Data Set Overview
  =================
    Mars Pathfinder bounced down and rolled to a stop on the surface
    of Mars on July 4, 1997.  It landed in an ancient floodplain in
    the Ares Vallis region of Chryse Planitia at 19.17 degrees North
    latitude, and 33.21 degrees West longitude.
 
    The total duration of the mission was 83 sols.  Surface
    meteorology data were collected by the MET instrument on 76 of
    these sols (sols 17, 31, 43, 45, 46, 48, and 51 contain no
    meteorology data).
 
    MET data are organized into science and housekeeping data
    records.  Each record is composed of 12 parameters, which for all
    intents and purposes are measured simultaneously.  The science
    data record consists of surface pressure (two instrument
    sensitivity ranges), atmospheric temperature at three vertical
    locations on the MET mast (referred to as TOP, MIDDLE, and
    BOTTOM), atmospheric temperature as measured by the Descent
    Thermocouple located near the top of the MET mast, and wind speed
    and direction at the top of the MET mast.  See [SEIFFETAL1997]
    for a complete description of the instruments.  The housekeeping
    data record consists of various instrument voltages, currents,
    and reference temperatures.
 
    At the time of writing (Sept.  1998), we have not yet derived
    reliable wind speeds and directions from the data received from
    the Martian surface.  We continue to re-calibrate the wind
    sensor, and hope to have the wind results released soon.
 
    The parameters contained in the science data record are the
    measured atmospheric pressure obtained with the two instrument
    sensitivity ranges (6-10 and 0-12 millibars).  Also included are
    the three temperatures measured along the MET mast, as well as
    the temperature measured by the Descent Thermocouple.  We have
    also included in this data set the measured temperatures of each
    of the six wire segments which comprise the wind sensor.  These
    temperatures, and their spatial variation around the sensor's
    central cylinder, are the data from which wind speed and
    direction will be derived.  The housekeeping data consist of
    several reference voltage measurements from which the stability
    of instrument power supplies and sensor constant current sources
    can be determined.  Additionally, thermocouple reference junction
    temperature, the internal temperature of the wind-sensor
    cylinder, the temperature of the pressure sensor, and the
    temperature of the MET circuit board are measured.  All science
    and housekeeping parameters are derived from sampled digitized
    voltages.
 
    The time intervals during which data were collected are referred
    to as sessions.  Each session is characterized by a session
    number, a duration, and science and housekeeping data record
    sampling rates.  Session numbers during the mission ranged from
    39 (~ 7 AM LST on Sol 1) to 1430 (Sol 83).  Session durations
    varied from 184 seconds to 89052 seconds.  This latter duration,
    used on 5 occasions, produced what we have referred to as
    'Presidential MET Sessions', during which data were continually
    collected at a 0.25 Hz rate for an entire sol.  The science data
    record sampling rate used during the mission varied per the
    specifications of the ASI/MET science team.  It ranged from 0.25
    Hz (a single point measurement of each variable once every four
    seconds) to 2 Hz (2 point measurements per second).
 
    For more info, see [SEIFFETAL1997, SCHOFIELDETAL1997].
 
 
  Processing
  ==========
    There are some time issues involved with this data reduction as
    follows:
 
      1) SCLK values as they came down were 14 characters in length, 10
         digits to the left of the decimal place representing the number
         of complete seconds since 1958-01-01. The three digits to the
         right of the decimal place were the number of 1/256th of a
         second, rather than the number of thousandths of a second.
         Thus, these right-of-the-decimal place values should have
         ranged from 000 to 255.
      2) Ground software (inadvertently) converted the decimal values to
         thousandths of a second, resulting in possible decimal values
         from 000 to 999.
      3) For the local time tags of each data sample, the following was
         done:
         a) using the CHRONOS time tool, the SCLK value for midnight of
            each sol of the mission was determined (SCLK at midnight ~3
            hours before landing, and subsequent local midnight SCLK
            values through the end of sol 83, which was the final sol of
            MET data)
         b) using these midnight SCLK values as anchors, SCLK time was
            linearly interpolated to a 24 hour day; thus, the difference
            between a sample SCLK value and the SCLK value of the
            preceding midnight was divided by the difference between the
            next and preceding SCLK values and this ratio was then
            converted to hours, minutes and seconds assuming a day
            length of 24 'hours'
         c) the derived hours, minutes and seconds as described above
            are the local time tag for each data sample.
 
    CHRONOS was not used on each data point (there were several
    hundred thousand during the course of the mission), but rather
    the interpolation routine above was used to generate a systematic
    time step on any given sol.  Mars orbital characteristics result
    in differing day lengths (in terms of Earth seconds) with
    advancing season.  Since the most useful form for meteorology
    data is knowing what the local time was, the data have been put
    into that form.
 
    For those who desire to know the precise time separation between
    samples in Earth seconds, these values can be obtained from the
    SCLK values and the time tools available at the Navigation and
    Ancillary Information Facility at JPL.
 
    A version 2 of this data set may be generated in which the local
    times are the precise conversion from SCLK to Local Time with the
    SCLK values having the digits to the right of the decimal point
    correctly indicating the number of 256ths of a second.
 
    The data as they are now are perfectly suitable for analysis.
    Direct comparisons with other data sets should be based upon SCLK
    values rather than the precise local times, as other groups/teams
    have used different conversions.
 
    A random sampling of local times in this data set with precise
    conversions between SCLK and Local Time indicates variations much
    less than one second in magnitude.
 
    There were some instances within sessions in which the indicated
    time stamp (SCLK value) for the first science data record of a
    packet would indicate the same time as the final record of the
    preceding packet.  (Data were packaged into packets consisting of
    one header record and 85 data records within each session prior
    to transmission from the lander).  This error does not propagate
    through to the next packet.  It has been corrected in this data
    set by identifying packets with this time problem, and adding an
    increment of time equal to the science sampling interval to all
    the science data record SCLK values in the packet.  Housekeeping
    data record SCLK values are not affected by this problem.
 
    The Ls (solar longitude) value associated with the START_TIME of
    each data collecting session was obtained by using the CHRONOS
    time tool.  The SCLK value at the session START_TIME was
    introduced into CHRONOS, which produced the corresponding Ls
    value.  Thus, the Ls values included in each *.TAB file header
    are precisely determined for the associated SCLK time at which
    the session began.
 
    In order to be able to determine the Ls value for any particular
    data point (SCLK time) during the mission, an analytic expression
    relating SCLK to Ls has been derived.  At each 1000 SCLK count
    interval during the mission, the CHRONOS tool was used to
    determine the corresponding Ls value.  A third-order curve was
    then fit to these Ls values, with the independent variable being
    the difference between the SCLK count at the time of interest and
    the SCLK value at midnight of the sol on which Pathfinder landed
    (approximately three hours prior to landing).  This midnight SCLK
    value is 1246715602.143.
 
    The 3rd-order curve and its coefficients are:
 
        Ls = 142.66 +
             (5.7837E-06  *  (SCLK - 1246715602.143)) +
             (5.6779E-14  *  (SCLK - 1246715602.143)**2) +
             (4.5717E-22  *  (SCLK - 1246715602.143)**3)
 
    With this curve-fit, Ls values for each SCLK count can be
    determined.  Note that Ls values derived with this curve-fit WILL
    NOT exactly correspond to the Ls values contained in the *.TAB
    file headers, since these header Ls values were each precisely
    determined with the CHRONOS tool.
 
 
  Data
  ====
    The data downlinked from the spacecraft are digital numbers
    derived from measured voltages.  Digitization is to 14 bits,
    compared with the 10 bit resolution available to the Viking
    meteorology instruments.  These numbers are stored in the
    Experiment Data Record (EDR) files, whereas the Reduced Data
    Records (RDR), contain results in scientific units (millibars,
    Kelvin, etc.).  EDR values are converted to volts and
    subsequently scientific units using calibration information.
 
    During the nominal mission (sols 1-30), the nominal MET
    observation strategy was 51 equally spaced 3 minute measurement
    sessions per Sol.  For each session science data records were
    sampled at a rate of 0.25 Hz.  Interspersed with these sessions
    were longer (15 minute, one hour, entire sol) sessions often
    sampled at a higher rate (1 Hz).  The purpose of this strategy
    was to characterize the full diurnal cycle throughout the sol, as
    well as higher frequency fluctuations at a variety of times
    during the sol.  These times were spread over a number of sols to
    remain within data downlink constraints.  Thus, we planned to
    characterize the diurnal cycle at 'low-frequency' (51 times each
    sol), and then over the period of ~10 sols to accumulate
    information at higher sampling frequencies so that every hour of
    the day would have such sampling.  Continuous sampling for a
    complete sol at 0.25 Hz (a 'Presidential MET Session') was first
    conducted on Sol 25 and subsequently on sols 32,38,55, and 68.
    These are the only sols during the extended mission (sols 31-83)
    when MET data were collected outside the hours of 0900 - 1500
    LST.
 
    Data are stored on a session-by-session basis.  Thus, data for a
    particular sol are included in multiple files, for sols 1-49.  On
    later sols, single sessions were specified to cover the duration
    of spacecraft activity on that sol (nominally 9 AM to 2:30 PM).
    Thus, from sol 50 (Session 1398) through sol 83 (Session 1430),
    data for each sol are contained in a single data file (session).
 
    Session numbers ranged from 39-1430, but no data for sessions
    1330, 1338, 1340, 1341, 1362, 1370, 1376, 1379, 1384, 1385, 1386,
    1387, 1389, and 1390 were returned.  Thus, there are no files for
    these sessions.
 
    All of the data in this data set are contained in ASCII tabular
    files with detached PDS labels.  Science and housekeeping data
    have been split into separate files, stored in the directories
    SURF_RDR/SCIDATA and SURF_RDR/HKPDATA, respectively.  Beneath
    that, the data files are separated by session number into
    fourteen directories labeled SR00XXS through SR14XXS (for the
    science data) and SR00XXH through SR14XXH (for the housekeeping
    data).
 
    Individual filenames are constructed as follows:
 
                         SR0559S.TAB
 
                         1234567.890
 
    1: The first character will always be an 'S', representing Surface
       data.
    2: The second character will always be an 'R', signifying the
       calibrated data (Reduced Data Record).  The corresponding raw
       data (ie., Experiment Data Record) can be found in the SURF_EDR
       directory, in a file with the same name, except that the 'R'
       will be an 'E'.
  3-6: The next four characters provide the session number of the data
       file.
    7: The seventh character will be either an 'S' or an 'H'.  'S'
       signifies science data; 'H' signifies housekeeping data.
  8-0: The file extension indicates which type of file it is.  The
       only two options are 'LBL' (the PDS label file), or 'TAB' the
       ASCII table containing the data.
 
    As an example, the data from session 559 is stored in the following
    locations on the ASI/MET CD:
 
    SURF_EDR/SCIDATA/SE05XXS/SE0559S.TAB - raw data file
    SURF_EDR/SCIDATA/SE05XXS/SE0559S.LBL - PDS label for SE0559S.TAB
 
    SURF_EDR/HKPDATA/SE05XXH/SE0559H.TAB - raw housekeeping file
    SURF_EDR/HKPDATA/SE05XXH/SE0559H.LBL - PDS label for SE0559H.TAB
 
    SURF_RDR/SCIDATA/SR05XXS/SR0559S.TAB - calibrated data file
    SURF_RDR/SCIDATA/SR05XXS/SR0559S.LBL - PDS label for SR0559S.TAB
 
    SURF_RDR/HKPDATA/SR05XXH/SR0559H.TAB - calibrated housekeeping file
    SURF_RDR/HKPDATA/SR05XXH/SR0559H.LBL - PDS label for SR0559H.TAB
 
    The tabular files are formatted so that they may be read directly
    into many database management systems (DBMS) or spreadsheet
    programs on various computers.  Each of the files contains two
    tables.  The first is the header table, and is only a single
    record in length.  The second table contains all of the data
    records for a session and varies in length.
 
    All fields in the tables are stored in columns of fixed width and
    are right justified.  The records are of fixed length; since the
    header records are shorter than the data records, they have been
    padded with blank spaces at the end of the record.  The last two
    bytes of each record contain the ASCII carriage return and line
    feed characters.  This allows the tables to be treated as fixed
    length record files on computers that support this file type and
    as normal text files on other computers.
 
    The PDS labels are object-oriented.  The object to which the
    labels refer (the tables) is denoted by a statement of the form:
 
        ^object = location
 
    in which the carat character ('^', also called a pointer in this
    context) indicates that the object starts at the given location.
    For an object located outside the label file (as in this case),
    the location denotes the name of the file containing the object,
    along with the starting record.  For example:
 
        ^DATA_TABLE = ('SR0559S.TAB', 2)
 
    indicates that the DATA_TABLE object begins at record 2 of the
    file SR0559S.TAB, in the same directory as the detached label
    file.  (Records are counted starting at 1, not 0.)
 
    The detached label files are stream format files, with a carriage
    return (ASCII 13) and a line feed character (ASCII 10) at the end
    of each record.  This allows the files to be read by the MacOS,
    DOS, UNIX, and VMS operating systems.
 
 
  Software
  ========
    The software employed to generate data in scientific/engineering
    units from the downlinked digital data was developed in IDL and
    FORTRAN, and the software used to create these data tables was
    FORTRAN.  The software took account of pre-flight and in-flight
    calibration of the sensors to convert the downlinked digital
    numbers to voltages, and ultimately to the desired units.  The
    software algorithms are available on the ASI/MET CD.
 
    Questions should be addressed to:
 
             Dr. Jim Murphy
             Dept. of Astronomy, MSC 4500
             New Mexico State University
             Box 30001
             Las Cruces, NM  88003
             Ph: (505) 646-5333
             FAX: (505) 646-1602
             email:  murphy@nmsu.edu
 
 
  Media / Format
  ==============
    The ASI/MET SRFEDR and SRFRDR data will be stored and distributed
    on compact disc-read only memory (CD-ROM) media.  The CDs are
    formatted according to ISO-9660 and PDS standards.

Confidence Level Overview
  =========================
 
    Review
    ------
      The contents of this CD have been peer reviewed by the
      following people:
 
      Lyle Huber        - PDS Atmospheres Node, New Mexico State
                          University
      Julio Magalhaes   - MPF ASI/MET Team, NASA Ames Research Center
      Jim Murphy        - MPF ASI/MET Team & PDS Atmospheres Node, New
                          Mexico State University
      Tim Schofield     - MPF ASI/MET Team Lead, Jet Propulsion
                          Laboratory
      Rob Sullivan      - MPF Participating Scientist, Cornell
                          University
      Betty Sword       - PDS Central Node Data Engineer, Jet Propulsion
                          Laboratory
      Joel Wilf         - PDS Central Node Data Engineer, Jet Propulsion
                          Laboratory
      Elizabeth Duxbury - PDS Imaging Node, Jet Propulsion Laboratory
      John Wilson       - Non-MPF scientist, Geophysical Fluid Dynamics
                          Laboratory/NOAA, Princeton University
 
 
  Data Coverage and Quality
  =========================
    During pre-flight testing, the pressure sensor was tested at
    temperatures (190 K) that were significantly lower than its
    design limits (220 K), and much colder than it experienced within
    the spacecraft on the Martian surface.  This extreme stressing of
    the sensor cast doubt upon the calibration which had been
    obtained prior to that point, and the schedule thereafter did not
    permit recalibration of the sensor in a controlled environment.
    During cruise to Mars, the ASI/MET system was periodically
    powered up and the pressure sensor signal, as well as its
    temperature, were measured.  These measurements, together with
    measurements obtained during free fall and entry, provided a data
    set from which the variation of sensor zero-offset as a function
    of the temperatures experienced during cruise (270-280K) could be
    determined.  It became clear that thermal stressing had changed
    the offset and increased its variation with temperature by a
    factor of 3.  It was not possible to verify gains in flight, but
    laboratory testing on the flight spare sensor suggested that
    sensor stressing would not produce significant gain changes.  The
    flight data, in addition to the pre-flight calibration data, have
    been used to determine pressure from the down-linked pressure
    sensor signals.
 
    The thermocouples appeared to work very well, and there is no
    reason to question the results from them.
 
    As stated above, the determination of wind speed and direction is
    ongoing, and a new wind data set will be issued following
    analyses of recent sensor recalibration work, and application of
    its results to wind-sensor data returned from the Martian
    surface.