, the ionosphere does not necessarily co-rotate with the Earth thus this effect may be reduced. It is necessary to do more studies on the importance of this effect. Figure 21.1 demonstrates the causal relationships between asymmetric ionospheric outflow and various asymmetric energy inputs. Each causal relationship is indicated by an arrow with a number labeled beside it. Some of correlations between energizations and the outflow have been quantitatively studied by Strangeway et al.  and Moore and Khazanov . In their studies, the ionospheric outflow is enhanced due to an increase in Poynting flux or electron precipitation. Below, we list the theoretical analysis, simulations, and observations (if available) of these causal relationships. Energization in the ionosphere can be classified as ion and electron energizations. For the ion energization of the thermal-type, increases in the ion temperature increases the scale height of ions, making more ion upwelling possible (illustrated as chart flow 5 and 6). In the kinetic-type energization, ions with higher velocities can easily overcome the gravitational drag. The electron scale height can increase with the electron temperature, which is correlated to the electron precipitation as depicted by chart flow 9. This generates a larger ambipolar electric field (chart flow 10), and enhances the ion upwelling as depicted by chart flow 11. Ion upwelling turns into ion outflow when waves interact with ions and enable them to escape in the form of ion beams and conics (chart flow 7). At high altitudes, ions accelerated by the ambipolar electric field are continuously accelerated by centrifugal acceleration (chart flow 1) [Cladis, 1986; Cladis et al., 2000]. Asymmetric ion outflow 1 ELF/VLF waves 7 Asymmetric centrifugal acceleration(polar wind) 6 Asymmetric ion upwelling Asymmetric increase of ion scale height Asymmetry in ionospheric convection 11 Asymmetric increase of electron scale height 10 5 2 Asymmetric electron heating Asymmetric Joule heating Asymmetric neutral wind of low altitude atmosphere 4 9 Asymmetric poynting flux 3 12 Asymmetric conductance of ionosphere 13 Asymmetric spatial distribution of cusp region 8 16 Asymmetric field-aligned currents Asymmetric electron precipitation 14 15 Asymmetric spatial distribution of aurora Figure 21.1 Chart of relationships between the dawn‐dusk asymmetry in ionospheric outflow and various asymmetries in energy inputs from the magnetosphere and the magnetosheath. Each relationship with a number labeled beside it is explained in the text. 276 DAWN-DUSK ASYMMETRIES IN PLANETARY PLASMA ENVIRONMENTS 21.2.1. Joule Heating in the Ionosphere Asymmetric ion heating can be found at lower altitudes. In chart flow 2, the Joule heating is caused by collisions between neutral particles and ions, which are driven by ionospheric convection and field aligned current (FAC) at lower altitudes in the ionosphere. In an investigation of collisions between neutral particles and ions, Bjoland et al.  conducted a statistical study of Joule heating in the ionospheric F-layer in the Northern Hemisphere, using data obtained by SuperDARN radar and CHAMP satellite. They found a significant dawn‐dusk asymmetry in this heating mechanism. The heating maximises at postnoon and postmidnight, which are collocated with maximum eastward and westward neutral wind velocity respectively. Noted by Strangeway et al. , casual relationship is found between ionospheric Joule heating and the Alfvénic Poynting flux (chart flow 4). The Alfvénic Poynting flux is highly related to both cusp and aurora region. Alfven waves propagating along open magnetic field lines transmit magnetic perturbations to the cusp. It is suggested that outflow rates in the cusp are enhanced during a coronal mass ejection pass. The location of the cusp may change with IMF orientation and show dawn‐dusk asymmetry; this is discussed in section 21.2.3. Alfvénic Poynting flux in the auroral region is associated with broadband precipitating electrons and ions, which originate from the plasma sheet. This is usually observed as wave/broadband aurora at about the time of substorm onset dipolarization. Two other types of aurora in a substorm cycle are discussed in section 21.2.2. They are (1) diffuse aurora associated with electron and ion precipitation from the plasma sheet, and (2) monoenergetic aurora in which the electron precipitation from the plasma sheet is accelerated by the global upward field‐aligned current system [Newell et al., 2010; Wing et al., 2013 and Chapter 20 in this volume]. In chart flow 15, asymmetric aurora distributions will result in asymmetric electron precipitation. Keiling et al.  compared 1-year’s worth of Alfvén wave data obtained by the Polar spacecraft at altitudes from 28,000 to 35,000 km with images of the northern aurora in both visible and UV spectrum. They found the average energy Poynting flux toward the Earth (downward) concentrates at regions in the ionosphere where the average aurora intensifies (Fig. 21.2a). They concluded that the Poynting flux is adequate to accelerate the ions within the auroral oval, in addition to FAC, which are also found over the auroral oval at those altitudes. This observation is reproduced by the LFM global simulation in Zhang et al. , as shown in Figure 21.2b. Here, clear asymmetry in the downward Alfvénic Poynting flux is shown in both observation (Fig. 21.2a) and simulation (Fig. 21.2b). To investigate the cause of the asymmetry, Zhang et al.  repeated the simulations with controlled spatial variation of conductance in the ionosphere. In Figure 21.2c, there are asymmetries in Alfvénic Poynting fluxes toward the Earth in an MHD simulation with enhanced conductance due to electron precipitation. The average downward Poynting flux on the nightside has a higher flux in the premidnight sector. This asymmetry disappears when the conductance is spatially uniform as we can see from the simulation in Figure 21.2d. Note that the fluxes in Figure 21.2a and b are 1 year and 24 h averaged fluxes, respectively. With occasionally uniform conductance in these time periods, they show relatively milder asymmetries than that in Figure 21.2c, which shows the fluxes in much shorter time periods (1 h). Also note that the Poynting flux in the cusp region in Figure 21.2c and d is minimized by controlled steady upstream driving parameters. Alfvénic wave activity is weak at the subsolar point, as well as at the cusp because the subsolar point connects the cusp with a magnetic field line. 21.2.2. Electron Heating in the Auroral Region Clear dawn‐dusk asymmetry can also be identified in electron energy precipitation of diffuse aurora and monoenergetic aurora [Wing et al., 2013 and Chapter 20 in this volume]. In Figure 21.3a and b are energy fluxes of diffuse aurora electron precipitation and monoenergetic electron precipitation, respectively. The data are obtained from DMSP at an altitude of 845 km. Diffuse aurora electron energy precipitation is more intense in the dawn sector of the ionosphere, with peak energy input almost twice the peak energy flux of monoenergetic electrons, while monoenergetic electron energy precipitation is more concentrated in the premidnight sector of the ionosphere. Monoenergetic precipitation is found to be enhanced in the winter hemisphere. They consider that the enhancement of monoenergetic electron precipitating energy flux may result from enhancement of the magnetotail stretching and the region 1 FAC increase during the substorm growth phase. Upward FAC accelerates electrons downward, enabling them to reach the upper atmosphere and creating discrete aurora. The FAC exhibits dawn‐dusk asymmetry, as studied by Iijima and Potemra . In the dawn/ dusk sector of the ionosphere, upward FAC is at relatively lower/higher latitude. The amplitude of FAC changes with solar wind and geomagnetic activities, and is indirectly influenced by magnetosphere‐ionosphere convection. Poleward of the region 1 FAC in the dayside, the region 0 FAC is believed to be associated with the mantle. It is found that the polarity of the region 0 current changes with direction of IMF By and upward region 0 current occurs more frequently in the prenoon sector [Wing et al., 2010]. Dawn‐Dusk Asymmetries of Ionospheric Outflow 277 Average Alfvénic Poynting flux 6–180 s bandpass, mapped to 100-km altitude POLAR 4–6 RE Kp = 2– 12 mW/m2 (a) 0.8 18 80 70 60 06 0.4 LFM 1.8 RE Kp = 3 12 mW/m2 1.1 (b) 18 80 70 60 0.5 06 0.0 Keiling et al.  24 Jan–Dec 1997 24 Simulated with enhanced conductance (c) 0.0 4–5 Feb 2004 Simulated with uniform conductance 5 mho 12 S‖ (d) 12 S‖ 60 06 0 24 70 18 60 0.5 mW/m2 70 18 mW/m2 1 06 0.0 24 Figure 21.2 Morphology of downward Alfvénic Poynting flux. (a) One year average flux measured by the Polar spacecraft [adopted from Keiling et al., 2003]; (b) 24 h averaged flux simulated by controlled global MHD model, as a comparison of observation in (a); (c) the dawn‐dusk asymmetry in Poynting flux with the same model but with asymmetric conductance in the ionosphere due to electron precipitation; (d) the same as (c), but the symmetric conductance of 5 mho is applied. Note that the fluxes are higher in (b) to (d) than in panel (a), since the altitude of simulation in (b) to (d) is lower than that of observation in (a). (b) to (d) adopted from Zhang et al. . Moreover, fast flows in the magnetotail may be associated with monoenergetic aurora, as they are observed faster in the dusk sector [Hori et al., 2000]. More recently, Zhang et al.  simulated the asymmetries in these two types of aurora with the Lyon‐Fedder‐Mobarry (LFM) global model. As shown in Figure 21.3c, both energy fluxes are simulated and mapped to 100 km reference altitude. These simulated energy fluxes are observed with similar morphologies. The causal relationship between electron heating and auroral electron precipitation or aurora is indicated as chart flow 9 or 15 in Figure 21.1. In principle, the conductivity and conductance can be enhanced when plasma density is high due to plasma precipitation (chart flow 8). Ionospheric conductance associated with electron precipitation is calculated by Robinson et al. , in which both Perdersen conductance and Hall conductance are positively related to the energy flux. High ionospheric conductance is observed in regions where the electron precipitation is enhanced [Ohtani et al., 2009]. With enhanced conductance, simulation by Zhang et al.  indicates a dawn‐dusk asymmetry in Alfvénic Poynting flux (Fig. 21.2c). This nightside asymmetric conductance induced asymmetry in the Poynting energy flux is indicated as chart flow 3 in Figure 21.1. 21.2.3. Cusp Precipitation Plasma precipitation to the dayside ionosphere comes directly from the magnetosheath through the cusp on open magnetic field lines [Meng, 1981; Wing et al., 1996, 2001], in addition to the uniform electron polar rain. The magnetosheath hot plasma precipitation is thus 278 DAWN-DUSK ASYMMETRIES IN PLANETARY PLASMA ENVIRONMENTS Diffuse aurora electron precipitation (a) Energy flux (c) 12 12 2.1 GW mW/m2 15 4.7 GW 09 60° 1.0 70° 0.5 18 06 06 18 80 75 70 65 0 60 21 55 03 50 00 00 (b) Mono Monoenergetic electron energy flux 12 mW/m2 4 15 Diffuse 0 4 mW/m2 09 0.5 0.25 06 18 80 75 70 65 0 60 21 55 03 50 00 Figure 21.3 Asymmetric energy flux of auroral electron precipitations. (a) Observed diffuse aurora electron precipitation with higher flux in the postmidnight and dawnward sectors; (b) observed monoenergetic electron precipitation favor in the premidnight sector of the ionosphere [after Wing et al., 2013]; (c) energy flux for both precipitations with 1 h averaging. The fluxes are simulated by Global MHD model, with fixed solar‐wind parameters. Numbers at the top of the panel are the total power input to the ionosphere. After Zhang et al. . collocated with the cusp and an important path for solar‐wind entry [Wing et al., 2014]. At high altitudes, the cusp energetic particle (CEP) precipitation is controlled by IMF directions [Woch and Lundin, 1992] and correlates with ULF waves [Chen and Fritz, 1998]. Escoubet et al.  found particle precipitation is immediately enhanced when IMF turns from southward to northward. Therefore, the location of the cusp is considered as a proxy for cusp precipitation and all accelerations involved in the cusp. In a study by Newell et al. , there is dawn‐dusk asymmetry in the location of the cusp, which is modulated by IMF directions. When IMF By is negative (positive), the peak probability of the observed cusp location projected onto the ionosphere is shifted to the prenoon sector for the Northern (Southern) Hemisphere. This shifting effect is more pronounced when IMF Bz is southward. The statistical location of the cusp is studied by Wing et al. , with DMSP observations. They suggest that the cusp location changes as a function of IMF By and Bz. Dawn‐Dusk Asymmetries of Ionospheric Outflow 279 Both high and low altitudes of the cusp show dawn‐dusk asymmetry with regulation by the IMF. In conjugate hemispheres, the aurora can be observed at places on both hemispheres connecting each other with a newly reconnected magnetic field line. This magnetic field line is reconnected at the reconnection site in the dayside magnetopause, which is influenced by the IMF orientation because it can change the configuration of the magnetosphere. This effect is studied by simultaneous imaging with IMAGE and Polar in ultraviolet wavelengths [Østgaard et al., 2007]. They show the evidence that IMF By and the geomagnetic dipole tilt angle are controllers of the substorm onset location. A MHD simulation has also been conducted, suggesting a dawn‐dusk asymmetry. In a study by Siscoe et al. , the cusp region at high altitudes moves towards lower latitudes when the IMF turns from northward to southward. The ion precipitation through the cusp to the ionosphere with dawn‐dusk asymmetry of ion‐neutral particle collisions causes asymmetric Joule heating (shown as chart flow 13). While the asymmetry of electron precipitation in the dayside also may be attributed to asymmetric spatial distribution of the cusp, as indicated as chart flow 14. 21.3. ASYMMETRIC TRANSPORTATION OF OUTFLOW IONS 21.3.1. Cusp Outflow Cusp outflow is found convecting with the opened magnetic field lines over the polar cap to the magnetotail lobes. Two‐dimensional kinetic simulation by Horwitz and Lockwood  shows that the higher the outflow velocity of the ion, the higher the altitude the ion can travel and thus the farther along the plasma sheet it can reach. At the specific region, the cusp outflow ions are monoenergetic, whereas the location of the cusp on the ionosphere is modulated by the IMF, as mentioned in section 21.2.3. In the cusp, O ions are efficiently accelerated in the ionosphere and transported to the lobes even during nonstorm times. Liao et al.  used Cluster measurements to study the transport path of cusp outflow in the magnetosphere. Their results show IMF By is a main controller. When IMF By is positive, cusp O from the northern/ southern hemisphere tends to supply the dawn/dusk side of the magnetotail. The asymmetry is obviously reversed when IMF By is strongly negative. This can be explained by magnetospheric convection, which is modulated by magnetic tension force after a magnetic reconnection in the dayside. Noda et al.  used Cluster data to conclude that IMF By controls the y component of magnetospheric convection velocity in the lobes. Their results show that a clockwise/counterclockwise convection (seen from the Sun) presents when IMF By is negative/positive. 21.3.2. Auroral Outflow: Ion Conics and Beams Ion conics with peak flux at an angle to the local magnetic field, as summarized by Yau and Andre , are associated with ion transverse heating. Ions with large pitch angles (at or close to 90°) are interpreted as transverse accelerated ions (TAI). On the nightside, they are observed at low altitudes from 400 km by sounding rockets during active aurora, to 1700 km frequently by the Freja satellite. In the dayside, they are observed at 3000 km by the Akebono satellite. On the other hand, ion beams with peak flux along the magnetic field are directed outward. These ions are observed at high altitudes (>5000 km), and may also be transformed from ion conics at low altitudes because of the magnetic mirror force. However, statistical altitude distributions of beams and conics reveal that the magnetic mirror force is not necessarily the only acceleration of ion beams. Both ion beams and conics have energies ranging from 10 eV to a few keV. Statistical studies of upflowing ion beams and conics have been conducted by Yau et al.  and Kondo et al. . They utilized DE 1 measurements to study upflowing ion beams and conics of H and O , in energy range from 0.01 to 17 keV (see Fig. 21.4). They found that the source of upflowing ion beams and conics is mostly confined to the auroral regions. Within the auroral regions, they also found clear dawn‐dusk asymmetry in the occurrence frequency of ion upward beams and conics. At geomagnetic quiet times (Kp ≤ 3+), the peak occurrence frequency of beams with energy higher than 1 keV is in the premidnight sector (21–24 MLT), whereas the minimum occurrence frequency is in the morning sector. This asymmetry is absent for lower‐energy ion beams ( 1 keV). The occurrence frequency of conics for both H and O peaks on the dayside around 11 MLT, indicating possible wave interactions in the cusp. At disturbed times (Kp 4−), two peaks of beam occurrence frequency are found in the dusk (18–21 MLT) and dawn sectors (03–09 MLT) for energy above 1 keV, although there is no obvious asymmetry found for beams with energy below 1 keV. Conics with energy above 1 keV are found frequently to appear in the dawn sector (06–09 MLT). 21.3.3. Ionospheric Convection The chart flow 1 indicates that the asymmetric convection velocities on the topside ionosphere can cause asymmetric outflow velocities, and consequently asymmetric outflow flux at high altitudes. In high‐latitude regions of the topside ionosphere, the ionospheric magnetic field 280 DAWN-DUSK ASYMMETRIES IN PLANETARY PLASMA ENVIRONMENTS Kp≤ 3+ 1.0 0.01– 1 keV Kp≥ 4– H+ 0+ H+ 4– 17 keV 0.5 0.0 Occurrence frequency Occurrence frequency 0.5 1– 4 keV 0.2 0.1 0.0 0.10 0.0 0.50 0.25 0.00 0.2 0.1 0.05 0.00 0+ 1.0 0 6 12 18 24 0 6 12 18 Magnetic local time 24 0.0 0 6 12 18 24 0 6 12 18 24 Magnetic local time Figure 21.4 Occurrence frequencies of dawn‐dusk asymmetries of upward ion beams (solid circles) and conics (open circles) for H and O in three energy ranges. Left and right panels are for Kp 3+ and Kp 4−, respectively. The occurrence frequencies are integrated over altitudes from 8,000 to 23,300 km. Adopted from Kondo et al. . lines connect to the magnetosphere, where the convection is highly controlled by the IMF direction [Dungey, 1961]. This leads to IMF modulation and dawn‐dusk asymmetry of the polar ionospheric convection (see also Chapters 9 and 10). For example, Haaland et al.  used Cluster EDI data to study the ionospheric convection and electric potentials by mapping magnetospheric convection. Figure 21.5 shows the electric potentials mapped to 400 km altitude for the northern hemisphere. The potentials are derived with Cluster EDI data from 2001 to 2006. The convection velocities VE B can be calculated by: VE B B /B 2. At places where the equipotential lines are closer to each other, the electric field is stronger. When the IMF is purely southward, the highest convection velocities can be found on the meridian line of noon‐midnight in the high‐latitude polar cap region with antisunward directions. The convection shows a pattern of symmetric convection cells. With positive/negative IMF By, the convection cells rotate clockwise/counterclockwise. The dawn‐dusk asymmetry of ionospheric convection is clear even for northward IMF, and this phenomena is opposite in the other hemisphere. Given the same magnetic field configuration, the asymmetric convection results in asymmetric centrifugal acceleration at high altitudes, where the centrifugal acceleration is the main energization process. ac VE VE B B dbˆ dt bˆ bˆ V VE t S (21.1) B bˆ In equation (21.1), ac is the centrifugal acceleration of the ion’s guiding center, VE B is the E B convection velocity, b̂ is unit vector of magnetic field, and S is unit length along the magnetic field line. The convection velocity is important since it highly effects the total centrifugal acceleration. Nilsson et al.  conclude that the average outward centrifugal acceleration is about 5 m/s2. Although it is small, the cumulative effect could result in a significant high velocity even for low‐energy ions. 21.3.4. Polar Cap Outflow In the polar‐wind scenario [Axford, 1968], charge separation is formed because ions usually have a lower scale height than electrons. Consequently, an ambipolar electric field is generated. The ambipolar electric field can accelerate the ions and enable them to escape. Rather than transverse heating by waves in the auroral region, ions escaping from polar cap are produced by centrifugal acceleration at high altitudes. Polar wind is characterized by low‐energy ions (few 10s eV), which are hard to detect by regular measurements due to the spacecraft charging problem [Chappell et al., 1987]. Below, we refer to this type of ion as cold ions. Few observations have been conducted before [e.g., Su et al., 1998; Seki et al., 2003], with artificially modulated low‐energy part of Maxwellian distribution, or measurements in the illumination shadow of the Earth where there is no spacecraft charging. With the “wake” technique to measure the electric field in the plasma wake formed by the positively charged Dawn‐Dusk Asymmetries of Ionospheric Outflow 281 Sector 7: Bz+/By– 12 Sector 0: Bz+ 12 ∆U = 18.3 kV Sector 1: Bz+/By+ 12 ∆U = 15.5 kV ∆U = 27.6 kV 0 18 06 3 3 –6 80 70 06 06 80 18 70 9 3 0 00 7.1 kV –8.7 kV Sector 6: By– 12 ∆U = 33.5 kV 80 18 70 06 –13.3 kV 0 10 20 30 3 –6 –10 0 Potential [kV] 0 Sector 5: Bz– /By– 12 ∆U = 51.3 kV 60 06 9 COR wmap < 5.0 km/s bias_> = _0.96 MP_dist_ > 2RE 15.5 kV 70 2 –1 –6 –12 –20 14.2 kV 00 80 18 –30 3 Sector 2: By+ 12 ∆U = 40.6 kV 06 9 00 6.8 kV North Polar Cap 2001/02–2006/03 EDI C1–C3 3 –17.9 kV 00 06 15 –11.2 kV 06 –6 0 06 –6 18 70 0 80 –19.2 kV 00 21.4 kV Sector 3: Bz–/By+ Sector 4: Bz– 12 ∆U = 61.9 kV 12 80 70 60 06 18 80 70 60 06 15 21 –6 9 9 15 21 –2 18 2 –1 –12 – –1 24 8 4 06 0 60 27 70 –18 –6 80 21 15 18 –12 –18 –6 ∆U = 54.3 kV 3 3 0 –27.3 kV 00 24.0 kV –27.3 kV 00 3 9 0 34.6 kV –26.4 kV 00 27.9 kV Figure 21.5 Electric potential of the ionosphere as a function of magnetic coordination in AACGM system and IMF orientation for the Northern Hemisphere. IMF orientation is indicated at the top of each sector. Potential in sectors is calculated with EDI measurements and mapped to 400 km altitudes, and color coded as in the colorbar at the center. Equipotential lines overlapping the colors are with a 3 kV spacing. Maximum and minimum potentials are indicated at the bottom of the sector. Total potential drop ∆U is shown at top right of each sector. Adopted from Haaland et al. . spacecraft, the bulk velocity of cold ions can be derived [Engwall et al., 2009], as well as the density by means of spacecraft electric potentials. Li et al.  studied the source of cold ions on the ionosphere by calculating ion trajectories. The trajectory calculation takes into account the centrifugal acceleration and ionosphere‐magnetosphere convection enters as indicated in equation (21.1). They found the cold ions originate from the polar cap region without significant dawn‐dusk asymmetry. However, in the follow‐up study, Li et al.  show a clear 282 DAWN-DUSK ASYMMETRIES IN PLANETARY PLASMA ENVIRONMENTS 21.4. SUMMARY (a) Y[RE ] IMF southward 10 X[RE ] –60 –50 –40 –30 –20 –10 0 –10 (b) Y[RE ] 10 X[RE ] –60 –50 –40 –30 –20 –10 0 –10 Flux [s–1/cm2] <=104 105 >=106 Figure 21.6 Cold ion outflow flux supply on the plasma sheet during a period of southward IMF. (a) or (b) the outflow fluxes, which can be traced back to the Northern or the Southern Hemisphere, with ion trajectory calculation by taking into account the accelerations in equation (21.1) and the magnetospheric convection. Consistent dawn‐dusk asymmetry is also found during periods with IMF in other directions. After Li et al. . asymmetry in the cold ion supply to the plasma sheet (see Fig. 21.6). The cold ions consistently tend to supply more to the dusk side of the plasma sheet regardless of the IMF direction. This phenomenon exhibits in the other hemisphere with similar pattern. As the orbit of Cluster is higher in the southern hemisphere, cold ion outflow is generally measured at higher altitudes in that hemisphere than in the northern hemisphere. Since ions at higher altitudes are transported to the plasma sheet further in the tail, ionospheric cold ion outflow from the southern hemisphere can be found further in the tail. This is attributed to the hemispheric asymmetry shown in two panels in Figure 21.6. This asymmetry, with more cold ion flow on the dusk plasma sheet, coincides with the fact that the signatures of reconnection are more commonly observed on the dusk side, as summarized in Walsh et al. . Asymmetries are found to be general in both ionospheric particle heatings and ionospheric outflow. In section 21.2, the asymmetric heatings are found to exist for both ions and electrons. The heating of electrons essentially increases the electron scale height, resulting in a larger ambipolar electric field that accelerates the ions. An important asymmetric heating of electrons depicted in previous studies is the electron precipitation associated with asymmetric spatial distribution of the cusp. Moreover, diffuse and monoenergetic aurora with dawn‐dusk asymmetry plays a role in asymmetric electron heating. In auroral and cusp regions, ion outflow composed of ion beams and conics are often observed with dawn‐dusk asymmetry. Besides, the electron precipitation increases the ionospheric conductance, which causes an asymmetric Poynting flux for both cusp and auroral regions. Asymmetric Poynting flux coupled with asymmetric atom‐neutral particle friction bring asymmetric Joule heating. With asymmetric scale height of ions, the ion upwelling is asymmetric. In the polar cap, although there is no evidence of asymmetry in ion energization, the asymmetries in the ionospheric convection driven by the IMF play a role for ion outflow at high altitudes. The ionosphere is heated up by solar irradiation on the dayside. When dayside ionosphere corotates with the Earth duskward, the plasma in the ionosphere with higher temperature fills into premidnight sector. Like wise, the cooler plasma from the postmidnight sector of the ionosphere rotates to the morning side. A dawn‐ dusk asymmetry with high temperature on the dusk side is consequently considered to exist. However, no clear evidence from observation of polar‐cap outflow could confirm this situation. This requires further investigation. It is interesting that the cold ion outflow from the polar cap is consistently found more on the duskside of the plasma sheet, even without asymmetry in its source region. This may cause duskside favored reconnection, and fast flow on the duskside of the magnetotail. The ionospheric conductance in the dusk sector may consequently be enhanced. This can be observed as an asymmetric Poynting flux, which is also simulated. The open question remains: is asymmetric auroral ion outflow caused by asymmetry in the Poynting flux, a trigger of asymmetric supply of cold ion on the plasma sheet? Another open question is, how long is the time taken by each energization process for the ion outflow? And how much asymmetric energy input to the ionosphere could generate the observable asymmetry in ion outflow? Dawn‐Dusk Asymmetries of Ionospheric Outflow 283 REFERENCES André, M. (1997), Waves and wave‐particle interactions in the auroral region, J. Atmos. Solar Terr. Phys., 59, 1687–1712; doi:10.1016/S1364‐6826(96)00173‐3. André, M., and C. M. Cully (2012), Low‐energy ions: A previously hidden solar system particle population, Geophys. Res. Lett., 39, L03101; doi:10.1029/2011GL050242. Axford, W. I. (1968), The polar wind and the terrestrial helium budget, J. Geophys. Res., 73, 6855–6859; doi:10.1029/ JA073i021p06855. Banks, P. M., and T. E. Holzer (1968), The polar wind, J. Geophys. Res., 73, 6846–6854; doi:10.1029/JA073i021p06846. Bjoland, L. M., et al. (2015), Interplanetary magnetic field and solar cycle dependence of Northern Hemisphere F region Joule heating, J. Geophys. Res., 120, 1478–1487; doi:10.1002/ 2014JA020586. Brambles, O. J., W. Lotko, B. Zhang, M. Wiltberger, J. Lyon, and R. J. Strangeway (2011), Magnetosphere sawtooth oscillations induced by ionospheric outflow, Science, 332, 1183; doi:10.1126/science.1202869. Chappell, C. R., T. E. Moore, and J. H. Waite, Jr. (1987), The ionosphere as a fully adequate source of plasma for the Earth’s magnetosphere, J. Geophys. Res., 92, 5896–5910; doi:10.1029/JA092iA06p05896. Chen, J., and T. A. Fritz (1998), Correlation of cusp MeV helium with turbulent ULF power spectra and its implications, Geophys. Res. Lett., 25, 4113–4116; doi:10.1029/1998GL900122. Cladis, J. B. (1986), Parallel acceleration and transport of ions from polar ionosphere to plasma sheet, Geophys. Res. Lett., 13, 893–896; doi:10.1029/GL013i009p00893. Cladis, J. B., H. L. Collin, O. W. Lennartsson, T. E. Moore, W. K. Peterson, and C. T. Russell (2000), Observations of centrifugal acceleration during compression of magnetosphere, Geophys. Res. Lett., 27, 915–918; doi:10.1029/ 1999GL010737. Dungey, J. W. (1961), Interplanetary magnetic field and the auroral zones, Phys. Rev. Lett., 6, 47–48; doi:10.1103/ PhysRevLett.6.47. Engwall, E., A. I. Eriksson, C. M. Cully, M. André, R. Torbert, and H. Vaith (2009), Earth’s ionospheric outflow dominated by hidden cold plasma, Nat. Geosci., 24–27; doi:10.1038/ NGEO387. Escoubet, C. P., et al. (2008), Effect of a northward turning of the interplanetary magnetic field on cusp precipitation as observed by Cluster, J. Geophys. Res., 113, A07S13; doi:10.1029/2007JA012771. Haaland, S., M. Förster, and G. Paschmann (2007), High‐ latitude plasma convection from cluster EDI measurements: Method and IMF‐dependence, Ann. Geophys., 239–253. Hamilton, D. C., G. Gloeckler, F. M. Ipavich, B. Wilken, and W. Stuedemann (1988), Ring current development during the great geomagnetic storm of February 1986, J. Geophys. Res., 93, 14,343–14,355; doi:10.1029/JA093iA12p14343. Hori, T., K. Maezawa, Y. Saito, and T. Mukai (2000), Average profile of ion flow and convection electric field in the near‐ Earth plasma sheet, Geophys. Res. Lett., 27, 1623–1626; doi:10.1029/1999GL003737. Horwitz, J. L., and M. Lockwood (1985), The cleft ion fountain: A two‐dimensional kinetic model, J. Geophys. Res., 90, 9749–9762; doi:10.1029/JA090iA10p09749. Howarth, A., and A. W. Yau (2008), The effects of IMF and convection on thermal ion outflow in magnetosphere‐ ionosphere coupling, J. Geophys. Res., 70, 2132–2143; doi:10.1016/j.jastp.2008.08.008. Iijima, T., and T. A. Potemra (1976), The amplitude distribution of field‐aligned currents at northern high latitudes observed by Triad, J. Geophys. Res., 81, 2165–2174; doi:10.1029/ JA081i013p02165. Keiling, A., J. R. Wygant, C. A. Cattell, F. S. Mozer, and C. T. Russell (2003), the global morphology of wave poynting flux: Powering the aurora, Science, 299, 383–386; doi:10.1126/ science.1080073. Kondo, T., B. A. Whalen, A. W. Yau, and W. K. Peterson (1990), Statistical analysis of upflowing ion beam and conic distributions at DE 1 altitudes, J. Geophys. Res., 95, 12,091–12,102; doi:10.1029/JA095iA08p12091. Korosmezey, A., C. E. Rasmussen, T. I. Gombosi, and G. V. Khazanov (1992), Anisotropic ion heating and parallel O(+) acceleration in regions of rapid E x B convection, Geophys. Res. Lett., 19, 2289–2292; doi:10.1029/92GL02489. Li, K., et al. (2012), On the ionospheric source region of cold ion outflow, Geophys. Res. Lett., 39, L18102; doi:10.1029/ 2012GL053297. Li, K., et al. (2013), Transport of cold ions from the polar ionosphere to the plasma sheet, J. Geophys. Res., 118, 5467–5477; doi:10.1002/jgra.50518. Liao, J., L. M. Kistler, C. G. Mouikis, B. Klecker, I. Dandouras, and J.‐C. Zhang (2010), Statistical study of O transport from the cusp to the lobes with Cluster CODIF data, J. Geophys. Res., 115, A00J15; doi:10.1029/2010JA015613. Lockwood, M., J. H. Waite Jr., T. E. Moore, C. R. Chappell, and M. O. Chandler (1985), The cleft ion fountain, J. Geophys. Res., 90, 9736–9748; doi:10.1029/JA090iA10p09736. Meng, C.‐I. (1981), Electron precipitation in the midday auroral oval, J. Geophys. Res., 86, 2149–2174; doi:10.1029/ JA086iA04p02149. Moore, T. E., and G. V. Khazanov (2010), Mechanisms of ionospheric mass escape, J. Geophys. Res., 115, A00J13; doi:10.1029/2009JA014905. Moore, T. E., and J. L. Horwitz (2007), Stellar ablation of planetary atmospheres, Reviews of Geophysics, 45; doi:10.1029/ 2005RG000194. Moore, T. E., et al. (1997), High‐altitude observations of the polar wind., Science, 277, 349–351; doi:10.1126/science.277.5324.349. Mozer, F. S., C. W. Carlson, M. K. Hudson, R. B. Torbert, B. Parady, J. Yatteau, and M. C. Kelley (1977), Observations of paired electrostatic shocks in the polar magnetosphere, Phys. Rev. Lett., 38, 292–295; doi:10.1103/PhysRevLett.38.292. Newell, P. T., A. R. Lee, K. Liou, S.‐I. Ohtani, T. Sotirelis, and S. Wing (2010), Substorm cycle dependence of various types of aurora,J.Geophys.Res.,115,A09226;doi:10.1029/2010JA015331. Newell, P. T., C.‐I. Meng, D. G. Sibeck, and R. Lepping (1989), Some low‐altitude cusp dependencies on the interplanetary magnetic field, J. Geophys. Res., 94, 8921–8927; doi:10.1029/ JA094iA07p08921. 284 DAWN-DUSK ASYMMETRIES IN PLANETARY PLASMA ENVIRONMENTS Nilsson, H., E. Engwall, A. Eriksson, P. A. Puhl‐Quinn, and S. Arvelius (2010), Centrifugal acceleration in the magnetotail lobes, Ann. Geophys., 28, 569–576. Noda, H., et al. (2003), Tail lobe convection observed by Cluster/EDI, J. Geophys. Res., 108, 1288; doi:10.1029/ 2002JA009669. Ohtani, S., S. Wing, G. Ueno, and T. Higuchi (2009), Dependence of premidnight field‐aligned currents and particle precipitation on solar illumination, J. Geophys. Res., 114, A12205; doi:10.1029/2009JA014115. Østgaard, N., S. B. Mende, H. U. Frey, J. B. Sigwarth, A. Åsnes, and J. M. Weygand (2007), Auroral conjugacy studies based on global imaging, J. Atmos. Solar Terr. Phys., 69, 249–255; doi:10.1016/j.jastp.2006.05.026. Pfaff, R., et al. (1998), Initial FAST observations of acceleration processes in the cusp, Geophys. Res. Lett., 25, 2037–2040; doi:10.1029/98GL00936. Robinson, R. M., R. R. Vondrak, K. Miller, T. Dabbs, and D. Hardy (1987), On calculating ionospheric conductances from the flux and energy of precipitating electrons, J. Geophys. Res., 92, 2565–2569; doi:10.1029/JA092iA03p02565. Seki, K., et al. (2003), Cold ions in the hot plasma sheet of Earth’s magnetotail, Nature, 422, 589–592. Seo, Y., J. L. Horwitz, and R. Caton (1997), Statistical relationships between high‐latitude ionospheric F region/topside upflows and their drivers: DE 2 observations, J. Geophys. Res., 102, 7493–7500; doi:10.1029/97JA00151. Siscoe, G., N. Crooker, K. Siebert, N. Maynard, D. Weimer, and W. White (2005), Cusp geometry in MHD simulations, Surv. Geophys., 26, 387–407; doi:10.1007/s10712‐005‐1902‐4. Strangeway, R. J., C. T. Russell, C. W. Carlson, J. P. McFadden, R. E. Ergun, M. Temerin, D. M. Klumpar, W. K. Peterson, and T. E. Moore (2000), Cusp field‐aligned currents and ion outflows, J. Geophys. Res., 105, 21,129–21,142; doi:10.1029/ 2000JA900032. Strangeway, R. J., R. E. Ergun, Y.‐J. Su, C. W. Carlson, and R. C. Elphic (2005), Factors controlling ionospheric outflows as observed at intermediate altitudes, J. Geophys. Res., 110(A9); doi:10.1029/2004JA010829. Su, Y.‐J., J. L. Horwitz, T. E. Moore, B. L. Giles, M. O. Chandler, P. D. Craven, M. Hirahara, and C. J. Pollock (1998), Polar wind survey with the Thermal Ion Dynamics Experiment/ Plasma Source Instrument suite aboard POLAR, J. Geophys. Res., 1032, 29,305–29,338; doi:10.1029/98JA02662. Walsh, B. M., T. D. Phan, D. G. Sibeck, and V. M. Souza (2014), The plasmaspheric plume and magnetopause reconnection, Geophys. Res. Lett., 41, 223–228; doi:10.1002/2013GL058802. Wing, S., et al. (2014), Review of solar wind entry into and transport within the plasma sheet, Space Sci. Rev., 184, 33–86; doi:10.1007/s11214‐014‐0108‐9. Wing, S., P. T. Newell, and J. M. Ruohoniemi (2001), Double cusp: Model prediction and observational verification, J. Geophys.Res.,106,25,571–25,594;doi:10.1029/2000JA000402. Wing, S., P. T. Newell, and T. G. Onsager (1996), Modeling the entry of magnetosheath electrons into the dayside‐ionosphere, J. Geophys. Res., 101, 13,155–13,168; doi:10.1029/96JA00395. Wing, S., M. Gkioulidou, J. R. Johnson, P. T. Newell, and C.‐P. Wang (2013), Auroral particle precipitation characterized by the substorm cycle, J. Geophys. Res., 118, 1022–1039; doi:10.1002/jgra.50160. Wing, S., S.‐I. Ohtani, P. T. Newell, T. Higuchi, G. Ueno, and J. M. Weygand (2010), Dayside field‐aligned current source regions, J. Geophys. Res. Space Physics, 115, A12215; doi:10.1029/2010JA015837. Woch, J., and R. Lundin (1992), Magnetosheath plasma precipitation in the polar cusp and its control by the interplanetary magnetic field, J. Geophys. Res. Space Physics, 97(A2), 1421–1430; doi:10.1029/91JA02487. Yau, A. W., and M. Andre (1997), Sources of ion outflow in the high latitude ionosphere, Space Sci. Rev., 80, 1–25; doi:10.10 23/A:1004947203046. Yau, A. W., B. A. Whalen, W. K. Peterson, and E. G. Shelley (1984), Distribution of upflowing ionospheric ions in the high‐altitude polar cap and auroral ionosphere, J. Geophys. Res., 89, 5507–5522; doi:10.1029/JA089iA07p05507. Zhang, B., W. Lotko, O. Brambles, M. Wiltberger, and J. Lyon (2015), Electron precipitation models in global magnetosphere simulations, J. Geophys. Res., 120, 1035–1056; doi:10.1002/2014JA020615. Zhang, B., W. Lotko, O. Brambles, P. Damiano, M. Wiltberger, and J. Lyon (2012), Magnetotail origins of auroral Alfvénic power,J.Geophys.Res.,117,A09205;doi:10.1029/2012JA017680.