Data Set Overview
  =================
    The REMS instrument is a meteorological suite of sensors designed
    to provide measurements of air and ground temperatures, wind speed
    and direction, pressure, humidity and ultraviolet radiation.

    The REMS ENVRDR data set contains processed REMS data converted to
    environmental magnitudes.

    Only some of the external factors affecting the degradation of the
    sensors (such as dust in the Ground Temperature and Ultraviolet
    Sensors, among others) have been modeled. Ultraviolet Sensor data
    are corrected by the Solar Zenith Angle since otherwise they would
    not be useful. Other factors (such as rover heat sources, shadows
    and ASIC electronics noise, among others) have been flagged with a
    confidence level code to indicate the quality of the measurement.

    Data is a time ordered sequence of rows organized into a table,
    taken at a maximum resolution of one second. Each data product
    contains one sol worth of activity and has information from all
    sensors.

    This data set is the result of the second processing step in the
    calculation of the real environmental values measured by
    REMS. Data is converted from electrical magnitudes and
    temperatures to environmental magnitudes with only few
    corrections.


  Parameters
  ==========
    The information included in this data set for each sensor is the
    following:

    - Wind Sensor: longitudinal and transversal differential thermal
      conductance for each of the 3 boards in each boom.
    - Ground Temperature Sensor: brightness temperature of the 3 
      thermopiles, and their estimated systematic uncertainties.
    - Air Temperature Sensor: temperatures measured by each one of the 
      PT1000 sensors on each boom (unmodified from the TELRDRs), and
      their estimated uncertainties.
    - Ultraviolet Sensor: ultraviolet radiation for each band and
      their estimated uncertainties.
    - Humidity Sensor: relative humidity from each of the 3 channels,
      their estimated calibration uncertainties and the Humidity
      Sensor operating temperature.
    - Pressure Sensor: pressure from each of the 2 barocaps, 2
      thermocaps temperatures and their estimated accuracy. Pressure
      Sensor configuration (oscillator and low/high resolution mode).
 
    Sampling is at 1Hz maximum, with a baseline operation of 5 minutes
    every hour. Additional measurements can be taken on an on-demand
    basis beyond those hourly observations. For these additional
    measurements, and besides tactical day to day conditions and
    resources, there is a general pattern that covers selected hours
    of the day built by the scientific team during operations. That
    pattern is shifted from sol to sol to cover the whole 24 hours
    after a few sols.

    Additionally, extended measurements can also be triggered
    automatically if event mode is activated, in which case the REMS
    computer will decide or not to continue measuring after the
    regular cadence, by comparing the previous measurements with the
    expected trend. The objective is to capture any ongoing transitory
    atmospheric event.

    During the first 72 sols, for each 5 minute block, the following
    measurement strategy was used: Wind Sensor is switched off for 60
    seconds, then it is switched on for 235 seconds, and then it is
    switched off again for the final 5 seconds. The rest of the
    sensors are switched on all the time. This strategy was based on
    results obtained during pre-flight testing. However, after
    evaluating flight data, it was determined that this strategy was
    not necessary, so from sol 73 onwards all sensors are switched on
    for each 5 minute block.

    From sol 793 onwards, a new measurement strategy for Humidity
    Sensor was introduced. It is called HS HRIM (Humidity Sensor High
    Resolution Interval Mode) and is only used on selected one-hour
    long observations. This new strategy intends to minimize heating
    of the Humidity Sensor, and consists of alternately switching on
    and off the sensor at periodic intervals. At the same time, Boom 2
    is switched off, which means that there are no Wind Sensor and Air
    Temperature Sensor measurements.


  Processing
  ==========
    Processing starts with the generation of the REMS EDRs. The REMS
    EDR data products are generated by the MIPL (Multimission Image
    Processing Laboratory) at JPL, under the OPGS, using the telemetry
    processing software called MSLEdrGen. This software will convert
    the binary data received from telemetry to ASCII. EDRs will then
    be retrieved at Centro de Astrobiologia (INTA-CSIC) using the File
    Exchange Interface (FEI).

    EDR data products have a first automatic process using calibration
    data. The result of this is the TELRDR data set, which contains
    electrical magnitudes and thermal data. In parallel, using
    ancillary data provided by JPL (such as rover location, sun
    position) the ADRs (Ancillary Data Records) are generated. EDRs,
    TELRDRs, ADRs, and calibration data are processed together to
    obtain the ENVRDRs.
    
    The level of processing of the ENVRDRs includes environmental
    magnitudes with minimal corrections (mainly based on the
    degradation of the sensors). Finally, applying models developed by
    the REMS team and refining them with the confidence levels, the
    MODRDR data set is created.

    The specific data used in the generation of ENVRDRs are:
    - Laboratory calibration data, which can be found in the CALIB 
      directory.
    - In-flight calibration data (of the Ground Temperature Sensor, 
      taken from EDRs).
    - ADRs. These Ancillary data are locations of the rover, the sun,
      and the masthead, and the estimated degradation factor of the UV
      sensor due to the dust.
    - Specific EDR data such as the ASIC temperature.

    ADRs are used mainly to populate the confidence level codes found
    in ENVRDRs. Some of those confidence levels take into account the
    projection of the shadows over the sensors using positions and
    locations contained in the ADRs.
    
    Finally, with a one to one time match between TELRDR and ENVRDR,
    the resulting data is integrated into a single table.


  Data
  ====
    Each REMS ENVRDR product in the data set is an ASCII table
    containing all sensors data. Each row contains an acquisition 
    session, and columns contain the sensors values.

    Columns with related information are placed together, in the
    following order: time references (REMS clock, LMST and LTST), Wind
    Sensor, Ground Temperature Sensor, Air Temperature Sensor,
    Ultraviolet Sensor, Humidity Sensor and Pressure Sensor.

    REMS clock is synced with the spacecraft clock (SCLK) with a
    maximum deviation of 30 sg. LTST is given with respect to the
    rover position, in contrast with LTST given in REMS EDR labels and
    in other MSL data, which is given with respect to landing site.

    Ground Temperature Sensor is corrected using in-flight calibration
    data (found in the TELRDR) and Ultraviolet Sensor responsivity is
    corrected depending on the Solar Zenith Angle and the estimation
    of the dust over the photodiodes.

    A confidence level code for each sensor is included. This code
    indicates the quality of the data by a string of zeroes and ones.
    Each character represents a factor by a '1' (good condition) or a
    '0' (bad condition). The higher the number of ones, the more
    reliable the measured magnitude is. The character 'X' may be
    present in some codes for factors whose value is not known at the
    moment of the data generation. A detailed description of the
    confidence levels can be found in the REMS RDR SIS.

    In addition to confidence level codes, some sensors include an
    estimation of the uncertainty in the provided data in another
    column.

    Columns are delimited by commas and are of fixed length. Rows are
    time ordered and are separated by a carriage return/line
    feed. Each table contains a sol of measurements.

    Data may be set to UNK if their value is not known and it will
    never be (such as saturation, or a specific sensor switched off
    during acquisition). They may also be set to NULL if their value
    is not known at the moment of the release of the data set, but it
    is expected to be known in a future release. If data from a sensor
    is set to UNK or NULL, its associated confidence level code will
    also be set to UNK or NULL.


  Ancillary Data 
  ==============
    The ancillary data used in the generation of ENVRDR data are
    laboratory calibration data and ADRs.
    
    Calibration data can be found in the CALIB directory. Calibration
    files are in plain ASCII text format. They are structured in a PDS
    label-like structure, with the form DATA = VALUE, each value
    taking up one line. A file per sensor is provided.

    The following information is contained in the ADRs: 

    Solar azimuth and elevation
    Rover azimuth, elevation, altitude, pitch, yaw and roll
    Rover speed
    Masthead azimuth and elevation
    Rover temperatures
    UV sensor dust attenuation


  Coordinate System
  =================
    This data set depends on several coordinate systems used in the MSL
    project. These are:

    - MSL_TOPO frame, fixed to landing site, is used to define rover
      position.
    - MSL_LOCAL_LEVEL, fixed to the rover, is used to define rover
      orientation (roll, pitch and yaw angles)
    - MSL_RSM_ZERO_AZ frame, fixed to the rover, is used to define
      rover remote sensor mast (RSM) azimuth.
    - MSL_RSM_ZERO_EL, fixed to the rover, is used to define rover
      remote sensor mast (RSM) elevation.
    - MSL_ROVER_MECH, fixed to the rover, is used to define Sun
      position angles (azimuth and elevation)


  Software
  ========
    No software is provided in this data set. The RDR tables are simple
    ASCII files that can be displayed on UNIX, Macintosh, and PC
    platforms with common software.

Confidence Level Overview
  =========================
    ENVRDR data depend on the reliability of the calibration data and
    the ADRs, although extended calibration campaigns have been
    performed to verify them. In addition, several factors affecting
    the data quality are represented by the confidence level codes and
    the uncertainty data provides an estimation of that influence.

    Wind Sensor data of boards 2 and 3 in boom 1 are not provided as
    they were damaged during landing. In addition, the following Wind
    Sensor boards in boom 2 are not operative starting on given sols:

    - Board 1, dice 1: since sol 854
    - Board 2: since sol 1485
    - Board 3: since sol 1491
    - Board 1: since sol 1504
    
    From sol 155 onwards, the Ultraviolet Sensor has been degraded
    beyond its operational functional requirements. This is caused by
    deposition of dust over it, which has made the attenuation of the
    signal to be greater than 10%.


  Review
  ======
    The REMS RDRs were reviewed internally by the REMS team prior to
    release to the PDS. PDS also performed an external peer review of
    the REMS RDRs.


  Data Coverage and Quality
  =========================
    Each ENVRDRs table provides sensors values measured during one
    sol. Data is dependent on REMS EDRs data products, pre-flight
    calibration data, SPICE data and ancillary data.
   
    Sensors switched off at the time of measurement are set to UNK.
    Saturated Wind Sensor data or those correspondent to damaged
    boards are set to UNK. Data from thermopiles B and C of the Ground
    Temperature Sensor are included but are too noisy to be considered
    useful.

    Confidence level codes whose correspondent values are UNK or NULL
    are also set to UNK and NULL respectively.

    Degradation of the Ultraviolet Sensor from sol 155 onwards is
    flagged in its confidence level code as having an attenuation above
    10%.

  Limitations
  ===========    
    These data are not corrected from several factors such as
    external heat sources or shadows, so confidence level codes shall
    be revised carefully before using them.
   
    Missing information from Wind Sensor in the boom 1 makes it
    difficult to estimate wind data when it is coming from behind the
    rover mast in the boom 2 direction.