Data Set Overview
    =================
    This data set tabulates local electron densities and local magnetic
    field strengths obtained from Mars Advanced Radar for Subsurface and
    Ionosphere Sounding (MARSIS) Active Ionospheric Sounding (AIS) mode
    ionograms.  The electron density is obtained by measuring the increment
    in frequency between the plasma frequency harmonics, visible as bright
    vertical lines at low frequency and delay time in many MARSIS
    ionograms. Similarly, the local magnetic field strength is found by
    measuring the difference in delay time between electron cyclotron
    echoes, visible as bright horizontal lines at low frequency on many
    ionograms.  Both the measured quantity and the derived result are
    included in the archive product.  We also include a data quality flag
    giving the impression of the archivist as to how reliable each result
    is.  This archive product includes only data from the fifth extended
    mission, which starts about orbit 13960 and ends on orbit 16459.

    Parameters
    ==========
    This data set consists of electron plasma density and magnetic field
    magnitudes along with spacecraft position information in a Mars - Sun
    coordinate system as well as data quality flags.

    Processing
    ==========
    The local electron densities are obtained from individual ionograms by
    measuring the frequency spacing between plasma frequency harmonics,
    primarily visible at sounding frequencies below 1 MHz.  These harmonics
    appear as bright vertical lines at approximately constant frequency
    intervals.  The local plasma frequency is equal to the frequency
    interval between the vertical lines.  An adjustable scale with digital
    readout coordinated with the computer screen cursor is used on the
    graphical representation of the ionogram to manually measure the
    interval between harmonic lines.  A quality flag is assigned by the
    operator to each measurement.  The criteria for assigning the various
    values of this flag are given elsewhere in this document.  The measured
    plasma frequency and its corresponding quality flag are written to the
    output file.

    During further processing the plasma frequency is converted to electron
    density using the well-known formula:

      n_e [cm^-3] = (f_pe [MHz] /8.980e-3) ^2

    where n_e is the electron density in particles per cubic centimeter and
    f_pe is the electron plasma frequency in megahertz.

    For more complete explanation of this process, and for discussion of
    how the plasma frequency harmonics are generated, please refer to
    Gurnett et al. (2005, 2008), Morgan et al. (2008, 2013a, 2013b), and
    Duru et al. (2008).  A fully automated method of obtaining the local
    plasma frequency is given by Andrews et al. (2013).

    The local magnetic field strengths are obtained from individual
    ionograms by measuring the time-delay spacing between electron
    cyclotron echoes, visible as equally spaced bright horizontal at low
    frequency on an ionogram.  The constant time-delay interval between the
    echoes is equal to the electron cyclotron period, i. e., the inverse of
    the electron cyclotron frequency.  An adjustable scale with digital
    readout coordinated with the computer screen cursor is used on the
    graphical representation of the ionogram to manually measure the time
    interval between electron cyclotron echoes.  A quality flag is assigned
    by the operator to each measurement.  The criteria for assigning the
    various values of this flag are given elsewhere in this document.  The
    measured electron cyclotron period and its corresponding quality flag
    are written to the output file.

    The electron cyclotron period is converted to magnetic field strength
    by using the well-known relation between magnetic field strength and
    electron cyclotron frequency, written here in terms of the electron
    cyclotron period:

      |B| [nT] = 1000/(28 * T [ms])

    where T is the measured electron cyclotron period in milliseconds and
    |B| is magnitude of the magnetic field in nanotesla.  For a more
    complete explanation of this process along with discussion of how the
    electron cyclotron echoes are generated, please refer to  Gurnett et
    al. (2005, 2008), Akalin et al. (2010), and Morgan et al. (2011).

    Data
    ====
    The electron plasma density and magnetic field magnitude data set is
    provided as simple ASCII SPREADSHEET objects with fields delimited via
    commas.  Each row in a spreadsheet contains a fixed number of fields,
    however some  fields may have a null entry, i.e. two adjacent commas
    with no value.  This occurs when one of the measurments (either the
    plasma density or B-field magnitude) was unclear.  In these case only
    the measurement which could be made with a mimimum of confidence is
    included and the other is left empty.  One row in each spreadsheet
    corresponds to a single frame in the upstream
    MEX-M-MARSIS-3-RDR-AIS-V1.0 dataset.


    Ancillary Data
    ==============
    Ancillary spacecraft position data are provided adjacent to the
    measurements to which they pertain.  In addition to the primary
    measurements, each row a product contains up to 2 data quality flags
    and a handful of spacecraft ephemerides considered relevent by the
    dataset producers.


    Coordinate System
    =================
    The primary data are scalar measurements that reflect conditions local
    to the spacecraft.  Because these quantities are scalars, no coordinate
    system is required for their intrinsic description; however the
    corresponding spacecraft position within the Mars-Sun system may be
    important in analyzing these products.  As a convenience to the archive
    user, the following position data fields are provided inline within the
    data product files:

      ALTITUDE ABOVE THE GEODE
      PLANETOGRAPHIC LATITUDE
      PLANETOCENTRIC LATITUDE
      LOCAL TRUE SOLAR TIME
      WEST LONGITUDE
      MARS-SUN DIRECTION DISTANCE (X)
      MARS-SUN PERPENDICULAR DIRECTION DISTANCE (RHO)
      SOLAR ZENITH ANGLE

    A description of each field may be found by reading any data label on
    this volume.  Further position information may be found in the
    GEO_INDEX.TAB file in the INDEX directory.  Also, the spacecraft event
    time field  provided as input to the NAIF Spice toolkit may be used to
    generate coordinate values in almost any imaginable system.


    Software
    ========
    As the data are formatted as simple ASCII spreadsheets, no
    programs are provided on this volume.


    Confidence Level Overview
    =========================
    For electron cyclotron echoes the data quality flags are given as follows:

    0 - NO echoes
    1 - 1 or 2 echoes, look faint
    2 - 1 clearer echo, with more echoes
    3 - Best (this usually is multiple, 5 or greater, that gives the best
        confidence aligning the trace

    For plasma frequency harmonics the data quality flags are given as
    follows:

    0 - NO plasma frequency harmonics
    1 - Harmonics are visible but there is significant doubt as to the correct
        value of the interval between harmonics.  This doubt occurs because
        the harmonics are blurred and difficult to locate precisely or because
        the frequency interval between harmonics is not unique.
    2 - The harmonics are well determined with a unique frequency interval
        between them.

    Review
    ======
    The MARSIS Electron Plasma Density and Magnetic Field Magnitude data will
    be reviewed internally by the Mars Express MARSIS team prior to release to
    the PSA. The data set will also be peer reviewed by the PSA and PDS.


    Data Coverage and Quality
    =========================
    Data coverage is effected by MARSIS being a dual-mode instrument.  In
    general, MARSIS coverage of the nightside of Mars is dedicated to
    Subsurface mode, since that is when the ionosphere interferes least with
    subsurface measurements.  This means that when Mars Express periapsis is
    primarily on the nightside, MARSIS AIS measurements are limited to
    altitudes above about 850 km, about 5 minutes on either side of periapsis.
    Early in the mission, there are also some partial orbits with a ten
    minute gap poleward of periapsis.  These partial orbital passes are not
    guaranteed to be included in this archive set.  A full orbital pass in
    the MARSIS nominal mission consists of approximately 35 min 50 seconds of
    continuously acquired ionograms.

    For both plasma frequency harmonic's electron density measurements and
    electron cyclotron echo-magnetic field strength measurements, all full
    orbital passes between orbits 13960 and 16459,
    2014-12-31T17:39:38 through 2016-12-29T08:31:18, are here archived.  These orbits
    and dates approximately bracket the beginning of regular data collection
    for the fifth extended mission of MARSIS and the end of the
    fifth extended mission.

    For the nominal mission volume the archived electron cyclotron echo
    magnetic field strength data include a number of orbits prior to orbit
    1994 which were taken during the commissioning phase of MARSIS.  Both
    archived electron cyclotron echo-magnetic field strength and plasma
    frequency harmonic-electron density data include a number of partial
    orbits that were collected as targets of opportunity.  We include such
    data with no guarantee of completeness of coverage.

    The uncertainty in frequency measurements for MARSIS AIS is discussed by
    Morgan et al. (2013). MARSIS AIS samples 160 frequencies over 1.26
    seconds in collecting data for an ionogram.  These frequencies are spaced
    at somewhat irregular intervals in order to avoid interference features
    and discontinuities in the receiver sensitivity; however, the general
    trend in spacing between sampled frequencies is quasilogarithmic.  The
    implication of this sampling regime is that over most of the sampled
    frequencies, between 0.5 MHz and 5.5 MHz, the fractional uncertainty in
    frequency is approximately 2%.  At frequencies below 0.5 MHz, down to the
    minimum sampling frequency of 0.1 MHz, the uncertainty increases sharply
    to around 12%.  The vertical lines that yield the local plasma frequency
    virtually always occur below frequencies of 1 MHz and above 0.3 MHz.
    Figure 3 of Morgan et al. (2013) indicates that measurement error will be
    below 4% in this range and will center about 2.5%.  Because the local
    electron density is proportional to the square of the plasma frequency,
    the fractional error on the electron density will be approximately twice
    that on the plasma frequency or about 5%.

    The time width of an electron cyclotron echo on an ionogram is usually
    about twice the time width of the sounding pulse, which is 91.4
    microseconds.  Therefore, we estimate the error in any single measurement
    of the electron cyclotron period as +/- 0.1828 milliseconds.  Because the
    fractional error in magnetic field strength is equal to the fractional
    error in time, and the absolute error in time is constant, the fractional
    error in magnetic field increases as the electron cyclotron period
    decreases and the magnetic field increases.  We can take into account the
    effect of having more electron cyclotron echoes to measure when the
    magnetic field is high by dividing by the square root of the number of
    echoes.  The resulting estimated percent errors are as follows:

         T_echo ms     |B| nT       Error on |B| %
         ---------     ------       --------------
           0.25         40.7            13%
           1.17         30.6             6.0%
           2.08         17.2             4.5%
           3.0          11.9             3.8%
           3.9           9.1             3.2%
           4.8           7.4             3.0%
           5.7           6.2             2.8%
           6.6           5.3             2.5%
           7.6           4.7             2.3%


    Limitations
    ===========
    Because the local electron density and magnetic field strength are
    collected at most one time from an ionogram, the frequency of collection
    is 1/7.543 Hz or 0.133 Hz.  Practical limits on the plasma frequency
    harmonics are 0.03 to .75 MHz or local electron density of 10 to 7000
    cm^-3.  Practical limits on the electron cyclotron period are 7.5 ms to
    200 microseconds or 5 nT to 180 nT.  The measurements described in this
    document are not observed when the Mars Express spacecraft is outside of
    the magnetosheath of Mars because (1) the values are typically too low to
    be measured and (2) the solar wind flow around the spacecraft convects the
    plasma away from the spacecraft before it can interact with the antenna
    and receiver of the MARSIS sounder.  This effect is explained in detail in
    the reference DURUETAL2008.

Standard Quality Flags:

      Confidence in the measurements provided by this
      dataset is indicated in a set of two quality flags.  For plasma
      frequency harmonics the data quality flag values are:

        1 - Harmonics are visible but there is significant doubt as to the
            correct value of the interval between harmonics.  This doubt
            occurs because the harmonics are blurred and difficult to locate
            precisely or because the frequency interval between harmonics is
            not unique.

        2 - Best quality.  Harmonics are clear with well-defined interval
            between them.

      For electron cyclotron echoes the data quality flag value are:

        1 - 1 or 2 echoes, look faint

        2 - 1 clearer echo, with more echoes

        3 - Best quality, with 5 or more clear echoes, giving the best
            confidence aligning the trace.

      Verification flags:

      For a limited set of data, the electron densities have been verified by
      a second independent measurement.  For the points where a verification
      has been attempted, the verification flag values are:

        empty field - No verification attempted

        1 - Verification attempted; difference between archived value and
            independently measured value is greater than or equal to one
            frequency pixel, which is 0.01 MHz in the frequency range where
            this method of measurement is applicable.

        2 - Verification attempted; difference between archived value and
            independently measured value is less than one frequency pixel,
            0.01 MHz.