The Solar Orbiter merged magnetic field dataset

(Solar Orbiter Nugget #75 by M. Kretzschmar¹, J.-Y. Brochot¹, T. S. Horbury², K. Rackovic¹, M. Maksimovic³, O. Alexandrova³, X. Bonnin³, G. Jannet¹, H. O’Brien², A. Crabtree², J. Morris², V. Krasnoselskikh^1,5, T. Dudok de Wit^1,6, O. Le Contel⁴, T. Chust⁴, P.-M. De Gelis¹, L. Da Silva Gonçalves¹ and E. Fauchon-Jones^2,7)

Introduction

Like other space missions, Solar Orbiter embarks two experiments to fully measure the varying magnetic field of the solar wind in-situ: the fluxgate magnetometer MAG [1], which is best suited for frequencies from DC to a few Hertz, and the search coil magnetometer SCM [2,3], which is best suited for frequencies above a few Hertz.

Measuring the magnetic field from DC to high frequencies is essential, for instance, in the case of turbulence and waves investigations, requiring analyses over a very broad range of frequencies [4,5]. In particular, the transition of turbulence from large magnetohydrodynamic (MHD) scales to kinetic, sub-ion scales and down to even smaller electron scales is still a matter of debate [6,7,8,9].

Studying these processes with data from the Solar Orbiter mission currently requires retrieving data from both magnetic sensors, plotting them together and deciding, depending on the frequency range of interest, which one to consider or where and how to transition from one to the other, eventually taking care of inter-calibration issues. To ease the analysis of Solar Orbiter in-situ data over different scales, we have built a merged magnetic field dataset that keep the best of the two instruments and that is available through the Solar Orbiter Archive. This nugget contains a brief description of the merging procedure and shows some examples of how this merged magnetic field can be used to study turbulence and waves.

Merging procedure

We have first determined the inflight noise of the two instruments, as shown in Fig.1.

SCM, which is part of the RPW experience, becomes more sensitive than MAG at frequencies above 5 Hz. SCM has interferences along the X_srf axis at 1.3 Hz and harmonics. The apparent decrease of the MAG noise above 10 Hz is actually caused by the anti-aliasing filter of the instrument. 

Figure 1: Inflight noise of MAG and SCM

The merging technique consists in doing a weighted sum in Fourier space : 
B_merged(ω)= α_mag(ω) B_mag(ω) + (1-α_mag)(ω) B_scm(ω) 
Minimizing the noise of the merged signal leads to α_mag = n²_scm/(n²_scm+ n²_mag). The α_mag we use to merge the MAG and SCM data is a slightly modified version of the one obtained directly by using the noise curve, in order to reduce the SCM interference at 1.3 and to decrease the weight of MAG at frequencies above 15 Hz.

Figure 2: Magnetic field time series filtered between 3.5 Hz and 8 Hz (left), and between 10 Hz and 20 Hz (right), as measured by MAG (red), SCM (black), and the merged product (cyan).

Next, we have intercalibrated the two instruments. Figure 2 shows time series of the magnetic fields as measured by MAG and SCM. At low frequencies, the two instruments agree very well, but it can be noticed that MAG data are slightly delayed with respect to SCM, and that SCM amplitudes are slightly underestimated at low frequencies. At higher frequencies, MAG is still delayed and MAG amplitudes are underestimated because of the MAG anti-aliasing filter.  This needs to be corrected before performing the merging, and for this we have determined the complex transfer function from one to the other instrument, in order to apply it before the merging. In short, the SCM amplitudes are multiplied by 1.144 to match the MAG amplitude at 1 Hz (therefore the DC magnetic field measured my MAG is unchanged in the merged data). Then the MAG amplitude and phase above a few Hz are changed in order to correct for the MAG antialiasing filter and to correct for the MAG delay. The overall merging procedure involves the following steps:

• Synchronize SCM and MAG time series, using SCM times as reference such that the time of the merged magnetic field will be the same as the one of the electric field.
• Perform FFT of the SCM and MAG time series.
• Increase SCM amplitudes by +14.4% and modify MAG amplitudes and phases with the transfer function
• Merge the time series using the weights determined from the respective inflight noise of the two instruments
• Go back in temporal domain

The process is done by blocks with 80% overlapping to avoid jumps. Whenever one MAG or SCM L2 data file is changed on the Solar Orbiter Archive, a new merged file is automatically regenerated. 
 

Results and availability of the merged magnetic field

The merged magnetic field is represented on Fig.2 with the SCM and MAG time series. At low frequencies, the amplitudes match the MAG ones and the MAG delay is corrected. At higher frequencies, the merged magnetic field resemble the SCM data. 

Figure 3: Top left: power spectral density of the magnetic field computed for September 6 2022 between 2:30pm and 3:30pm, using L2 MAG burst data (black), L2 SCM data (orange), and the merged magnetic field (red). The purple curve is the noise level of the merged product. The f^−5/3 and f^−8/3 scaling are indicated for visual comparison. Bottom left: power spectral density for the same day, but between 21:30 and 22:30. Top right: time series, power spectral density, and reduced helicity σ_r of the merged magnetic field at 4096 Hz on September 4 2022. Bottom right: two power spectral densities computed from the above panel with and without waves. Noise level is indicated in grey. Scaling of f^−5/3 and f^−8/3 are indicated for visual comparison.

Figure 3 compares spectra of the merged magnetic field with spectra obtained from MAG and SCM separately. Top left panels shows that the merged magnetic field spectra is identical to the MAG one at low frequencies and to the SCM one at high frequencies, where MAG reaches its noise floor. Bottom left panel shows similar spectra for a higher level of turbulence. The MAG spectrum is lower than the merged and SCM ones because of the effect of the anti-aliasing filter. The right column shows an example of merged magnetic field with a sampling frequency of 4096 Hz, revealing that the turbulence and waves can be analyzed over a broad range of frequencies (more than five orders of magnitude here).

The merged magnetic field data product is distributed as Common Data Format (CDF) files on the Solar Orbiter Archive, with a sampling frequency of either 256 Hz or 4096 Hz and in either the SRF or RTN reference. It is a daily level 3 data product, and the filename starts with solo_L3_multi-mag-rpw-scm-merged. On average, the merged magnetic field at 256 Hz exists for 8h/day and the one at 4096 Hz for about 20 minutes/day.

A Python code to read the data and produce plots like the ones shown on the right column of Figure 3, both with merged magnetic field at 256 Hz and 4096 Hz, is available on this webpage, where one can also find the noise level of the merged data product.
 

Conclusion

Studying dynamic processes in the solar wind like waves and turbulence requires the analysis of the magnetic field over a broad range of frequencies. We have used calibrated data from the MAG and RPW/SCM instruments to produce a validated merged magnetic field available for the community though the Solar Orbiter Archive. The MAG amplitudes are corrected for its anti-aliasing filter starting at frequency above 2 Hz, with a factor increasing smoothly to match and follow the SCM amplitudes around 10 Hz. Above that frequency, the merged magnetic field resemble the magnetic field measured by SCM.  The timestamps of the merged magnetic field are the same than for the electric field data of RPW. A Python procedure to read and plot the merged magnetic field data as well as the inflight sensitivity (noise curve) of the merged magnetic field are available on the LPC2E Solar Orbiter webpage.

This work has been published in Kretzschmar et al., 2025 : https://doi.org/10.1051/0004-6361/202554731
 

Affiliations

(1) LPC2E UMR7328, OSUC-University of Orléans-CNRS-CNES, 3A avenue de la recherche scientifique, Orléans, France 
(2) Department of Physics, Imperial College London, SW7 2AZ London, UK 
(3) LESIA, Observatoire de Paris, Université PSL, CNRS, Sorbonne Université, Université de Paris, Meudon, France 
(4) LPP UMR7648, CNRS, Ecole Polytechnique, Sorbonne Université, Observatoire de Paris, Université Paris-Saclay, Palaiseau, Paris, France 
(5) Space Sciences Laboratory, University of California, Berkeley, CA, USA 
(6) International Space Science Institute, Bern, Switzerland 
(7) Department of Engineering Science, University of Oxford, Parks Road, Oxford OX1 3PJ, UK 

References

[1] Horbury et al., 2020 “The Solar Orbiter magnetometer”, DOI: 10.1051/0004-6361/201937257
[2] Jannet et al., 2021 “Measurement of magnetic field fluctuations in the parker solar probe and solar orbiter missions”, DOI: 10.1029/2020JA028543
[3] Maksimovic et al., 2020 “The Solar Orbiter Radio and Plasma Waves (RPW) instrument”, DOI: 10.1051/0004-6361/201936214
[4] Verscharen et al., 2019 “The multi-scale nature of the solar wind” DOI: 10.1007/s41116-019-0021-0
[5] Bruno & Carbone, 2013 “The Solar Wind as a Turbulence Laboratory” DOI: 10.12942/lrsp-2013-2
[6] Chen et al., 2010 “Anisotropy of Solar Wind Turbulence between Ion and Electron Scales”, DOI: 10.1103/PhysRevLett.104.255002
[7] Alexandrova et al., 2009 “Universality of Solar-Wind Turbulent Spectrum from MHD to Electron Scales” DOI: 10.1103/PhysRevLett.103.165003
[8] Sahraoui et al., 2009 “Evidence of a Cascade and Dissipation of Solar-Wind Turbulence at the Electron Gyroscale” DOI : 10.1103/PhysRevLett.102.231102
[9] Bowen et al., 2024 “Mediation of collisionless turbulent dissipation through cyclotron resonance”, DOI: 10.1038/s41550-023-02186-4

Acknowledgements

We acknowledge the support from the Centre National d’Etudes Spatiales (CNES, French space agency) for this study and the development, calibration, and operation of the search coil magnetometer of RPW. Solar Orbiter magnetometer (MAG) operations are funded by the UK Space Agency (grant ST/T001062/1).