General Comments and Hints for Using EFW data
- Please contact John Wygant for all RBSP EFW data requests with a copy of the e-mail to John Bonnell and Aaron Breneman. You may want to include members of the EFW team as co-authors if a significant EFW team effort is involved
- In the subject line of emails regarding EFW data, please include the phrase “RBSP EFW Data Request”. This will allow us to find the data requests which are mixed in with all our other e-mail messages.
- If you are planning to use our CDFs, please inform John Wygant of the nature of your study and which days you are looking at. This is important because there are a number of issues associated with the data that we may be aware of that will impact your study. We strongly recommend contacting us at the early stages of your work with the EFW data to avoid wasted effort. It is best not to wait until the time of submission or just before an AGU talk.
- The electric field instrument can be sensitive to the low energy plasma environment and spacecraft effects in ways that other instruments are not. Effects that can impact the data include: 1) spacecraft charging intervals; 2) wake effects; 3) time tagging problems; 4) incorrect biasing; 5) attitude maneuvers; 6) probe saturation; 7) spin harmonics and sub-harmonics; 8) changes in effective antenna lengths; 9) Earth shadowing, and 10) residual errors in the Lorentz transformation of the measured electric field from spacecraft to Earth inertial coordinates (E-Vsc x B) due to attitude and sensor coordinate system uncertainties. You might want our opinion on these matters. You should probably ask if there are any EFW team members or associated students working on the same problem or event. You might want some fraction of these people to be co-authors. There are checks on the data quality you should perform. Some of these are discussed below.
- EFW density estimates come from the spacecraft potential (V1+V2)/2, calibrated from the EMFISIS upper hybrid line. This line is typically only visible inside of or near
the boundary of the plasmasphere. Outside of this region the EFW density value cannot be directly compared to the upper hybrid line. Note that EFW density estimates in the plasmasheet, where energetic (T>10 eV) electrons are present, are not accurate and can be off by an order of magnitude. In order to identify these times see the HOPE electron data at http://www.rbsp-ect.lanl.gov/data_pub/rbspa/hope/
- Check this website periodically, the data files maybe reprocessed and improved. The last major improvement occurred on November 15, 2013. This is discussed below.
- If you plan on using our data the Acknowledgements in the publication should include: “The work by the EFW team was conducted under JHU/APL contract 922613 (RBSP-EFW).”
Comments on EFW data and CDFs
Level 3 Processed EFW data are on-line and available as level 3 CDFs on this website. This includes spin-fit electric field data in the inertial and co-rotational frame, density, and various ephemeris values. A brief discussion of the measurements and issues is also presented below. These can be opened using IDL tools and Autoplot.
Level 2 Processed EFW data are on-line and available as level 2 CDFs on this website This includes spin-fit electric field data, higher resolution electric field data, spacecraft potential measurements, spectral data, and filter bank data. These measurements are discussed in the EFW instrument paper. A brief discussion of the measurements and issues is also presented below. These can be opened using IDL tools and Autoplot.
Level 1 EFW data are accessible on this website and from the SSL Berkeley Website. The primary analysis tools are TDAS which is an IDL based series of programs and libraries. SDT is another tool used to display data in a wide variety of coordinate systems and frames of reference under different assumptions. Links to TDAS and its extensive array of downloadable software tools are available on this website.
Spin-Fit Electric Fields
The spin-fit electric field (spin period resolution) tends to be our most accurate electric field measurement. However, as always, due diligence by the user/scientists must be exercised in using these data. We provide the spin-fit electric field in both an inertial and co-rotating frame. These are discussed in the next section.
The electric field is presented in an MGSE (modified GSE) coordinate system. We provide two components (Y MGSE and Z MGSE) of the electric field. Both components are in the spin plane of the spacecraft and are measured with the long 50 m booms. The X MGSE component is along the spin-axis. The spin-axis of the spacecraft is oriented within 37 degrees of the Earth sun line. So the Y GSE direction is most nearly aligned with the Y MGSE axis and the Z MGSE direction is most nearly aligned with Z GSE.
Sometimes the spin axis electric field component is estimated using the assumption that that E • B = 0 or the parallel electric field is zero.This is used most frequently for large scale convection electric fields, MHD structures and ULF waves, and small scale waves for which perpendicular electric fields are larger than parallel. The following document discusses this calculation, the constraints on the magnetic field geometry (the ratios of |By/Bx| and |Bz/Bx|), and a very discussion of the kinds of situations the E • B assumption is likely to be appropriate.
Lorentz Transformation to an Earth-Fixed Non-Corotational Frame
The spin-fit electric field data are Lorentz transformed from the spacecraft frame to a frame at rest with the center of the Earth, giving us the inertial frame electric field. Thus it is in a non-corotating frame (i.e. not co-rotating with the Earth). To be explicit, in a direct comparison of the EFW electric field data to radar-measured electric field, the electric field must be transformed into the Earth’s co-rotational frame. This co-rotational frame electric field is also supplied in our L3 product.
To transform from the spacecraft frame to the non-corotational Earth frame we subtract the motional electric field of the spacecraft (Vsc x B), where Vsc is the orbital velocity of the spacecraft. In this frame, the electric field observed consists of the electric field due to the Earth’s rotation and the electric field due to any additional error electric fields, possibly due to attitude uncertainty, errors in the orientation of electric or magnetic booms, and EFW effective antennae length uncertainties. The residual VxB error is typically 3 mV/m (out of 300 mV/m of Vsc xB) at perigee. The error decreases with altitude faster than R3, where R is the radius of the orbit. At R>3Re, the VxB subtraction error is reduced to <0.3 mV/m.
Spacecraft Charging Issues
Below we discuss quality flags that identify the most intense intervals of spacecraft charging and earth eclipse. Spacecraft charging can occur when the spacecraft is in sunlight during storm times, when there are intense fluxes of 10eV to 1 keV electrons. The measurement is compromised during these times.
Spacecraft charging can be evaluated by checking the fifth panel down on the first page of our publically available survey plots, which can be accessed via a link on the left hand menu of the home page. This panel is labeled (V1+V2)/2. Charging above 20 volts has the potential for degrading the data. Measurements of electric fields during intervals of charging above 100 volts should be used with extreme care and never without consultation with the EFW team. Such charging tends to occur during some of the large storms in October and November 2012 and also during the large storm of March 17, 2013.
A power point presentation showing some specific examples of spacecraft charging is available here. This power point should be instructive in showing how to identify charging intervals.
A special note:
1) (V1+V2)/2 is positive when the spacecraft charges negative and vice versa. Vi=Vis-VSC where Vis is the potential of a given spin plane sensor (i=1,4) and VSC is the potential of the spacecraft. So, when the spacecraft charges negative due to intense fluxes of 10 eV-1 keV electrons, Vi can become positive.
2) Charging also occurs during Earth eclipse. These times are indicated by black vertical lines in the survey data. Actually, electric field data are compromised about 10 minutes before and after the delineated eclipse times.
To ensure the spin-fit electric field data are accurate before publishing or beginning a laborious analysis, it is worth ensuring the data in question are not subject to “wake effects.” This can be done by also downloading the V1-V6 cdf files, which we have posted. They provide 16 sample/s measurements of the electric field waveforms as the spacecraft spins. Systematically distorted spin period sine waves are evidence for a wake effect. Examples of wake effects may be found in (Bonnell et al., 2008).
Special Care for High Time Resolution Electric Fields
The EFW instrument also measures the spin plane electric field at 32 samples/s and stores it in cdf files which should be on our website soon (mid November 2013) . These data are called the “high time resolution de-spun” data. These data are presented in the MGSE system. Special care should be taken in its use. The effects discussed in Point 4 of the general comments are applicable. In addition, mismatches between the responses of the different probes can introduce error signals. A preliminary check of the quality of this data may be carried out by comparing the rotated 32 sample/s electric field in the high time resolution CDF to the spin fit data found in the spin fit CDFs. In order to carry out this comparison the high time resolution data should be averaged down to the spin period to eliminate the higher frequency turbulence.
EFW Spectogram Data
Level 2 CDFs of wave electric field and search coil wave magnetic fields spectrograms between ~40 Hz and ~8 kHz are available on this website. The spectrograms are calculated on-board the EFW instrument every 4 seconds. These data include one component of the ac-coupled electric field from the rotating V12 sensors (V12ac) The electric field quantity is labeled EFW/L2. Also included are two components of the wave magnetic field from the University of Iowa EMFISIS sensors. The search coil quantities are labeled EFW/L2 spec64 smu. In the labels “smu” and “smv”, the suffixes “u” and “v” refer to the u and v sensor coordinates rotating with the spacecraft. All these quantities, the nature of the on board signal processing, their filtering, time resolution and their respective coordinate systems are described in the EFW instrument paper (Wygant et al. 2013). When the spectral CDF is opened (in Autoplot, for example) many other possible quantities are listed which have no data in them. These are “place holder quantities” which are used by other instrument modes not currently selected. Examples of the spectral data may be found in our on-line survey data plots. EMFISIS instrument wave data may be found on the University of Iowa website. This includes valuable spectral electric field data up to 400 kHz and magnetic field data up to about ~6 kHz.
EFW Burst Waveform Data
Burst waveform CDF files are available from the links in the “Data” tab. All quantities are in “uvw” coordinates where “u” and “v” are the sensor coordinates rotating with the spacecraft and “w” points along the spin axis. The three data types available are electric field “E”, searchcoil magnetic field “MSC”, and antenna potential “V”. The suffix on each of these is either “B1” or “B2”. B1, or burst 1 is the human-in-the-loop burst, meaning that both collection and playback (for arbitrary lengths of time) are requested on the ground. B2, or burst 2 is automatically telemetered as short bursts based on an onboard triggering algorithm, typically set so trigger on large amplitude signals near 1 kHz. Sample rates for B1 and B2 can and are changed depending on varying science goals.
CDFs for spin fits have quality flags that can be selected and plotted. Some are used and some are not yet used. The following are the flag quantities in the L2 cdfs:
We initially default all the values to -1. This is the “DO_NOT_KNOW” value.
For example, we don’t at the moment look into whether the data are affected by a magnetic wake or not. The values for flags 14-19 are set to -2, which is the “DO_NOT_CARE” value.
For the spinfit MGSE L2 files, we are flagging eclipse times as well as saturated data (voltages > 195 volts). The logic for flagging each few-second chunk of data is as follows:
If (eclipse eq 1) or (V saturation eq 1) or (any other flag thrown) then global_flag = 1
Update and Reprocessing Notes
November 15, 2013: A time tagging error in the conversion from instrument time to UT was discovered in late October, 2013. The timing error varied between 0 and about 0.5 seconds. It was typically 0.1 to 0.3 seconds. It existed for all EFW data (Survey and Burst) processed prior to late October 2013.
The reprocessing to correct this error is complete as of November 15 and all Level 1 and Level 2 data are now corrected on the EFW website. This correction significantly improved the accuracy of VscxB subtraction in electric field spin fits data. Timing of EFW relative to other instrument is also improved. Timing of EFW measurements of whistlers is improved when compared to the world wide ground lightning detection networks. If you are using old Level 1 files (version 1) it is important to delete these and download the new version 2 data.
As part of the development of collaboration with the broader Heliophysics community, the mission has also drafted a “Rules of the Road” to govern how Van Allen Probes instrument data is used.