Time domain structures and dust in the solar vicinity: Parker Solar Probe observations

On April 5, 2019, while the Parker Solar Probe was at its 35 solar radius perihelion, the data set collected at 293 samples/sec contained more than 10,000 examples of spiky electric-field-like structures having durations less than 200 milliseconds and amplitudes greater than 10 mV/m. The vast majority of these events was caused by plasma turbulence. Defining dust events as those having similar, narrowly peaked, positive, single-ended signatures, resulted in finding 135 clear dust events, which, after correcting for the low detection efficiently, resulted in an estimate consistent with the 1000 dust events expected from other techniques. Defining time domain structures (TDS) as those having opposite polarity signals in the opposite antennas resulted in finding 238 clear TDS events which, after correcting for the detection efficiency, resulted in an estimated 500-1000 TDS events on this day. The TDS electric fields were bipolar, as expected for electron holes. Several events were found at times when the magnetic field was in the plane of the two measured components of the electric field such that the component of the electric field parallel to the magnetic field was measured. One example of significant parallel electric fields shows the negative potential that classified them as electron holes. Because the TDS observation rate was not uniform with time, it is likely that there were local regions below the spacecraft with field-aligned currents that generated the TDS.

Unified Astronomy Thesaurus concepts: Interplanetary dust (821); Solar electromagnetic emission (1490); Solar coronal waves (1995) Time domain structures (TDS) are several millisecond duration, intense, electron scale, and electric field spikes that have significant components parallel to the local magnetic field Hutchison 2017). At least five different types of TDS exist, including electrostatic and electromagnetic double layers, electrostatic and electromagnetic electron holes, and nonlinear whistlers (Drake et al. 2015;Mozer et al. 2015;Agapitov et al. 2018). Electron hole TDS were first studied in numerical models of the instability of two electron beams (Roberts & Berk 1967;Morse & Nielson 1969a, 1969b. Saeki et al. (1979) first observed their generation in the lab in a Q-machine. Generation by spontaneous reconnection in a toroidal device was studied (Fox et al. 2008), and laboratory observations of electron holes in a plasma device have been presented (Lefebvre et al. 2011). Double layer-type TDS were first discussed in connection with magnetospheric physics and astrophysics by the Stockholm group under Hannes Alfvén (Alfvén & Carlqvist 1978, and references therein;Raadu 1989), and they were first observed in the magnetosphere along auroral zone magnetic field lines by the S3-3 satellite (Mozer et al. 1977;Temerin et al. 1982). They were more thoroughly studied on the Fast Auroral Snapshot (FAST) mission (Ergun et al. 1998a). They have been seen in the magnetospheric tail (Matsumoto et al. 1994;Franz et al. 1998;Streed et al. 2001), the plasma sheet (Ergun et al. 2009;Deng et al. 2010), at shocks (Bale et al. 1998;Cattell et al. 2003), at magnetic field reconnection sites (Cattell et al. 2002;Mozer & Pritchett 2009;Khotyaintsev et al. 2010;Li et al. 2014), in the solar wind (Bale et al. 1996;Williams et al. 2005;Malaspina et al. 2013a), and at Saturn (Williams et al. 2006). They have been invoked to explain particle heating, scattering, and acceleration (Ergun et al. 1998b;Bale et al. 2002;Cranmer & van Ballegooijen 2003;Mozer et al. 2016;Vasko et al. 2017Vasko et al. , 2018. Dust impacts have been observed on many spacecraft in the solar wind since they were first found in Saturnʼs rings (Aubier et al. 1983;Gurnett et al. 1983;Meyer-Vernet et al. 2009, 2017Malaspina et al. 2013b;Zaslavsky 2015;Vaverka et al. 2018). Szalay et al. (2020) and Page et al. (2020) describe properties of dust striking the Parker Solar Probe (PSP). Their dust data was obtained from microsecond resolution, infrequently recorded bursts of fields data. Another view of the same phenomenon is obtained by analyses, reported here, of lower time resolution (293 samples/s) full-time coverage of the spikes observed in the electric field detector on the PSP. An additional result of this analysis is the discovery of abundant TDS and the determination of their properties. The electric field detector is described elsewhere (Bale et al. 2016). Figure 1 presents >10 Hz filtered data collected by the Fields Digital Fields Board (DFB) waveform receiver at 293 samples/s on 2019 April 5 and at the 35 solar radius perihelion of the spacecraft orbit. The data are in the spacecraft coordinate system, which has X perpendicular to the Sun-spacecraft line, in the ecliptic plane, and pointing in the direction of solar rotation (against the ram direction); Y perpendicular to the ecliptic plane, pointing southward; and Z pointing sunward. The Z-component of the electric field was not measured. The four single-ended potentials V1, V2, V3, and V4, whose differences (V1-V2) and (V3-V4) determine the X and Y components of the electric field, are in a plane parallel to the satellite heat shield. In Figure 1, there are 9800 10-20 mV m −1 amplitude electric field spikes with durations <200 milliseconds, 750 20-30 mV m −1 spikes, and 760>30 mV m −1 spikes. By far, the largest fraction of these events resulted from differences in sensor floating potentials driven by density fluctuations of the type illustrated in Figure 2. Panels 2(a) and 2(b) of Figure 2 give the apparent electric field measured due to potential differences between the single-ended antenna measurements V1 (not shown because it was measured at a too-low frequency) and V2 (panel 2(c)), as well as V3 (panel 2(d)) and V4 (panel 2(e)). Although the single-ended potentials all have the same shape, the ∼20% differences in their 100 millivolt amplitudes produces the apparent 20 mV m −1 electric field seen in panels 2(a) and 2(b). The task in the following data analysis is to distinguish between these turbulent electric fields, those produced by dust, and those produced by TDS. Figure 3 illustrates a dust impact seen simultaneously in the sub-microsecond resolution channels (panels 3(a)-3(d)), and the 293 samples/s channels (panels 3(e)-3(j)). As seen in this figure, a dust hit produces a <0.1 msec electric field spike in the singleended potentials V1, V2, V3, and V4, when these voltages are measured with sufficiently high time resolution (panels 3(a)-3 (d)). This spike results from the rapid release of spacecraft ions by the impact (Meyer-Vernet et al. 2009). Because this release causes the potential of the spacecraft to change, V1-V4 experience similar changes because their measured quantity is the potential of the antenna minus the potential of the spacecraft. The system recovers from this dust hit in the time required to charge the spacecraft capacitance to its previous condition. This charging of the spacecraft capacitance depends on the spacecraft shape and the plasma thermal current, but it is typically of the order of one millisecond. This signal is processed by electronics that produce an overshoot for two reasons. First, the data is AC coupled, which results in the average signal being zero. So the spike must be followed by a signal of the opposite sign and  . Panels 2(a) and 2(b) give the apparent electric field measured by the potential differences between the single-ended measurements V1 (not measured) and V2 (panel 2(c)), as well as V3 (panel 2(d)) and V4 (panel 2(e)). Although the single-ended potentials have the same shape, their ∼20% differences in their 100 millivolt amplitudes produces the apparent 20 mV m −1 electric field seen in panels 2(a) and 2(b). Figure 3. Panels 3(a)-3(d) give three milliseconds of high time resolution single-ended voltages measured on the four individual antennas during a dust impact. Significantly, the signals on all of the antennas are essentially identical. Panels 3(e) and 3(f) give 100 milliseconds of the <10 mV m −1 electric field signals measured at 293 Hz for the dust event and panels 3(g)-3(j) give the four antenna signals, which are also essentially identical. identical area. The shape and duration of this overshoot depends on the electronics. Second, if the spike is sufficiently large, it saturates the electronics, which then take longer to recover, depending again on the electronic design. Panels 3(e)-3(j) are the high-rate data of panels 3(a)-3(d) after passing through a 100 Hz low-pass filter associated with the lower data rate. This filter both attenuates and spreads the input signal in time due to the fact that the low-rate channel cannot see components of the input signal faster than ∼10 msec. For this reason, a millisecond dust pulse generally appears in the 293 samples/s, single-ended voltage channels as a single point peak in a >10 msec signal with a following overshoot whose duration depends on the amplitude of the signal and the electronic design.
The single-ended potentials in panels 3(g)-3(j) are similar in shape but different enough in amplitude to produce the apparent 10 mV m −1 electric field in panels 3(e) and 3(f). To understand how general the data of Figure 3 are, 104 dust spike plots of V2-V4 (V1 was not measured), measured at 293 Hz at times when known dust pulses were simultaneously observed in the sub-microsecond resolution data, were studied. They show that, in 72% of the cases, the single-ended signals were like those illustrated in panels 3(h)-3(j), lasting ∼100 milliseconds, although the dust signature at a high time resolution (panels 3(a)-3(d)) lasted ∼1 millisecond. While there are dust impacts that do not fit this description, this empirical result is used to define dust events in the 293 Hz DFB channel as those for which all the positive, sharply peaked, and similarly shaped but different amplitude single-ended potentials with <200 millisecond durations, such as those of panels 3(h)-3(j), are observed. This resulted in a detection efficiency of 72%. Previous observations of dust impacts on other spacecraft differ from this definition because such data were measured at different data rates, involved other than the several single-ended potentials, did not involve current-biased antennas, and involved electronics with different saturation properties from those on the PSP. Figure 4 presents another dust event observed at 293 samples/s. As expected from the above discussion, the three single-ended voltages are positive, peaked at a single point, and somewhat different in amplitude (panels 4(c)-4(e)). In this case, the apparent electric field of panel 4(b) was 1000 mV m −1 , which is difficult to understand unless the input pulse saturated the low frequency electronics. As seen in the two dust examples of Figures 3 and 4, the apparent electric fields due to dust can be small (∼10 mV m −1 ) or, occasionally, larger than 1000 mV m −1 . Because the dust duration is defined as the time interval during which the 293 Hz signal exceeds a fixed voltage threshold, the observed durations depend mainly on the amplitude, and the duration of the overshoot, which is discussed above. For example, the differential signal in Figures 3(e) and (f) is above a 5 mV m −1 threshold for about 15 msec while the signal in Figures 4(a) and (b) is above the same threshold for more than 100 msec. Figure 5 presents an example of an electric field signal produced by a TDS. Panels 5(a) and 5(b) of this figure give the electric field components in the X-Y plane, and the remaining panels give the single-ended voltages V2, V3, and V4. What is uniquely different from dust pulses in this example is that the three single-ended voltages are not similar in form. Instead, V3 and V4 have opposite polarity components, which means that there is a real electric field across the spacecraft because the antenna on one side of the spacecraft has an increasing (decreasing) voltage at the same time that the opposite antenna has a decreasing (increasing) voltage. Furthermore, the electric field signal is bipolar, which is the known structure of most of the previously observed TDS in the solar wind and elsewhere throughout the heliosphere. The TDS duration is of the order of several Debye lengths divided by the electron thermal velocity, which is ∼50 msec for the plasma parameters at the observation time. The similarity of the dust and the TDS durations is a coincidence associated with the plasma parameters for the TDS and the electronic design and data rate of the dust.
Using the requirement that dust produces similar and narrowly peaked single-ended potentials on the three antenna while TDS have opposite polarity V3 and V4 values, the 1500 examples of >20 mV m −1 electric fields observed on the day of interest were examined to determine those due to dust and those due to TDS. The result of this search is shown in Figure 6, which plots the amplitude and width of each event.  . Electric field (panels 5(a) and 5(b)) and the individual antenna signals (panels 5(c)-5(e)) for a time domain event, which is distinguished from a dust event because the signals from the opposite antennas, V3 and V4, have opposite polarity components.
As expected, dust produced both the largest and smallest amplitude events and the longest and shortest duration events because the dust signature depended on the overshoot and saturation properties of the incident signal and the electronics, as discussed earlier. This data set includes 238 TDS and 135 dust strikes.
To further characterize the TDS, the bipolar nature of each event is plotted in Figure 7 as the ratio of the amplitude of the first peak to the amplitude of the opposite polarity peak, with the sign of the ratio being the sign of the first peak. Most bipolar structures had roughly equal amplitudes of their positive and negative maxima and the negative polarity maximum appeared before the positive polarity maximum in the majority of cases. This latter result may be an artifact of the selection of TDS using only voltages V3 and V4, or it may contain information on the TDS propagation direction.
The TDS data were searched for events that occurred when B Z was close to zero because, for such events, the magnetic field was in the X-Y plane (as was the measured electric field) and the component of the electric field parallel to the magnetic field could be determined. Figures 8(a) and (b) give one example of the perpendicular and parallel electric fields measured when the magnetic field was within 3°of the X-Y plane. Panel 8(c) gives V3 and V4, which have opposite polarity components, as is required for a TDS. That V3 and V4 are dissimilar in magnitude is the result of the potential of the spacecraft changing during the event to add to one of the signals and to subtract from the other. Assuming that the TDS was moving in the −B Z direction away from the Sun, this is an example of a negative potential structure, which makes it an electron hole. Figure 9 gives the TDS observation rate versus time of day of interest, during which the spacecraft distance from the Sun    varied by less than 1 solar radius. This non-uniform distribution suggests that TDS events appeared in spatially or temporally confined regions. It also shows that this distribution was not produced by dust impacts or turbulence because they would be uniform in time.

Discussion
TDS and dust events are identified for one day at the 35 solar radius perihelion of the PSP by examining the single-ended voltages on three electric field antennas in the X-Y plane when the amplitude of the electric field exceeded 20 mV m −1 . The criterion for identifying a dust impact was that the three antennas have similar positive and narrow voltage peaks and the criterion for a TDS was that V3 and V4 had opposite polarity components. This method found 238 TDS. The detection efficiency for TDS was less than 50% both because turbulence (plasma density perturbations) can be large enough to hide the anticorrelation and because only V3 and V4 were used for identification. Considering these effects, between 500 and 1000 TDS are estimated to be present on this day. This estimate is supported by the observation of a similar number of bipolar electric field events to those in the 238 identified events, although they did not have opposite polarity V3 and V4 signals. Such events may have had opposite V1 and V2 signals.
For a 73% detection efficiency of >20 mV m −1 dust events, the 135 identified dust events correspond to a probable 185 dust events in the day of data. This estimate is low because most of the dust impacts had amplitudes <20 mV m −1 (Figure 3) and were lost in the huge number of turbulence events. The number of dust events estimated in other ways (Page et al. 2020;Szalay et al. 2020) is suggested to be about 1000. Thus, because most dust events were too low in amplitude to be detected by the detection technique in this paper, as illustrated in Figure 2, the observation of 185 events with amplitudes >20 mV m −1 seems reasonable.
A few TDS were observed at times when B Z was small, such that the magnetic field was in the plane of the electric field measurement and the component of the electric field parallel to the magnetic field was measured. Under the assumption that the TDS was moving away from the Sun, such structures were electron holes, which is in agreement with TDS properties in the magnetosphere (Mozer et al. , 2016Vasko et al. 2017Vasko et al. , 2018 and from numerical simulations (Drake et al. 2015) Because the TDS appeared in spatially confined regions and because electron holes are likely associated wih field-aligned currents Hutchison 2017), this data suggests the presence of regions containing field-aligned currents at lower altitudes. While electron holes can significantly heat and accelerate electrons (Ergun et al. 1998b;Bale et al. 2002;Cranmer & van Ballegooijen 2003;Mozer et al. 2016;Vasko et al. 2017Vasko et al. , 2018, they may not occur in sufficient numbers at 35 solar radii to cause a local effect. However, closer to the Sun, in regions that will be examined in later orbits, their numbers might increase to levels of interest for the electron dynamics.