The published article (Kellogg et al. 2018) presented some STEREO measurements of electric fields in the cascade and dissipation region of the solar wind and promised further work using data from the Artemis and MMS missions, which are expected to be able to measure electric fields in three dimensions in the same frequency range. This further work did not support the conclusions of Kellogg et al. (2018). We therefore retract many of the conclusions, especially the large contribution of the observed electric fields to the total electric field power spectrum. The Artemis and MMS data do show some electric fields, sometimes nearly as intense as those reported in Kellogg et al. (2018), but there are far fewer and their shapes tend to be different. Evidence is presented here that the signals were not due to electric fields but were due to density fluctuations. For the work presented in Kellogg et al. (2018), a Carrington rotation was searched for likely looking events, and about 300 were found in 9 hr of data. The data used were from the burst mode of the Low Rate Science part of the SWaves experiment. For MMS, so far, about 4.5 hr of data have been searched, and only about a dozen events found. Most of the MMS data were from MMS1 during the period 2018 March 8 0000 to March 9 0900, when MMS1 was beyond the bow shock. Some Artemis data were from a period around 2014 June 4. The instruments of Artemis and of MMS were designed, as were those of STEREO, for measuring electric fields. Both of these missions, MMS and Artemis, had their principle aims as phenomena in the magnetosphere, but the distant regions of the magnetosphere are not much different from the solar wind, the habitat for which STEREO was designed There were differences, such as the biasing of the antennas to compensate for photoemission. Biasing reduces the response to density fluctuations and only enhances the reinterpretation of the signals as density fluctuations. There is another important difference, that the STEREO had only monopole antennas pointing away from the Sun to stay out of the field of view of the optics experiments. The electric field antennas of Artemis and of MMS are operated as dipoles. Dipoles are insensitive to density fluctuations already, a discrimination against density fluctuations that is even more important than the biasing. Changes in ambient electron density in the solar wind cause changes in the potential balance between photoemission and electron pickup on an antenna. Density fluctuations in the relevant frequency range have very long wavelengths, and the changes in potential are balanced out on dipoles. The two, essentially identical, STEREO spacecraft carry three short (6 m) monopole antennas mounted on the antisunward side of the spacecraft. Their effective lengths for detection of electric fields are about 1 m. The data returned are obtained by operating the antennas as monopoles, but dipoles can be constructed by taking the differences of the potentials on pairs of antennas. This was done in Kellogg et al. (2018), but the dipole mode was used mainly as a check. As the monopole antennas on STEREO are not directly opposite, there was some concern about the accuracy of dipole construction. Further, as the monopoles have different shading by the spacecraft body, their response to density fluctuations is not identical. Some care was taken to try to calibrate out this difference. If the interpretation that the difference between the MMS and Artemis observations and those reported in Kellogg et al. (2018) is due to a failure to recognize density fluctuations, this calibration was apparently not successful. We have done some preliminary work to check these ideas. Figure 1 shows the beginning of an attempt to verify that there are variations due to density fluctuations. First is to check the potentials of a single probe. However, the probe potentials are subject to large unwanted signals, mainly due to variations of the potential of the irregular spacecraft itself as it presents various surfaces to sunlight. On the left side, a series of panels shows the potential of one of the elements of each dipole, with the previous (in time) set of rotations subtracted. That is to say, this attempts to show the variations with time after subtracting the nearly time invariant variations due to rotation (these rotation variations change slowly with time). It will be seen that these curves show some irregular variations starting at 35 s (with a large peak due to shadowing of the probe and also some regular variations probably due to conducted interference). On the right side is shown the series of differences between the two elements of the corresponding dipoles. It will be seen that much of the irregular variation has disappeared, though the procedure has failed to get rid of the large spikes due to shadowing of the antennas. We therefore attribute these irregular variations, similar to what was shown in Kellogg et al. (2018) from STEREO data, to density variations that make the same signal on both arms of the dipole. The general data used for the conviction that the STEREO measurements were taken in the free solar wind, upstream of the Earth and west of the foreshock in the case of MMS, upstream of the Moon when the Moon was also upstream of the Earth in the case of Artemis, in orbit around the Moon. Thus, the regions of measurement were as nearly the same as could be managed. Conclusions: Evidence has been presented that the signals studied in Kellogg et al. (2018) are not all due to electric fields, as was claimed there. It seems that they are sometimes due to density fluctuations, to which the monopole antennas of STEREO are sensitive, but the dipole antennas of MMS and Artemis are not. The evidence is simply that fewer events like those reported in Kellogg et al. (2018) are seen on the spacecraft of MMS and Artemis, and further that signals similar to those on STEREO are seen on the monopole antennas of MMS but are canceled when the dipole is considered. It is apparent, therefore, that many, probably most, of the signals reported in Kellogg et al. (2018) were of density fluctuations, not of electric fields. Therefore, much of the data and interpretation presented in Kellogg et al. (2018) are not measurements of electric fields. In particular, the very large contribution of the selected STEREO waveforms to the electric power spectrum of the solar wind cascade seems not to be correct. The data used in the correcting work, Artemis and MMS data, are from the the CDAWeb and are publicly available. The first author sincerely thanks Forrest Mozer, University of California at Berkeley, for discussions leading to this erratum. The authors thank CDAWeb and J. Burch, R. Ergun, and P. Lindqvist for the MMS data. This work was supported, as was the original, by NASA grants NNX16AF86G and NNX14AK73G. (Figure Presented).
Bibliographical notePublisher Copyright:
© 2019. The American Astronomical Society. All rights reserved.