J. Geophys. Res., 101, 17,313-17,324, 1996
(Received October 30, 1995; revised March 15, 1996; accepted March 21, 1996)
Copyright 1996 by the American Geophysical Union.
Paper number 96JA00927.
Plasma waves, observed above the dayside Venus ionosphere by the Pioneer Venus Orbiter (PVO) plasma wave instrument, have been attributed to whistler mode waves, lower hybrid waves, or ion acoustic waves. In order to clarify the nature of the waves, we have performed both case study and statistical analyses of the plasma wave and magnetic field data. We find that the plasma wave data are well ordered by altitude with respect to the location where the PVO Langmuir probe, or orbiter electron temperature probe (OETP), measures a density of 100 cm, known as the OETP ionopause. The dominant signature in the wave data appears to be a change in the instrument noise level because of changes in the plasma Debye length. However, there is a burst of wave activity near the OETP ionopause. Also, we find that there is a rotation in the magnetic field at or near this location. By casting the magnetic field data into a coordinate system ordered by the presumed magnetosheath flow, we find that the rotation of the field tends to orient the field in a more flow-aligned direction at lower altitudes. We attribute this to mass loading at lower altitudes. We further suggest that the field-aligned current associated with the field rotation corresponds to a shear Alfvén wave standing in the magnetosheath flow. The field-aligned currents are present because of boundary conditions imposed on the flow, and it is not clear if the waves are actually associated with the field-aligned currents or are simply coincidental. Since the waves are observed at the OETP ionopause, further progress in understanding these waves will be made though determining what underlying plasma structure, if any, is related to the OETP ionopause, which is defined by a specific instrument threshold. Nevertheless, our study confirms that the wave activity, field-aligned currents, and OETP ionopause all occur within the plasma mantle above the ionosphere. As such, the plasma waves are not an energy source for the dayside ionosphere.
The ionosphere forms the principal obstacle to the solar wind for an unmagnetized planet, allowing direct interaction between the solar wind and ionospheric plasma. At Venus, part of this interaction occurs in a region known as the plasma mantle [Spenner et al., 1980] , which lies above the dayside ionopause of Venus. The mantle contains a mixture of ionospheric and solar wind plasma. VLF and ELF plasma waves are observed in this region by the Pioneer Venus Orbiter (PVO) [Crawford et al., 1993] , and it appears likely that these waves are a consequence of the mixed plasmas. A similar region is found at Mars [Nagy et al., 1990; Sagdeev et al., 1990] , although it is referred to as the planetosphere rather than the mantle, and plasma waves are also observed within this region [Grard et al., 1991] .
At Venus the VLF and ELF waves were originally thought to be whistler mode waves generated at the bow shock, propagating through the magnetosheath to be absorbed within the ionosphere [e.g., Scarf et al., 1979, 1980a; Taylor et al., 1979]. However, Szegö et al.  pointed out that the magnetic field near the ionopause (in the mantle) is draped over the ionopause. Since whistler mode waves tend to carry energy parallel to the ambient magnetic field, very little wave energy is consequently transported to the ionosphere. Instead, Szegö et al.  suggested that the waves were lower hybrid waves generated by the relative drift between ions of planetary and solar wind origin.
Unfortunately, lower hybrid waves also carry most of their wave energy flux parallel to the ambient field [Strangeway and Crawford, 1993]. Moreover, thermal effects, such as ion and electron Landau damping, were neglected by Szegö et al. . Thus Strangeway and Crawford  concluded that an ion acoustic-like instability [Taylor et al., 1981] may best explain the waves. Huba  followed up on this suggestion and found that an ion-acoustic instability could be generated by the relative drift between shocked solar wind electrons and cold planetary oxygen ions. The group velocity of these waves is (nT/nm) 10 km s for typical mantle values, where n is the oxygen ion density, T is the magnetosheath electron temperature, n is the electron density, and m is the ion mass. Electrostatic ion acoustic waves have lower energy density than electromagnetic whistler mode waves for the same electric field amplitude. This, combined with the low group velocity, implies that ion acoustic waves would not transport significant energy from the mantle to the ionosphere.
More recently, however, Shapiro et al.  have argued that cold electrons of planetary origin would tend to quench the ion acoustic instability discussed by Huba . Instead, they suggested that the modified two-stream instability generates waves in two branches. One, the lower hybrid (or hydrodynamic) branch, is below the lower hybrid resonance frequency and is not detectable by the Pioneer Venus wave instrument. The other, which they refer to as the kinetic branch (absent in the earlier analysis of Szegö et al. ) would be detectable by PVO. It should be noted, however, that this so-called kinetic branch appears to be simply the whistler mode resonance cone. Inspection of equation (11) and Figure 6 of Shapiro et al.  shows that for this branch, (m/m)(k/k) cos, where is the wave frequency, m and m are the proton and electron masses, respectively, is the electron gyrofrequency, is the lower hybrid frequency (= (m/m)), k is the wave vector, is the angle of the wave vector with respect to the magnetic field, and k = kcos. For this mode the assumption that << is no longer valid, and kinetic effects due to gyroresonance, as well as Landau damping by the cold (planetary) electrons, should be included. The group velocity of both wave modes considered by Shapiro et al.  is primarily along the magnetic field and hence mainly tangential to the ionopause.
It therefore appears unlikely that the waves observed in the Venus plasma mantle are a significant source of energy for the dayside Venus ionosphere, regardless of the wave mode considered. However, the waves may play an important role within the mantle. If they are generated by the relative drift between magnetosheath and planetary ions, they may provide momentum coupling and also cause local heating of the planetary ions. On the other hand, Crawford et al.  showed that the waves are often associated with field-aligned currents within the mantle region. It is therefore possible that the waves are generated by the field-aligned currents and may be a source of anomalous resistivity within the plasma. It is the purpose of this paper to determine, from a statistical viewpoint, where the waves occur within the mantle and what other phenomena are associated with the waves. This will hopefully provide an initial step in constraining the various hypotheses put forward to explain the waves.
In section 2 we will discuss the different definitions of the ionopause at Venus. We will show that the ionopause as defined by the Langmuir probe (the orbiter electron temperature probe (OETP) ionopause) orders the plasma wave data. Of the various ionopause definitions the OETP ionopause is at highest altitude and generally lies within the plasma mantle, as defined by Spenner et al. . In section 3 we will describe a coordinate system used to order the magnetic field data. This coordinate system, which is aligned along the radial and assumed magnetosheath flow directions, allows us to show that the magnetic field is deflected above the ionosphere, in a manner consistent with mass loading at lower altitudes. In section 4 we will present a statistical analysis that further supports the results of sections 2 and 3, that the waves tend to occur at the OETP ionopause and that a field-aligned current is often observed near this ionopause. In section 5 we will study the observed magnetic field rotation in more detail, showing that the rotation can be explained as being due to field lines that are convected to lower altitudes in the subsolar region, where they are transported more slowly to the nightside than at higher altitudes. We will further show that the field rotation can be interpreted as a shear Alfvén wave standing in the magnetosheath flow. Last, in section 6 we will summarize our results. The OETP ionopause clearly orders the data, but it is not yet obvious what is the ultimate source for the waves. The next step appears to be determining the nature of the OETP ionopause.
One possible source of confusion concerning the nature of the plasma waves observed on the dayside of Venus is due to different definitions of the ionopause. Phillips et al.  show in their Figure 4 the altitude of the ionopause using different definitions. The four definitions discussed by Phillips et al.  are (1) the altitude at which the PVO Langmuir probe (orbiter electron temperature probe, OETP) measures a density of 100 cm (the OETP ionopause); (2) the altitude at which the retarding potential analyzer (ORPA) measures 100 cm (the ORPA ledge); (3) the altitude where plasma thermal pressure equals magnetic pressure (the pressure balance ionopause); and (4) the altitude where the ORPA observes a break in the density profile (the ORPA top). These different "ionopauses" are usually encountered in the order given when passing from high to low altitude. At low solar zenith angles they are close together in altitude, but they can be hundreds of kilometers apart at the terminator.
Crawford et al.  noted that the wave bursts detected by the orbiter electric field detector tended to occur at or near the OETP ionopause. This can be seen clearly in Figure 1a, 1b. In Figure 1a, 1b we show the 100-Hz wave amplitude plotted using a gray-scale representation as a function of altitude and local time. The data are taken from the Unified Abstract Data System (UADS) database for orbits 125-248, with Figure 1a showing the inbound leg for each orbit and Figure 1b showing the outbound leg. UADS was originally designed as an on-line data system [Ferandin et al., 1980] , using a common 12-s format for all instrument data. Later, the on-line system was discontinued, and UADS data were subsequently submitted by experimenters to the National Space Science Data Center, as described in the appendix to Brace and Kliore . The circles in Figure 1a, 1b give the location of the OETP ionopause (L. H. Brace, personal communication, 1993) for each orbit.
Figure 1a, Figure 1b. 100 Hz peak amplitude as a function of altitude and local time for orbits 125-248. The data are shown for (a) inbound and (b) outbound portions of each orbit. The gray scale indicates the peak amplitude per 24-s interval, with 12-s spacing between samples. The circles indicate the altitude of the orbiter electron temperature probe (OETP) ionopause.
In the UADS database the wave amplitude is given as both a peak and an average, and we have used the peak in Figure 1a, 1b. The peak amplitude allows us to more clearly discern the association between the waves and the underlying plasma structure, but some caution should be employed in interpreting the data. In particular, both the peaks and averages are calculated using a 24-s window with 12-s spacing. Thus there is a potential for aliassing of the data, especially in point-by-point comparisons. However, statistical studies will be less prone to aliassing, and Figure 1a, 1b shows that statistically the wave amplitude tends to peak at or near the OETP ionopause. Thus we will use altitude with respect to the OETP ionopause to order the data, rather than altitude from the surface of the planet. While subsequent analysis presented in this paper will show that the OETP ionopause is a useful reference altitude for ordering the data, this methodology is also supported by Crawford . He showed that for a sample of some 30 orbits the OETP ionopause altitude ordered the plasma wave data better than the = 1 altitude, which tended to occur always below the wave intensity peak altitude. Since the other ionopause definitions discussed earlier usually occur below the = 1 altitude, they will similarly be less useful in ordering the plasma wave data.
In the introduction we stated that the waves are observed within the plasma mantle. Although we refer to the altitude at which the Langmuir probe measures a density of 100 cm as the OETP ionopause, we consider this as occurring within the plasma mantle. At higher altitudes the plasma is essentially shocked solar wind plasma. At much lower altitudes the plasma is ionospheric in origin. The OETP ionopause is therefore probably within a region of mixed plasmas. Whether or not the OETP ionopause corresponds to a sharp boundary, or is simply a point on a more gradual transition, requires further analysis.
Before discussing the magnetic field and plasma wave signatures at the OETP ionopause in more detail, we note that the other striking feature in Figure 1a, 1b is the relatively low level of plasma wave noise at lower altitude. We will discuss this later. However, our main conclusion concerning the low level of wave noise is that this is mainly an instrument artifact associated with the changes in the plasma Debye length and is not due to the absorption of waves as postulated by Scarf et al. [1980a].
Besides using the OETP ionopause as a reference altitude, we will also use a coordinate system for the magnetic field data that is based on the assumed flow direction within the magnetosheath. This coordinate system, which is similar to radial-east-north coordinates, is shown in Figure 2. In deriving the coordinate system we have assumed that the magnetosheath flow is symmetric about the Venus-Sun line and is mainly tangential to the obstacle, which we further assume is spherical in shape. Thus most of the flow is assumed to be along the direction labeled "flow," with very little or no radial component, except at the subsolar point. The direction labeled "perp" completes the triad and is perpendicular to both the radial and assumed flow direction. The transformation from the Venus solar orbital (VSO) coordinate system to the "radial-flow-perp" system can be obtained as follows: If the spacecraft position is given by R in VSO, then the perpendicular direction , where is the unit vector along the X axis. The flow direction is given by . The radial direction is of course given by . The dot product of any vector in VSO with the unit vectors , , and gives the vector in the radial-flow-perp coordinates. A similar coordinate system was employed by Law and Cloutier  in their study of the magnetic field structure.
Figure 2. Relationship between flow-aligned coordinates and Venus solar orbital (VSO) coordinates.
The usefulness of this coordinate system can be seen in Figure 3, where we present high-resolution plasma wave and magnetic field data for the outbound portion of orbit 169. This orbit was also shown by Crawford et al. . The top four panels show the wave intensity for the four wave channels, while the bottom four channels show the magnetic field rotated into radial-flow-perp coordinates. The vertical line in each panel indicates the location of the OETP ionopause.
Figure 3. High resolution plasma wave and magnetic field data from orbit 169 outbound. Fifteen minutes of data are shown, with the top four panels showing wave intensity for the four wave channels, using a logarithmic scale. The bottom four panels show the magnetic field, cast into radial-flow-perp components. The vertical line in Figure 3 marks the OETP ionopause.
Before discussing the magnetic field data, we will repeat some of the points made by Crawford et al.  concerning the plasma wave data. First, the dominant change in intensity, especially at 100 Hz, is the change in instrument background. This is due to the change in plasma Debye length, which is short (a few centimeters) within the ionosphere and long (a few meters) within the magnetosheath. The antenna length is 0.76 m [Scarf et al., 1980b] , while the spacecraft is 2.54 m in diameter. Hence the antenna is within the Debye sphere of the spacecraft when the spacecraft is in the magnetosheath, and the wave instrument is sensitive to noise caused by photoemission of electrons from the spacecraft and antenna elements. In the ionosphere the antenna is effectively shielded from photoemission noise. Thus any analysis of the wave data near the dayside ionopause should take into account the change in background. In particular, the decrease in wave intensity as the spacecraft enters the ionosphere cited by Scarf et al. [1980a] is not evidence for whistler mode absorption in the ionosphere but is simply the result of the change in instrument background. However, in addition to the change in background, Figure 3 shows a wave burst at all four channels at the OETP ionopause. This burst does not display any of the characteristics of the background noise and is probably due to plasma waves generated at or near the OETP ionopause.
Turning now to the magnetic field data, Figure 3 shows that the field is quite large near the OETP ionopause. The magnetic barrier [Elphic et al., 1980] extends to altitudes well below the OETP ionopause. Throughout this region the radial component is small, and the field is mainly draped over the obstacle. However, the field direction changes just below the OETP ionopause. Above the OETP ionopause the field has large components both parallel and perpendicular to the assumed flow direction. Below the OETP ionopause the field rotates to a direction parallel to the assumed magnetosheath flow. Luhmann  discussed the presence of a magnetic field rotation within the mantle region. In her model the rotation of the field was due to convection of the field to lower altitudes, where the horizontal plasma flow was reduced. This velocity shear causes the magnetic field to be more flow-aligned at lower altitudes, resulting in the "weathervaning" of the field described by Law and Cloutier  in their analysis of magnetic field data from the dayside of Venus.
Since there is little or no change in magnetic field magnitude, the current in the magnetic shear layer is mainly field-aligned. The magnetic field is mainly horizontal, and hence the field-aligned currents are horizontal. At lower altitudes the magnetic field is shielded from the ionosphere by perpendicular currents, which also flow horizontally. We therefore assume that there are only vertical gradients in the field, and the variation of the horizontal components of magnetic field as a function of altitude will be used to calculate the current density.
Additional examples of high resolution wave and field data are given in Figures 4, 5, and 6. Each of Figures 4, 5, and 6 show a wave burst near the altitude of the OETP ionopause. In all the examples the burst is detected at 100 Hz, but sometimes the burst extends to higher frequencies. This argues against a lower hybrid wave mode and for an acoustic mode [Taylor et al., 1981; Strangeway and Crawford, 1993; Huba, 1993], since the lower hybrid resonance frequency is typically only a few tens of Hertz. Because lower hybrid waves are thought to be in resonance with ions of planetary origin, there would be little or no Doppler shift of the waves when observed in the spacecraft frame, which moves relatively slowly with respect to the planet. The spacecraft speed is 10 km s, while the shocked solar wind speed is 100 km s.
Figure 4. Wave and magnetic field data for the inbound portion of orbit 157. Similar in format to Figure 3.
Figure 5. Wave and magnetic field data for the periapsis portion of orbit 201. Similar in format to Figure 3. This orbit was also discussed by Luhmann .
Figure 6. Wave and magnetic field data for the outbound portion of orbit 171. Similar in format to Figure 3. This orbit was also discussed by Law and Cloutier .
In each of the examples shown in Figures 3, 4, 5, 6, the magnetic field deflects to a more flow-aligned direction at lower altitude. The tendency for the field to become more flow aligned below the OETP ionopause will be a conclusion of the statistical analysis presented in sections 4 and 5. However, it should be noted that not every orbit displays this behavior. There are orbits in which the field rotation is in the opposite sense, becoming transverse to the assumed flow direction at lower altitudes. This could be because the actual magnetosheath flow is not as assumed or because the interplanetary magnetic field changes direction. Clearly, these exceptions should be investigated, but we will defer such analysis to future studies.
In Figure 7a, 7b we show UADS data from the first three dayside periapsis seasons of the Pioneer Venus Orbiter (i.e., orbits 125-248, 345-475, and 570-700). In Figure 7a, 7b we have restricted the data to solar zenith angles (SZA) < 30°. The top panel of Figure 7a shows the peak wave amplitude binned as a function of altitude with respect to the OETP ionopause, using 25-km bins. For each channel the symbols indicate the median peak amplitude per bin, with the shaded area indicating the upper and lower quartile. Clearly, the largest signal is measured at 100 Hz, but both the 730-Hz and 5.4-kHz channels also observe a wave burst at the OETP ionopause. However, as noted earlier, the major signature associated with the OETP ionopause appears to be a change in the instrument background. For example, apart from the peak at the OETP ionopause, we would conclude that the 100-Hz background changed from 10 V m Hz within the ionosphere to 10 V m Hz within the magnetosheath. At the OETP ionopause the peak wave amplitude is 10 V m Hz.
Figure 7a, Figure 7b. Wave and plasma variations as a function of altitude with respect to the OETP ionopause for solar zenith angles (SZA) <30°. In Figure 7a, 7b and subsequently (Figures 8a, 8b and 9a, 9b) we have used data from the Unified Abstract Data System database for the first three seasons of dayside periapsis. The wave data are peak amplitudes. The plasma data are from the Langmuir probe. In calculating the plasma beta we assume that the ion temperature (T) = T/1.8 for altitudes > 350 km, where T is the electron temperature. The top panel of Figure 7a shows the wave amplitude for the four wave channels, while the bottom panel shows magnetic field strength, electron density, and plasma beta. The symbols show the median values per 25-km-altitude bin, while the shaded regions mark the upper and lower quartiles per bin. In Figure 7b we show the angle the magnetic field makes with the presumed flow direction () and the angle the field makes with respect to the vertical (). Both angles have been folded into the range 0°- 90°. The bottom panel of Figure 7b shows the parallel and perpendicular current density, calculated assuming horizontal currents and neglecting the vertical component of the magnetic field.
The bottom panel of Figure 7a shows the ambient magnetic field strength and plasma parameters. From the magnetic field profile we deduce that the OETP ionopause is probably within the magnetic barrier region; there is little change in field strength across the OETP ionopause.
The plasma beta is deduced from the electron data, assuming that the ion temperature is related to the electron temperature. Miller et al.  presented average profiles of the electron and ion temperature for the dayside ionosphere of Venus, and we have used a function attributable to J. L. Phillips to model the temperature ratio. For altitudes >350 km the ratio is constant, with the ion temperature (T) = T/1.8, where T is the electron temperature. The peak temperature ratio (T/T) is 4 at 200 km altitude. For reference, the electron temperature increases with increasing altitude, with T 0.1 eV at altitudes 160 km, 0.3 eV at 250 km, and then gradually increasing up to the OETP ionopause altitude, where T as measured by the Langmuir probe 1 eV. At lowest altitude the deduced ion temperature T 0.06 eV. Because of the enhanced T/T ratio at 200 km, T does not begin to increase until 250 km altitude. At 350 km altitude, T 0.2 eV, increasing to around 0.5 eV at the OETP ionopause. It should be remembered, however, that the ion temperature is deduced from the electron temperature and does not take into account the presence of superthermal ions often detected at or above the ionopause [Taylor et al., 1981] . Beta is generally <1 at the OETP ionopause.
By definition the density at the OETP ionopause is 100 cm, and no electron data are included in the UADS data for lower values, when the spacecraft is in sunlight. However, electron data from higher altitudes may be included if the density is >100 cm. Thus the apparent ledge in the density profile could simply be an effect of the density threshold for the Langmuir probe.
Figure 7b shows the magnetic field orientation and the current densities as a function of altitude with the respect to the OETP ionopause. From the top panel of Figure 7b we see that the field is mainly horizontal at and above the OETP ionopause. However, the orientation with respect to the presumed flow direction shows no preferred orientation as a function of altitude for this SZA range, with median value around 45° and the lower and upper quartiles at 25° and 65°. Because the top panel of Figure 7b does not show how the field changes direction along an individual orbit, we show the current density in the bottom panel of Figure 7b. As noted earlier, we assume only vertical gradients in the field. The field-aligned current indicates regions where the field rotates but does not change magnitude, while the perpendicular currents indicate where the field is shielded from the lower ionosphere. In presenting current densities in Figure 7b we have excluded data for which the vertical separation used to compute the current density is less than 20 km. This is done to exclude data near periapsis, where the spacecraft is traveling mainly horizontally, and the assumption that the observed gradients are vertical is probably not valid. At these low altitudes, fine-scale structures known as flux ropes tend to occur, further invalidating the assumption of vertical gradients.
From Figures 7a and 7b we therefore conclude that the OETP ionopause orders the wave data, with the largest-amplitude wave bursts occurring at the OETP ionopause. Furthermore, the OETP ionopause marks a region where the magnetic field rotates and field-aligned currents flow. The causal relationship, if any, between the waves and the currents has yet to be determined, but at this stage we would conclude that the currents flow in response to the magnetic field geometry imposed by the different flow regimes within the magnetosheath and ionosphere.
The peak current density in Figure 7b is 1 A m. This corresponds to a relative drift velocity of 60 km s for an electron density of 100 cm, which is the nominal plasma density at the OETP ionopause. This velocity is much smaller than the thermal velocity of the shocked solar wind protons (temperature 100 eV) and the ambient electrons, be they hot shocked solar wind (50 eV) or cold planetary (1 eV) electrons. In order to drive an instability, we expect the drift velocity of the current-carrying species to be larger than its thermal velocity. If the current is to be the source of an instability, this suggests that cold planetary electrons are the current carriers, and these electrons are a small fraction of the total density. If, for example, we assume that the electrons of planetary origin are only 3% of the total density, then the electron drift velocity 1800 km s, which is much larger than the cold electron thermal speed. This current could generate an obliquely propagating acoustic mode with phase speed of the order of the shocked solar wind sound speed but with parallel phase velocity of the order of the electron drift velocity. However, such an instability will be subject to Landau damping by the solar wind protons, unless the protons are cooler than the solar wind electrons. This is usually not thought to be the case, although Shapiro et al.  do present an example of one orbit in which the electron temperature is 100 eV. Clearly, determining whether or not the currents are a source of the waves requires detailed knowledge of the ambient ion and electron populations.
Figures 8a, 8b and 9a, 9b show plasma and field data for SZA in the range 30° < SZA < 60° and 60° < SZA < 90°, respectively. The format of Figures 8a, 8b and 9a, 9b is similar to Figure 7a, 7b, and again, current density data obtained for vertical separations less than 20 km have been excluded. As the SZA increases, the vertical scale of the plasma and field gradients also increases. Thus the current densities tend to be smaller at higher SZA. At higher SZA the change in orientation of the magnetic field becomes more clear, with the lowest occurring at the OETP ionopause. Figures 8a, 8b and 9a, 9b therefore reinforce the conclusions drawn from Figure 7a, 7b.
Figure 8a, Figure 8b. Wave and plasma variations as a function of altitude with respect to the OETP ionopause for 30° < SZA < 60°. Similar in format to Figure 7a, 7b.
Figure 9a, Figure 9b. Wave and plasma variations as a function of altitude with respect to the OETP ionopause for 60° < SZA < 90°. Similar in format to Figure 7a, 7b.
While the altitude profiles of current density and magnetic field orientation indicate that the magnetosheath magnetic field rotates at or near the OETP ionopause, the data shown in Figures 7a, 7b, 8a, 8b, and 9a, 9b do not indicate the sense of the rotation. In Figure 10 we show the orientation and the rotation of the magnetic field near the OETP ionopause. The magnetic field orientation just below the OETP ionopause is shown in the top panel of Figure 10. The data are binned in 10° bins, with the solid circles giving the median angle and the error bars marking the upper and lower quartile per bin. Noting that < 90° means that the field is pointing to higher altitude, Figure 10 shows that when the field is pointing away from the subsolar point (-90° < < 90°) the field is also pointing to higher altitude. The reverse is the case when the field points toward the subsolar point.
Figure 10. Magnetic field orientation and rotation at the OETP ionopause. The solid circles indicate the median, while the error bars mark the upper and lower quartile.
The bottom panel of Figure 10 shows the median and quartiles of the field rotation on passing through the OETP ionopause from above. The angle has been folded into the range 0°-90°, while the angle is defined as <0 if the net orientation of the field is more flow-aligned at lower altitude. Thus, for the majority of the data, the field rotates to a more flow-aligned orientation at lower altitude.
The magnetic field orientation and rotation shown in Figure 10 can be explained in terms of velocity shear across the OETP ionopause. Near the subsolar point the magnetosheath flow tends to transport magnetic flux to lower altitude, where the field is "hung up" on the obstacle, presumably because of the presence of ionospheric plasma. A flux tube that has been convected by the magnetosheath flow will be at lower altitude near the subsolar point and at higher altitude away from the subsolar point where the magnetosheath flow is more nearly tangential to the obstacle. Because the flow at higher altitudes is faster, while the subsolar portion of the flux tube moves more slowly, the flux tube will tend to be aligned along the flow as it is convected away from the subsolar region. This "weathervaning," discussed by Law and Cloutier  and modeled by Luhmann , is shown schematically in Figure 11.
Figure 11. Schematic of the magnetic field geometry in the mantle and upper ionosphere.
In Figure 11 the light gray surface marks a layer in which field-aligned currents flow. The solid arrows show the magnetosheath flow above this layer. The dark gray curves are magnetosheath flux tubes. When the flux tubes pass through the current layer, indicated by becoming darker gray, they encounter a region of reduced flow. As the flux tubes are transported away from the subsolar region, the shear in the flow causes the bend in the field to become larger, and the flux tube immediately below the current layer tends to be flow-aligned. In passing, we note that a similar process has been invoked for the formation of flux ropes [e.g., Luhmann and Cravens, 1991, and references therein], but flux rope formation occurs at lower altitude. Flux ropes are observed deep within the ionosphere below the magnetic barrier.
We can explain the field-aligned current layer as a shear Alfvén wave standing in the magnetosheath flow. This is in many ways analogous to the slow-mode standing wave observed in the terrestrial magnetosheath, as reported by Song et al. , and discussed theoretically by Southwood and Kivelson . For a shear Alfvén wave the flow perturbation (u) is related to the magnetic field perturbation (b) by
where B is the ambient magnetic field, V is the Alfvén speed = B/(), is the permeability of free space, is the mass density, and is the angle between the wave vector (k) and B. Since the wave is standing in the magnetosheath flow, k points upward, and . From Figure 10, if B points away from the subsolar point, then B also has an upward component. Thus > 0, and u is antiparallel to b. We illustrate the geometry of this situation in the two left-hand panels of Figure 12 for two orthogonal planes: the plane defined by the radial direction and the flow vector (top) and the horizontal plane (bottom). Since the horizontal component of the flow velocity (v) is also away from the subsolar point, > 0. The direction of u is such that the horizontal flow velocity is reduced below the field-aligned current layer. Taking primed vectors as being below the current layer, then , , and to first order = - . Since < 0, > , and the field is more flow-aligned below the current layer. We note that the closer alignment occurs with a change in both the field and the flow. When the magnetic field points toward the subsolar point, < 0 and < , but < 0, as illustrated in the two panels on the right-hand side of Figure 12. Again, the field is more flow-aligned below the current layer, and again, both field and flow rotate to accomplish this alignment.
Figure 12. The relationship between the magnetic field perturbation and velocity perturbation for a shear Alfvén wave standing in the magnetosheath flow. Figure 12 shows (a) magnetic field pointing away from the subsolar point, and (b) magnetic field pointing toward the subsolar point. The unprimed vectors are above the current layer, while the primed vectors are below. For each case the upper plot shows the projection in the radial-flow plane, with the lower plot showing the projection in the horizontal plane.
Figure 10 shows that the median rotation of the field is 10°, and hence b/B 0.18. From Figure 7a, B 100 nT at the OETP ionopause, while the plasma density is 100 cm. If protons are the dominant ion, then V 220 km s. Thus u 40 km s, which should be compared with a nominal sheath flow speed of 100 km s, and the flow velocity can be significantly reduced across the current layer. Increasing the amount of O present reduces V and hence u. However, an increase in O density would suggest that some slowing of the magnetosheath flow will have already occurred, and the flow velocity above the shear layer could be lower than 100 km s. Nevertheless, even if the flow is reduced because of mass loading at higher altitudes, our discussion here indicates that there is a marked reduction in flow below the OETP ionopause, presumably because of higher planetary ion densities.
Through analysis of data from individual orbits and also from statistical studies we find that the plasma wave data are well ordered by altitude with respect to the OETP ionopause, which is the altitude at which the Langmuir probe on PVO measures an electron density of 100 cm. Of the various definitions of the ionopause the OETP ionopause is usually at highest altitude. The dominant signal in the wave data is a change in the background noise of the instrument, which we attribute to changes in the plasma Debye length. At low altitudes, within the ionosphere, the wave instrument is shielded from noise due to photoelectron emission from the spacecraft. At higher altitudes, in the magnetosheath, the antenna is within the Debye sphere of the spacecraft and is probably more sensitive to photoelectron emission noise and other sources of noise on the spacecraft. Inspection of the high-resolution plasma wave data does suggest that the wave burst often observed near the OETP ionopause is due to naturally occurring waves, since the data are qualitatively different. However, some caution may be warranted since the Debye length is 0.74 m when n = 100 cm and T = 1 eV, this density and temperature being appropriate for the plasma at the OETP ionopause. Since the antenna separation is 0.76 m, a resonance between the antenna and the Debye sheath is possible.
The statistical studies show that field-aligned currents flow at or perhaps slightly below the OETP ionopause. These field-aligned currents occur above the topside density gradient and well above the perpendicular currents which mark the bottom of the magnetic barrier. On comparison with the data presented by Spenner et al. , this places the OETP ionopause and the field-aligned currents in the mantle.
The field rotation associated with the field-aligned currents tends to align the field more closely with the magnetosheath flow. It appears that the flow alignment occurs because of a shear in the magnetosheath flow [see Luhmann, 1988; Law and Cloutier, 1995] which is probably due to mass loading by ionospheric plasma at lower altitudes. The magnetic field geometry is hence dictated by the field and flow boundary conditions imposed within the magnetosheath and the ionosphere, and the field-aligned currents occur in response to this imposed geometry. We have suggested that the field-aligned current layer is a shear Alfvén wave standing in the magnetosheath flow. The plasma waves observed at the OETP ionopause may be a consequence of the field-aligned currents but are certainly not a cause.
Recently, Sauer et al.  argued that a composition boundary should be present above the ionopause of Venus and Mars. Perhaps the OETP ionopause is this boundary. However, the two-dimensional simulations of Sauer et al.  cannot explain the field rotation we have discussed here. This rotation in the field should only be present in three-dimensional simulations. We also note that the composition boundary discussed by Sauer et al. , and observed at Mars by Dubinin and Lundin , is associated with a decrease in the ambient magnetic field strength. At Venus the magnetic field strength tends to decrease at altitudes below the location of the field-aligned currents. The field rotation discussed here strongly suggests that the magnetosheath magnetic field passes into a region dominated by plasma of planetary origin (i.e., below the composition boundary), as the increased mass density would result in a velocity shear across the composition boundary.
An additional complication in determining the relationship between the various signatures observed in the mantle is the "intermediate transition" (IT) [Pérez-de-Tejada et al., 1991, 1993, 1995]. The IT is usually observed above the ionopause near or behind the terminator and is often associated with both a reduction of the magnetic field strength and a rotation of the field to a more Venus-Sun-aligned orientation. Sauer et al.  suggested that the IT is an example of the composition boundary found in their simulations. We suggest here that the transition within the field and plasma observed on the dayside evolves downstream, ultimately becoming the IT. Whether the OETP ionopause, which may be a plasma boundary, or the field-aligned current, which may mark a shear in the flow, evolves into the IT has yet to be determined.
Indeed, the relationship between the OETP ionopause and the field-aligned current is unclear. Both occur within the mantle. The mantle provides the transition from magnetosheath to ionosphere which requires a change in magnetic field orientation, marked by the field-aligned current, and a change in plasma density and composition, perhaps corresponding to the OETP ionopause. Near the subsolar point we might expect these two transitions to be close together, since a change in the plasma mass density could introduce the velocity shear that results in the Alfvén wave. However, further downstream these two signatures could separate, since the shear is carried by a standing Alfvén wave, while a mass density change could be carried by a slow-mode wave.
In conclusion, while the OETP ionopause clearly orders the magnetic field and plasma wave data, the relationship between each of these is not yet obvious. It is possible that the plasma waves are generated by the field-aligned currents which we attribute to a standing Alfvén wave associated with the shear in the plasma flow. On the other hand, if the OETP ionopause is a composition boundary, then we might expect pickup ion related instabilities to be present, be they lower hybrid [Szegö et al., 1991; Shapiro et al., 1995] or ion acoustic [Huba, 1993] . Alternatively, if the OETP ionopause corresponds to a density gradient, then perhaps gradient-drift instabilities generate the waves [cf. Huba, 1992]. Thus it is necessary to determine the nature of the OETP ionopause, which is an instrument-defined boundary, as being the altitude at which the Langmuir probe measures a density of 100 cm. Unfortunately, most of the plasma instrumentation on board the Pioneer Venus Orbiter was designed to operate optimally in a dense cold plasma, the ionosphere, or in a supersonic beam, the solar wind, but not in both. The possible exception is the orbiter retarding potential analyzer (ORPA). Regardless of the ultimate source of the waves, our analysis confirms that the waves occur within the plasma mantle and are hence not a direct source of heating for the topside ionosphere.
The Editor thanks K. Sáuer and J. D. Huba for their assistance in evaluating this paper.
Crawford, G. K., A study of plasma waves arising from the solar wind interaction with Venus, Ph. D. thesis, Univ. of Calif., Los Angeles, 1993.
Crawford, G. K., R. J. Strangeway, and C. T. Russell, VLF emissions at the Venus dayside ionopause, in Plasma Environments of Non-Magnetic Planets, edited by T. I. Gombosi, pp. 253-258, Pergamon, Tarrytown, N. Y., 1993.
Dubinin, E., and R. Lundin, Mass-loading near Mars, Adv. Space Res., 16, (4)75-4(79), 1995.
Elphic, R. C., C. T. Russell, J. A. Slavin, and L. H. Brace, Observations of the dayside ionopause and ionosphere of Venus, J. Geophys. Res., 85, 7679-7696, 1980.
Ferandin, J. A., C. L. Weeks, and R. D. Pak, Pioneer Venus unified abstract data library and quick look data delivery system, IEEE Trans. Geosci. Remote Sens., GE-18, 19-27, 1980.
Grard, R., C. Nairn, A. Pedersen, S. Klimov, S. Savin, A. Skalsky, and J. G. Trotignon, Plasma and waves around Mars, Planet. Space Sci., 39, 89-98, 1991.
Huba, J. D., Theory of small-scale density and electric field fluctuations in the nightside Venus ionosphere, J. Geophys. Res., 97, 43-50, 1992.
Huba, J. D., Generation of waves in the Venus mantle by the ion acoustic beam instability, Geophys. Res. Lett., 20, 1751-1754, 1993.
Law, C. C., and P. A. Cloutier, Observations of magnetic structure at the dayside ionopause of Venus, J. Geophys. Res., 100, 23,973-23,981, 1995.
Luhmann, J. G., A three-dimensional diffusion/convection model of the large scale magnetic field in the Venus ionosphere, J. Geophys. Res., 93, 5909-5914, 1988.
Luhmann, J. G., and T. E. Cravens, Magnetic fields in the ionosphere of Venus, Space Sci. Rev., 55, 201-274, 1991.
Miller, K. L., W. C. Knudsen, and K. Spenner, The dayside Venus ionosphere, 1, Pioneer- Venus retarding potential analyzer experimental observations, Icarus, 57, 386-409, 1984.
Nagy, A. F., T. I. Gombosi, K. Szego, R. Z. Sagdeev, V. D. Shapiro, and V. I. Shevchenko, Venus mantle - Mars planetosphere: What are the similarities and differences?, Geophys. Res. Lett., 17, 865-868, 1990.
Pérez-de-Tejada, H., D. S. Intriligator, and R. J. Strangeway, Steady-state plasma transition in the Venus ionosheath, Geophys. Res. Lett., 18, 131-134, 1991.
Perez-de-Tejada, H., D. S. Intriligator, and R. J. Strangeway, Magnetic field properties of the intermediate transition of the Venus ionosheath, Geophys. Res. Lett., 20, 991-994, 1993.
Perez-de-Tejada, H., D. S. Intriligator, and R. J. Strangeway, Intermediate transition in the Venus ionosheath, J. Geophys. Res., 100, 14,523-14,535, 1995.
Phillips, J. L., J. G. Luhmann, W. C. Knudsen, and L. H. Brace, Asymmetries in the location of the Venus ionopause, J. Geophys. Res., 93, 3927-3941, 1988.
Sagdeev, R. Z., V. D. Shapiro, V. I. Shevchenko, A. Zacharov, P. Király, K. Szegó, A. F. Nagy, and R. J. L. Grard, Wave activity in the neighborhood of the bowshock of Mars, Geophys. Res. Lett., 17, 893-896, 1990.
Sauer, K., A. Bogdanov, and K. Baumgärtel, Evidence of an ion composition boundary (protonopause) in bi-ion fluid simulations of solar wind mass loading, Geophys. Res. Lett., 21, 2255-2258, 1994.
Scarf, F. L., W. W. L. Taylor, and I. M. Green, Plasma waves near Venus: Initial observations, Science, 203, 748-750, 1979.
Scarf, F. L., W. W. L. Taylor, C. T. Russell, and R. C. Elphic, Pioneer Venus plasma wave observations: The solar wind-Venus interaction, J. Geophys. Res., 85, 7599-7612, 1980a.
Scarf, F. L., W. W. L. Taylor, and P. V. Virobik, The Pioneer Venus orbiter plasma wave investigation, IEEE Trans. Geosci. Remote Sens., GE-18, 36-38, 1980b.
Shapiro, V. D., K. Szegö, S. K. Ride, A. F. Nagy, and V. I. Shevchenko, On the interaction between the shocked solar wind and the planetary ions on the dayside of Venus, J. Geophys. Res., 100, 21,289-21,305, 1995.
Song, P., C. T. Russell, and M. F. Thomsen, Slow mode transition in the frontside magnetosheath, J. Geophys. Res., 97, 8295-8305, 1992.
Southwood, D. J., and M. G. Kivelson, On the form of the flow in the magnetosheath, J. Geophys. Res., 97, 2873-2879, 1992.
Spenner, K., W. C. Knudsen, K. L. Miller, V. Novak, C. T. Russell, and R. C. Elphic, Observation of the Venus mantle, the boundary between solar wind and ionosphere, J. Geophys. Res., 85, 7655-7662, 1980.
Strangeway, R. J., and G. K. Crawford, On the instability and energy flux of lower hybrid waves in the Venus plasma mantle, Geophys. Res. Lett., 20, 1211-1214, 1993.
Szegö, K., V. S. Shapiro, V. I. Shevchenko, R. Z. Sagdeev, W. T. Kasprzak, and A. F. Nagy, Physical processes in the plasma mantle of Venus, Geophys. Res. Lett., 18, 2305-2308, 1991.
Taylor, H. A., Jr., R. E. Daniell, R. E. Hartle, H. C. Brinton, S. J. Bauer, and F. L. Scarf, Dynamic variations observed in thermal and superthermal ion distributions in the dayside ionosphere of Venus, Adv. Space Res., 1, 247-258, 1981.
Taylor, W. W. L., F. L. Scarf, C. T. Russell, and L. H. Brace, Absorption of whistler mode waves in the ionosphere of Venus, Science, 205, 112-114, 1979.
Return to the top of the document
To R. J. Strangeway's homepage
Text and figures by R. J. Strangeway
Converted to HTML by R. J. Strangeway
Last modified: Sept. 19, 1996