Data Set Information
DATA_SET_NAME VG1 J PLASMA WAVE SPECTROMETER DENSITY 1S V1.0
DATA_SET_ID VG1-J-PWS-5-DDR-ELE-DENSITY-1S-V1.0
NSSDC_DATA_SET_ID
DATA_SET_TERSE_DESCRIPTION
DATA_SET_DESCRIPTION Data Set Overview ================= This data set consists of ASCII formatted plasma wave frequency and electron plasma density measurements as measured by the Plasma Waves Science instrument and calculated from the equations of cold plasma theory. These frequency measurements were taken from Voyager 1 electric field waveform samples from Voyager 1 Plasma Wave Science (PWS) waveform receiver obtained during its Jupiter flyby. The data set includes select measurements from spacecraft event time (SCET) 1979-03-01T19:58:22.500Z to 1979-03-21T14:25:32.500Z. The data are separated into day files and the individual data points are taken every 1-second. As will be explained in more detail below, the frequency measurements and therefore the density measurements in this data set are measured from high-resolution wideband plasma wave spectra. To learn more about these spectra and their specific submitted volumes, see the Related PDS Products section in the AAREADME files found in the root directory of this volume. Parameters ========== While the data essential to this volume are the electron plasma densities, there are a number of other plasma parameters included with this data. The data set consists of ASCII files with one record per time step, occurring in 1-second increments. Each record includes the time, magnetic field strength (obtained from the Voyager 1 magnetometer), the electron cyclotron frequency (if available), the frequency of the cutoff or resonance measured, a code indicating the name of the frequency measured, the calculated electron density, and a set of position coordinates for the spacecraft at the time of the observation. Also included in each record are the electron plasma frequency fpe, extraordinary mode cutoff frequency fR=0, ordinary mode cutoff frequency fL=0, upper hybrid resonance frequency fUH, and a quality index. One of these four frequencies is just a copy of the measured cutoff or resonance frequency while the remaining frequencies are calculated using the magnetic field data and the equations of cold plasma theory. Different files are used for each day. Processing ========== The ASCII density data files produced in this volume were derived from measuring the characteristic frequencies from the local plasma. The density was calculated from these data, along with cyclotron frequency data derived from magnetic field data, using the equations of cold plasma theory. In order to measure these characteristic frequencies, this work utilizes a new program that allows the operator to highlight the general vicinity of the cutoff or resonance on a frequency-time spectrogram. Then, an algorithm finds the cutoff or resonance in the region and records the frequency at 1 second intervals. Hence, the automated procedure has a high temporal resolution (1 second) and requires a relatively low level of both manual effort and subjective judgment by the operator. There are two different algorithms used: one for cutoff detection and one for resonance or peak detection. The cutoff detection algorithm is controlled by a small number of parameters that can be set by the operator. The first parameter is the cutoff level. In determining possible cutoff candidates, the algorithm scans the region highlighted by the operator and records two separate points, one above the cutoff level and one below. The closer the two points are, temporally, the steeper the slope will be. Therefore, the operator can change the location of the cutoff level to manipulate where the algorithm looks for cutoffs within the highlighted region of interest. The next parameter is the slope magnitude, which designates the minimum magnitude of the finite difference slope where the cutoff must reside. The operator may raise the slope level in order to scan only for sharp cutoffs, or lower it in order to accommodate less steep slopes, depending on the quality of the spectrum data. When there is more than one possible cutoff, the detection program will display them as cutoff candidates. The cutoff level, slope magnitude and cutoff candidates are displayed by the program for viewing by the operator. While the algorithm chooses the lowest frequency cutoff by default, the operator may override the algorithm and choose any of the possible cutoffs to be recorded. While most of the characteristic frequencies are, by definition, the cutoff of propagating wave modes, there are certain circumstances when the characteristic frequency is denoted as the peak of a wave mode in the spectrum. Because of this, there is an algorithm specifically for resonance or peak detection. Many spectra of interest to this study include Z-mode radiation, which has a low-frequency cutoff at fL=0. Barbosa et al. [BARBOSAETAL1990B] demonstrated that taking the peak of the Z-mode as fL=0 yields the highest consistency in the determination of fpe. Hence, when the Z-mode is enhanced, we utilize the peak detection algorithm to identify fL=0 from which fpe and the electron density can be derived. This algorithm can also be used to determine fUH when an enhancement at that frequency is present in the spectrum. In order to measure this resonance or spectral peak, the peak detection algorithm fits a Gaussian curve to the highest peak within the region specified by the program operator. The algorithm then records the frequency of the Gaussian's peak as the peak frequency in the spectrum. The algorithm displays the spectrum and a darker line which is the Gaussian. Because there may be noise which exhibits a large peak in the highlighted spectrum, the spectrum is displayed along with the Gaussian curve and a vertical line designating where the peak was measured. The operator always has the ability to manually change the peak's location and alter the measurement in such cases. While the operator utilizes a color spectrogram to guide the cutoff and peak detectors, we emphasize that this is only used as a means of identifying the appropriate range in frequency for the algorithm to search. The direct use of color spectrograms tends to mislead an operator to perceive a cutoff that is not equivalent to the cutoff in the actual power spectrum [BARBOSAETAL1990B]. Because this may lead to a systematic error in the data, the algorithm utilizes the spectrum itself, and does not depend on a color scale to determine the characteristic frequencies. This should reduce systematic error and lead to more accurate results. Data Coverage ============= This data set does not provide complete coverage of the time intervals when Voyager 1 was within Jupiter's magnetosphere. Two criteria were necessary in order for density measurements to be obtained. First, wideband waveform data must exist. These data are waveforms sampled at 4-bit resolution and have a frequency range from 40 Hz to 12 kHz. 1600 samples are collected for 55.56-msec at a rate of 28,800 samples per second, and are then followed by a 4.44-msec gap. These 60-msec intervals make up one line of waveform samples. The spectrograms then consist of 48-second 'frames', consisting of up to 800 lines or 48 seconds each. Because these data compete with Voyager imaging data for the downlink resource, data is available only a fraction of the time. Plasma wave spectra data can be either continuous, periodic (such as one 48-second frame out of every four such frames), or simply present sporadically. Second, spectral features which exhibit a cutoff or resonance related to the electron density must be observable in the spectrum. By far the most prevalent emission of use is the non-thermal continuum radiation, whose low frequency cutoff is at fpe. This radiation literally fills the magnetosphere between the magnetopause and higher density regions of the inner magnetosphere. However, some regions include other emissions and regions when the plasma frequency does not have a clear cutoff. If the plasma frequency is not measured directly from the spectrum for any reason, it may be calculated from the local magnetic field data (essentially fce) and one of three other characteristic frequencies using the equations of cold plasma theory. Thus, magnetic field data must exist for regions when fpe is not present, or the electron density cannot be calculated. Typically, continuum radiation is not present inside of approximately 20 to 25 Jovian radii. There is at least twice as much data derived from Voyager 1 than Voyager 2 and data exists typically from approximately 20 to 65 Jovian radii. Interpretations =============== High resolution wideband waveform data were used to measure characteristic frequencies (peaks and cutoffs) which relate to the electron plasma density. When dealing with a variety of spectrograms and plasma conditions found in different regions of the Jovian magnetosphere, it is necessary to interpret the present modes and characteristic frequencies correctly in order to determine the most accurate value for the electron plasma density. Below, we will briefly discuss the methods used for interpreting different spectra. A more thorough explanation of the interpretations including examples of different spectra can be found in the DENSITY.DOC and DENSITY.PDF files within the DOCUMENTS directory accessible from the root directory of this volume. The simplest spectra to interpret for the purposes of determining the electron plasma density are those that include non-thermal continuum radiation with a clear low frequency cutoff and with no other emissions obscuring the cutoff. For the purposes of this data set, we agree with the Gurnett et al. [1981] interpretation that based upon PWS data it is appropriate to assume that the continuum radiation cutoff is fpe and we can accurately determine the local electron plasma density using the appropriate equation from cold plasma theory. The electron plasma density is directly proportional to the square of the electron plasma frequency and therefore in this case the determination of the density does not depend on magnetic field measurements. When there is only one cutoff present in the continuum radiation, we assume that the continuum radiation is propagating in the ordinary mode and that the cutoff is indeed the plasma frequency. An alternate possibility would be to identify this cutoff as the R=0 cutoff at fR=0. But, most theories [Shaw and Gurnett, 1980; Moses et. al., 1987; Barbosa et al., 1990] favor the L,O mode as the most likely continuum radiation component, hence, we assume that there is always at least some L,O component present when the continuum radiation is detected. Sometimes, more than one wave mode cutoff is present at different frequencies for the same time period. When this is the case, we use a guess-and-check system called Consistency Checks. The method is as follows: 1. Assume that one frequency cutoff/peak present in the spectrum is a particular characteristic frequency. 2. Use the local magnetic field data (which determines fce) along with the equations from cold plasma theory to calculate the remaining characteristic frequencies. 3. Look for consistency between the calculated frequencies and the remaining spectral features (cutoffs/peaks) present. A consistent interpretation is one where the calculated frequencies match the cutoffs/peaks present in the spectrum. In some examples there exist two spectral cutoffs which need identification in order to calculate the electron plasma density. By using a Consistency Check, (mainly, assuming fpe is the lower cutoff and calculating the remaining characteristic frequencies) it is found that when the plasma frequency is assumed to be the lower frequency cutoff, the cutoff at higher frequencies matches the calculated R=0 frequency. In addition to the non-thermal continuum radiation with a low frequency cutoff at fpe, another mode of propagating waves (called Z-mode) is sometimes also present. Based upon results from our consistency checks and in agreement with the previous work of Barbosa et. al. [1990], we conclude that there are two types of Z-mode radiation: weak, broadband Z-mode and intense, narrowband Z-mode. We interpret the L=0 frequency as the cutoff of the weak, broadband radiation, however, when the Z-mode emission becomes intense we have found that taking the L=0 frequency as the peak of the intense emission gives the most consistent estimate for fpe. This is because as the intensity of this peak increases, the width of the emission appears to broaden due to limitations in the Fourier transform and the low dynamic range of the 4-bit waveform system on Voyager 1. Barbosa et al. [1990] demonstrated that taking the peak of the Z-mode as fL=0 yields the highest consistency in the determination of fpe, which concurs with our consistency checks. Thus, in regions where Z-mode is present, we can determine the electron density by either measuring the cutoff of the broadband Z-mode as fL=0 or the peak of the narrowband intense Z-mode as fL=0, and using the equations of cold plasma theory. As Voyager approaches the lobe of the magnetosphere, the density as well as fpe drops precipitously and approaches or even drops below the cyclotron frequency. Perraut et al. [PERRAUTETAL1998] studied one such case obtained by the Galileo plasma wave instrument. In order to determine the proper identification of the characteristic frequencies and determine the electron plasma density, we utilize our previous consistency check method. When an interpretation is found to be consistent with the spectrum, we assume temporal continuity of the cutoffs, and extend the interpretation into regions where a consistency check is not possible due to lack of features in the spectrum or a lack of magnetic field data. While there is no method for determining the density certainly in these regions, we believe that assuming that the spectrum does not change greatly in a span on the order of minutes is appropriate and suitable to determine the density. A more thorough look into time periods where fpe << fce, including multiple examples, is included in the DENSITY.DOC and DENSITY.PDF files in the DOCUMENTS folder found in the root directory of this volume. Ancillary Data ============== None Coordinate System ================= Included in this data set are two coordinate systems that ensure an accurate location of the density data points. The first system of coordinates consists of distance, longitude and magnetic latitude and is referred to commonly as the Jovigraphic coordinate system, or one that is fixed to the rotation of the planet. We have used the System III Jovigraphic coordinate system which uses the planet's magnetic field to measure the rotation. The radial distance is defined as the distance from the center of Jupiter to the spacecraft (in kilometers) divided by the radius of Jupiter at the equator (71492km). In the usual astronomical convention, the longitude is a west longitude which increases with time from an observer above the system, rather than just the angle of rotation about the z-axis. The other coordinate system included is referred to as the Jovicentric Solar Ecliptic (JSE) system. This coordinate system has its x-axis point from Jupiter toward the Sun, and its y-axis is chosen to be in the ecliptic plane pointing toward dusk (thus opposing planetary motion). Its z-axis is parallel to the ecliptic pole. All X, Y, and Z coordinates are measured in Jovian radii (1 Rj = 71492km). Software FIX FIX FIX FIX FIX ======== Included in this volume is Java BROWSE software that allows a user to view the high resolution spectrogram data. Also, the Browse tool allows the user to print the cutoff/resonance data used to determine the electron density. This allows all users to see firsthand the interpretations invoked in this study, and determine for themselves the accuracy of the density determination for a particular time period. The resonance/peak and cutoff detection algorithms are also included in this section to insure replicability by all users. Media/Format ============
DATA_SET_RELEASE_DATE 2007-11-05T00:00:00.000Z
START_TIME 1979-03-01T07:58:22.500Z
STOP_TIME 1979-03-21T02:25:32.500Z
MISSION_NAME VOYAGER
MISSION_START_DATE 1972-07-01T12:00:00.000Z
MISSION_STOP_DATE N/A (ongoing)
TARGET_NAME JUPITER
TARGET_TYPE PLANET
INSTRUMENT_HOST_ID VG1
INSTRUMENT_NAME RADIO AND PLASMA WAVE SCIENCE
PLASMA WAVE RECEIVER
INSTRUMENT_ID PWS
INSTRUMENT_TYPE PLASMA WAVE SPECTROMETER
NODE_NAME planetary plasma interactions
ARCHIVE_STATUS ARCHIVED
CONFIDENCE_LEVEL_NOTE Confidence Level Overview ========================= In order to determine the density, we used an analysis tool that measured characteristic frequencies from PWS wideband spectrograms. This frequency detection tool contained both a cutoff and peak algorithm which measured either the cutoff of the spectrum or the peak, respectively. Two different semi-interchangeable systems of confidence analysis were used depending on which algorithm was used. In order to record the confidence of the cutoff frequency measurements, data quality indices were given to each data point. The indices range from 0 to 3 with 0 being a cutoff with the highest confidence and 3 being the least. The amount of noise in the spectrum dictates the quality rating given to a particular data point. Points with a minimal amount of noise are given values of 0, whereas points where the spectral noise is so great that the cutoff is unclear are given a 1 or 2 index depending on the severity of the noise. A data quality index of 3 is given when there are obstructions in the spectrum blocking the ability for an accurate measurement, or when there is such a significant amount of background spectral noise that the spectra must be time-averaged over a longer period of time in order to interpret the cutoff. For example, the Voyager wideband receivers have notch filters that are used to suppress interference at the power supply frequency of 2.4 kHz and its third harmonic. When a cutoff resides near a notch filter, a data quality index of 3 is used. While the notch filters do not span more than 200 Hz, any measurement made near them is considered uncertain. In order to determine an estimate of the error of a peak measurement, the full-width at half maximum of the Gaussian curve is measured and recorded along with its peak frequency. The narrower the peak is, the more accurate the measurement will be. This quantitative method is superior to the quality data indices described above. However, to provide a consistent data set we have also translated the width of the peak to quality indices by assigning an index of 0 for 0 < deltaf <= 20 Hz, 1 for 20 < deltaf <= 35 Hz, 2 for 35 < deltaf <= 50 Hz, and 3 for deltaf > 50 Hz. Limitations =========== Since the Fourier transform parameters used to produce the spectra yield a spectral resolution of 28 Hz, this represents the ultimate accuracy with which a cutoff can be determined. Previously, we discussed the Consistency Check method when two or more characteristic frequencies were identified consistently with the spectrum and used to determine fpe, whether the plasma frequency is visible or not. For all the density data compiled using this process with Voyager 1 and 2, the calculated frequencies typically agree with the features observed in the spectrum to several 10's of Hz. Almost all cases agree to within 100 Hz. The uncertainty in fpe translates into a variable uncertainty in the plasma density and is greater in a relative sense when the plasma frequency is low. For example, if we assume an error of 100 Hz in any measurement, a measurement at 300 Hz produces a 70% fractional uncertainty in the density. However, if the measurement is made at 3 kHz, then the fractional uncertainty of the density is much smaller, approximately 7%. It is important to note that if the plasma frequency is not present in the spectrum, the electron plasma density must be calculated using another characteristic frequency and the cyclotron frequency, which is directly proportional to the local magnetic field. If magnetic field data do not exist for a region where the plasma frequency cannot be measured, the density can not be determined. There are a small number of regions containing anomalous magnetic field data which may affect the density calculation. These magnetic field dropouts are sharp, momentary decreases or increases in the magnetic field which are instrumental effects and do not represent realistic occurrences. They are characterized by a magnetic fluctuation of several orders of magnitude in a span of less than one minute. Regions of anomalous magnetic field data which affected density measurements were changed manually from their anomalous values to the floater value -1 and the subsequent freqeuency and density calculations were changed appropriately. During any given time step in Voyager spectrogram data, there may be more than one measurable frequency which can be used to calculate the density. To obtain the final data set included in this volume, a number of steps were followed to take many measured frequencies and conclude with only one density measurement for each time step. Firstly, each frequency cutoff and/or peak that could be measured from Voyager spectrogram data was measured separately and saved in individual folders designating the frequency which had been measured (e.j. p_files, l_files, r_files, u_files). The digitized frequency data files contain only a time stamp, a frequency measurement, and either a data quality index or a full-width at half-maximum value. Next, this data were run through a program which creates new files (1 for each measured frequency) with the same structure as in the final data of this volume. This includes compiling magnetic field data and coordinate data for each time step, and calculating the characteristic frequencies and density based upon the appropriate measured frequency. This output a maximum of four separate versions (based upon how many frequencies were measured) of the final data for each particular time step. It is necessary to use only one interpretation (one measured frequency) of the spectra for a particular time step in order to obtain only one calculation of the density for that time step. Therefore, an 'index' file is created which specifies the correct interpretations to be used for every Voyager 1 and 2 time period. Then, a program was run which read the index file, selected the correct time record (from either p, L, R, or u files), and copied the record into the final data day-files. The final day-files were then checked for measurement errors or anomalous magnetic field data which caused incorrect density calculations. Data points found were either deleted or the fce values, based directly upon magnetic field measurements, were switched from their anomalous value to -1 floater values. For a complete list of manually altered data values, see POSTPROD_V1.TXT and POSTPROD_V2.TXT which are located in the EXTRAS directory accessible from the root directory of this volume. The final step converted the files into .CSV files which are the final data found in this volume. The present investigation focuses only on the outer magnetosphere. However, the work of Scarf et al. in 1979 [SCARFETAL1979] has shown that in the inner magnetosphere, other wave phenomena such as the upper hybrid resonance band can be used to infer the plasma density. For future investigations, our resonance/peak algorithm may be used to accurately determine the plasma density in regions inside ~20 Jovian radii.
CITATION_DESCRIPTION Barnhart, B.L., Kurth, W.S., Groene, J.B., Piker, C.W., Faden, J.B., Santolik, O., and D.A. Gurnett, VG1-J-PWS-5-DDR-ELE-DENSITY-1S-V1.0, VG1 J PLASMA WAVE SPECTROMETER DENSITY 1S V1.0, NASA Planetary Data System, 2007.
ABSTRACT_TEXT Electron plasma density data set for Jupiter's outer magnetosphere derived from high-resolution wideband measurements of low-frequency radio and plasma waves obtained by Voyagers 1 and 2 during their 1979 flybys. Density data were measured at 1-second intervals.
PRODUCER_FULL_NAME DR WILLIAM S. KURTH
SEARCH/ACCESS DATA
  • Planetary Plasma Interactions Website
    		•