SCC Documentation: netcdf_file.rst@58ac47f4bbea (annotated)

netcdf_file.rst@58ac47f4bbea (annotated)

netcdf_file.rst

Thu, 15 Mar 2012 10:45:22 +0100

author: ulalume3 <binietoglou@imaa.cnr.it>
date: Thu, 15 Mar 2012 10:45:22 +0100
changeset 0: 58ac47f4bbea
permissions: -rw-r--r--

Adding first files.

 The SCC netCDF file format
 ==========================
 Rationale
 ---------
 The Single Calculus Chain (SCC) is composed by two different modules:
 -  pre-processing module ( scc\_preprocessing)
 -  optical processing module ( ELDA)
 To perfom aerosol optical retrievals the SCC needs not only the raw
 lidar data but also a certain number of parameters to use in both
 pre-processing and optical processing stages. The SCC gets these
 parameters looking at two different locations:
 -  Single Calculus Chain relational database (SCC\_DB)
 -  Input files
 There are some paramenters that can be found only in the input files
 (those ones changing from measurement to measurement), others that can
 be found only in the SCC\_DB and other ones that can be found in both
 these locations. In the last case, if a particular parameter is needed,
 the SCC will search first in the input files and then in SCC\_DB. If the
 parameter is found in the input files the SCC will keep it without
 looking into SCC\_DB.
 The input files have to be submitted to the SCC in NetCDF format. At the
 present the SCC can handle four different types of input files:
 . Raw Lidar Data
 . Sounding Data
 . Overlap
 . Lidar Ratio
 As already mentioned, the  Raw Lidar Data file contains not only the
 raw lidar data but also other parameters to use to perform the
 pre-processing and optical processing. The  Sounding Data file
 contains the data coming from a correlative radiosounding and it is used
 by the SCC for molecular density calculation. The  Overlap file
 contains the measured overlap function. The  Lidar Ratio file contains
 a lidar ratio profile to use in elastic backscatter retrievals. The
 Raw Lidar Data file is of course mandatory and the  Sounding Data,
 Overlap and  Lidar Ratio files are optional. If  Sounding Data file
 is not submitted by the user, the molecular density will be calculated
 by the SCC using the “US Standard Atmosphere 1976”. If the  Overlap
 file is not submitted by the user, the SCC will get the full overlap
 height from SCC\_DB and it will produce optical results starting from
 this height. If  Lidar Ratio file is not submitted by the user, the
 SCC will consider a fixed value for lidar ratio got from SCC\_DB.
 The user can decide to submit all these files or any number of them (of
 course the file  Raw Lidar Data is mandatory). For example the user
 can submit together with the  Raw Lidar Data file only the  Sounding
 Data file or only the  Overlap file.
 This document provides a detailed explanation about the structure of the
 NetCDF input files to use for SCC data submission. All Earlinet groups
 should read it carefully because they have to produce such kind of input
 files if they want to use the SCC for their standard lidar retrievals.
 Every comments or suggestions regarding this document can be sent to
 Giuseppe D’Amico by e-mail at ``damico@imaa.cnr.it``
 This document is available for downloading at ``www.earlinetasos.org``
 In table tab:rawdata is reported a list of dimensions, variables and
 global attributes that can be used in the NetCDF  Raw Lidar Data input
 file. For each of them it is indicated:
 -  The name. For the multidimensional variables also the corresponding
    dimensions are reported
 -  A description explaining the meaning
 -  The type
 -  If it is mandatory or optional
 As already mentioned, the SCC can get some parameters looking first in
 the  Raw Lidar Data input file and then into SCC\_DB. This means that
 to use the parameters stored in SCC\_DB the optional variables or
 optional global attributes must not appear within  Raw Lidar Data
 file. This is the suggested and recommended way to use the SCC. Please
 include optional parameters in the  Raw Lidar Data only as an
 exception.
 In table tab:sounding, tab:overlap and tab:lr are reported all the
 information about the structure of  Sounding Data,  Overlap and
 Lidar Ratio input files respectively.
 Example
 -------
 Let’s now consider an example of  Raw Lidar Data input file. Suppose
 we want to generate NetCDF input file corresponding to a measurement
 with the following properties:
 +----------------------+-------------------------------------------+
 | Start Date           | :math:`30^{th}` January 2009              |
 +----------------------+-------------------------------------------+
 | Start Time UT        | 00:00:01                                  |
 +----------------------+-------------------------------------------+
 | Stop Time UT         | 00:05:01                                  |
 +----------------------+-------------------------------------------+
 | Station Name         | Dummy station                             |
 +----------------------+-------------------------------------------+
 | Earlinet call-sign   | cc                                        |
 +----------------------+-------------------------------------------+
 | Pointing angle       | 5 degrees with respect to the zenith      |
 +----------------------+-------------------------------------------+
 Moreover suppose that this measurement is composed by the following
 lidar channels:
 . 1064 lidar channel
    +------------------------------+-------------------------------+
    | Emission wavelength=1064nm   | Detection wavelength=1064nm   |
    +------------------------------+-------------------------------+
    | Time resolution=30s          | Number of laser shots=1500    |
    +------------------------------+-------------------------------+
    | Number of bins=3000          | Detection mode=analog         |
    +------------------------------+-------------------------------+
    | Range resolution=7.5m        | Polarization state=total      |
    +------------------------------+-------------------------------+
 . 532 cross lidar channel
    +-----------------------------+---------------------------------+
    | Emission wavelength=532nm   | Detection wavelength=532nm      |
    +-----------------------------+---------------------------------+
    | Time resolution=60s         | Number of laser shots=3000      |
    +-----------------------------+---------------------------------+
    | Number of bins=5000         | Detection mode=photoncounting   |
    +-----------------------------+---------------------------------+
    | Range resolution=15m        | Polarization state=cross        |
    +-----------------------------+---------------------------------+
 . 532 parallel lidar channel
    +-----------------------------+---------------------------------+
    | Emission wavelength=532nm   | Detection wavelength=532nm      |
    +-----------------------------+---------------------------------+
    | Time resolution=60s         | Number of laser shots=3000      |
    +-----------------------------+---------------------------------+
    | Number of bins=5000         | Detection mode=photoncounting   |
    +-----------------------------+---------------------------------+
    | Range resolution=15m        | Polarization state=parallel     |
    +-----------------------------+---------------------------------+
 . 607 :math:`N_2` vibrational Raman channel
    +-----------------------------+---------------------------------+
    | Emission wavelength=532nm   | Detection wavelength=607nm      |
    +-----------------------------+---------------------------------+
    | Time resolution=60s         | Number of laser shots=3000      |
    +-----------------------------+---------------------------------+
    | Number of bins=5000         | Detection mode=photoncounting   |
    +-----------------------------+---------------------------------+
    | Range resolution=15m                                          |
    +-----------------------------+---------------------------------+
 Finally let’s assume we have also performed dark measurements before the
 lidar measurements from the 23:50:01 UT up to 23:53:01 UT of
 :math:`^\mathrmth` January 2009.
 Dimensions
 ~~~~~~~~~~
 Looking at table tab:rawdata we have to fix the following dimensions:
 ::
     points
     channels
     time
     nb_of_time_scales
     scan_angles
     time_bck
 The dimension ``time`` is unlimited so we don’t have to fix it.
 We have 4 lidar channels so:
 ::
     channels=4
 Regarding the dimension ``points`` we have only one channel with a
 number of vertical bins equal to 3000 (the 1064nm) and all other
 channels with 5000 vertical bins. In cases like this the dimension
 ``points`` has to be fixed to the maximum number of vertical bins so:
 ::
     points=5000
 Moreover only one channel (1064nm) is acquired with a time resolution of
 seconds, all the other channels have a time resolution of 60 seconds.
 This means that we have to define two different time scales. We have to
 set:
 ::
     nb_of_time_scales=2
 The measurement is performed only at one scan angle (5 degrees with
 respect to the zenith) so:
 ::
     scan_angles=1
 We have 3 minutes of dark measurements and two different time scales one
 with 60 seconds time resolution and the other one with 30 seconds time
 resolution. So we will have 3 different dark profiles for the channels
 acquired with the first time scale and 6 for the lidar channels acquired
 with the second time scale. We have to fix the dimension ``time_bck`` as
 the maximum between these values:
 ::
     time_bck=6
 Variables
 ~~~~~~~~~
 In this section it will be explained how to fill all the possible
 variables either mandatory or optional of  Raw Lidar Data input file.
 Raw_Data_Start_Time(time, nb_of_time_scales)
    This 2 dimensional mandatory array has to contain the acquisition
    start time (in seconds from the time given by the global attribute
    ``RawData_Start_Time_UT``) of each lidar profile. In this example we
    have two different time scales: one is characterized by steps of 30
    seconds (the 1064nm is acquired with this time scale) the other by
    steps of 60 seconds (532cross, 532parallel and 607nm). Moreover the
    measurement start time is 00:00:01 UT and the measurement stop time
    is 00:05:01 UT. In this case we have to define:
    ::
        Raw_Data_Start_Time =
 , 0,
 , 30,
 , 60,
 , 90,
 , 120,
          _, 150,
          _, 180,
          _, 210,
          _, 240,
          _, 270 ;
    The order used to fill this array defines the correspondence between
    the different time scales and the time scale index. In this example
    we have a time scale index of 0 for the time scale with steps of 60
    seconds and a time scale index of 1 for the other one.
 Raw_Data_Stop_Time(time, nb_of_time_scales)
    The same as previous item but for the data acquisition stop time.
    Following a similar procedure we have to define:
    ::
        Raw_Data_Stop_Time =
 , 30,
 , 60,
 , 90,
 , 120,
 , 150,
          _, 180,
          _, 210,
          _, 240,
          _, 270,
          _, 300 ;
 Raw_Lidar_Data(time, channels, points)
    This 3 dimensional mandatory array has to be filled with the
    time-series of raw lidar data. The photoncounting profiles have to
    submitted in counts (so as integers) while the analog ones in mV. The
    order the user chooses to fill this array defines the correspondence
    between channel index and lidar data.
    For example if we fill this array in such way that:
    +-------------------------------------+------------------------------------------------------------+
    | Raw_Lidar_Data(time,0,points        | :math:`\rightarrow` is the time-series of 1064 nm          |
    +-------------------------------------+------------------------------------------------------------+
    | Raw_Lidar_Data(time,1,points        | :math:`\rightarrow` is the time-series of 532 cross        |
    +-------------------------------------+------------------------------------------------------------+
    | Raw_Lidar_Data(time,2,points        | :math:`\rightarrow` is the time-series of 532 parallel     |
    +-------------------------------------+------------------------------------------------------------+
    | Raw_Lidar_Data(time,3,points        | :math:`\rightarrow` is the time-series of 607 nm           |
    +-------------------------------------+------------------------------------------------------------+
    from now on the channel index 0 is associated to the 1064 channel,
 to the 532 cross, 2 to the 532 parallel and 3 to the 607nm.
 Raw_Bck_Start_Time(time_bck, nb_of_time_scales)
    This 2 dimensional optional array has to contain the acquisition
    start time (in seconds from the time given by the global attribute
    ``RawBck_Start_Time_UT``) of each dark measurements profile.
    Following the same procedure used for the variable
    ``Raw_Data_Start_Time`` we have to define:
    ::
        Raw_Bck_Start_Time =
 , 0,
 , 30,
 , 60,
          _, 90,
          _, 120,
          _, 150;
 Raw_Bck_Stop_Time(time_bck, nb_of_time_scales)
    The same as previous item but for the dark acquisition stop time.
    Following a similar procedure we have to define:
    ::
        Raw_Bck_Stop_Time =
 , 30,
 , 60,
 , 90,
          _, 120,
          _, 150,
          _, 180 ;
 Background_Profile(time_bck, channels, points)
    This 3 dimensional optional array has to be filled with the
    time-series of the dark measurements data. The photoncounting
    profiles have to submitted in counts (so as integers) while the
    analog ones in mV. The user has to fill this array following the same
    order used in filling the array ``Raw_Lidar_Data``:
    +---------------------------------------------+----------------------------------------------------------+
    | Background_Profile(time_bck,0,points        | :math:`\rightarrow` dark time-series at 1064 nm          |
    +---------------------------------------------+----------------------------------------------------------+
    | Background_Profile(time_bck,1,points        | :math:`\rightarrow` dark time-series at 532 cross        |
    +---------------------------------------------+----------------------------------------------------------+
    | Background_Profile(time_bck,2,points        | :math:`\rightarrow` dark time-series at 532 parallel     |
    +---------------------------------------------+----------------------------------------------------------+
    | Background_Profile(time_bck,3,points        | :math:`\rightarrow` dark time-series at 607 nm           |
    +---------------------------------------------+----------------------------------------------------------+
 channel_ID(channels)
    This mandatory array provides the link between the channel index
    within the  Raw Lidar Data input file and the channel ID in
    SCC\_DB. To fill this variable the user has to know which channel IDs
    in SCC\_DB correspond to his lidar channels. For this purpose the
    SCC, in its final version will provide to the user a special tool to
    get these channel IDs through a Web interface. At the moment this
    interface is not yet available and these channel IDs will be
    communicated directly to the user by the NA5 people.
    Anyway to continue the example let’s suppose that the four lidar
    channels taken into account are mapped into SCC\_DB with the
    following channel IDs:
    +----------------+--------------------------------------+
    | 1064 nm        | :math:`\rightarrow` channel ID=7     |
    +----------------+--------------------------------------+
    | 532 cross      | :math:`\rightarrow` channel ID=5     |
    +----------------+--------------------------------------+
    | 532 parallel   | :math:`\rightarrow` channel ID=6     |
    +----------------+--------------------------------------+
    | 607 nm         | :math:`\rightarrow` channel ID=8     |
    +----------------+--------------------------------------+
     In this case we have to define:
    ::
        channel_ID = 7, 5, 6, 8 ;
 id_timescale(channels)
    This mandatory array is introduced to determine which time scale is
    used for the acquisition of each lidar channel. In particular this
    array defines the link between the channel index and the time scale
    index. In our example we have two different time scales. Filling the
    arrays ``Raw_Data_Start_Time`` and ``Raw_Data_Stop_Time`` we have
    defined a time scale index of 0 for the time scale with steps of 60
    seconds and a time scale index of 1 for the other one with steps of
 seconds. In this way this array has to be set as:
    ::
        id_timescale = 1, 0, 0, 0 ;
 Laser_Pointing_Angle(scan_angles
    This mandatory array contains all the scan angles used in the
    measurement. In our example we have only one scan angle of 5 degrees
    with respect to the zenith, so we have to define:
    ::
        Laser_Pointing_Angle = 5 ;
 Laser_Pointing_Angle_of_Profiles(time, nb_of_time_scales)
    This mandatory array is introduced to determine which scan angle is
    used for the acquisition of each lidar profile. In particular this
    array defines the link between the time and time scales indexes and
    the scan angle index. In our example we have a single scan angle that
    has to correspond to the scan angle index 0. So this array has to be
    defined as:
    ::
        Laser_Pointing_Angle_of_Profiles =
 , 0,
 , 0,
 , 0,
 , 0,
 , 0,
          _, 0,
          _, 0,
          _, 0,
          _, 0,
          _, 0 ;
 Laser_Shots(time, channels)
    This mandatory array stores the laser shots accumulated at each time
    for each channel. In our example the number of laser shots
    accumulated is 1500 for the 1064nm channels and 3000 for all the
    other channels. Moreover the laser shots do not change with the time.
    So we have to define this array as:
    ::
         Laser_Shots =
 , 3000, 3000, 3000,
 , 3000, 3000, 3000,
 , 3000, 3000, 3000,
 , 3000, 3000, 3000,
 , 3000, 3000, 3000,
 , _, _, _,
 , _, _, _,
 , _, _, _,
 , _, _, _,
 , _, _, _ ;
 Emitted_Wavelength(channels)
    This optional array defines the link between the channel index and
    the emission wavelength for each lidar channel. The wavelength has to
    be expressed in nm. This information can be also taken from SCC\_DB.
    In our example we have:
    ::
        Emitted_Wavelength = 1064, 532, 532, 532 ;
 Detected_Wavelength(channels)
    This optional array defines the link between the channel index and
    the detected wavelength for each lidar channel. Here detected
    wavelength means the value of center of interferential filter
    expressed in nm. This information can be also taken from SCC\_DB. In
    our example we have:
    ::
        Detected_Wavelength = 1064, 532, 532, 607 ;
 Raw_Data_Range_Resolution(channels)
    This optional array defines the link between the channel index and
    the raw range resolution for each channel. If the scan angle is
    different from zero this quantity is different from the vertical
    resolution. More precisely if :math:`\alpha` is the scan angle used
    and :math:`\Delta z` is the range resolution the vertical
    resolution is calculated as :math:`\Delta
    z'=\Delta z \cos\alpha`. This array has to be filled with
    :math:`\Delta z` and not with :math:`\Delta z'`. The unit is
    meters. This information can be also taken from SCC\_DB. In our
    example we have:
    ::
        Raw_Data_Range_Resolution = 7.5, 15.0, 15.0, 15.0 ;
 ID_Range(channels)
    This optional array defines if a particular channel is configured as
    high, low or ultranear range channel. In particular a value 0
    indicates a low range channel, a value 1 a high range channel and a
    value of 2 an ultranear range channel. If for a particular channel
    you don’t separate between high and low range channel, please set the
    corresponding value to 1. This information can be also taken from
    SCC\_DB. In our case we have to set:
    ::
        ID_Range = 1, 1, 1, 1 ;
 Scattering_Mechanism(channels)
    This optional array defines the scattering mechanism involved in
    each lidar channel. In particular the following values are adopted:
    +------+---------------------------------------------------------------------------------------------+
    | 0    | :math:`\rightarrow` Total elastic backscatter                                               |
    +------+---------------------------------------------------------------------------------------------+
    | 1    | :math:`\rightarrow` :math:`N_2` vibrational Raman backscatter                               |
    +------+---------------------------------------------------------------------------------------------+
    | 2    | :math:`\rightarrow` Cross polarization elastic backscatter                                  |
    +------+---------------------------------------------------------------------------------------------+
    | 3    | :math:`\rightarrow` Parallel polarization elastic backscatter                               |
    +------+---------------------------------------------------------------------------------------------+
    | 4    | :math:`\rightarrow` :math:`H_2O` vibrational Raman backscatter                              |
    +------+---------------------------------------------------------------------------------------------+
    | 5    | :math:`\rightarrow` Rotational Raman Stokes line close to elastic line                      |
    +------+---------------------------------------------------------------------------------------------+
    | 6    | :math:`\rightarrow` Rotational Raman Stokes line far from elastic line                      |
    +------+---------------------------------------------------------------------------------------------+
    | 7    | :math:`\rightarrow` Rotational Raman anti-Stokes line close to elastic line                 |
    +------+---------------------------------------------------------------------------------------------+
    | 8    | :math:`\rightarrow` Rotational Raman anti-Stokes line far from elastic line                 |
    +------+---------------------------------------------------------------------------------------------+
    | 9    | :math:`\rightarrow` Rotational Raman Stokes and anti-Stokes lines close to elastic line     |
    +------+---------------------------------------------------------------------------------------------+
    | 10   | :math:`\rightarrow` Rotational Raman Stokes and anti-Stokes lines far from elastic line     |
    +------+---------------------------------------------------------------------------------------------+
    This information can be also taken from SCC\_DB. In our example we have:
    ::
         Scattering_Mechanism = 0, 2, 3, 1 ;
 Acquisition_Mode(channels)
    This optional array defines the acquisition mode (analog or
    photoncounting) involved in each lidar channel. In particular a value
    of 0 means analog mode and 1 photoncounting mode. This information
    can be also taken from SCC\_DB. In our example we have:
    ::
         Acquisition_Mode = 0, 1, 1, 1 ;
 Laser_Repetition_Rate(channels)
    This optional array defines the repetition rate in Hz used to
    acquire each lidar channel. This information can be also taken from
    SCC\_DB. In our example we are supposing we have only one laser with
    a repetition rate of 50 Hz so we have to set:
    ::
        Laser_Repetition_Rate = 50, 50, 50, 50 ;
 Dead_Time(channels)
    This optional array defines the dead time in ns associated to each
    lidar channel. The SCC will use the values given by this array to
    correct the photoncounting signals for dead time. Of course for
    analog signals no dead time correction will be applied (for analog
    channels the corresponding dead time values have to be set to
    undefined value). This information can be also taken from SCC\_DB. In
    our example the 1064 nm channel is acquired in analog mode so the
    corresponding dead time value has to be undefined. If we suppose a
    dead time of 10 ns for all other channels we have to set:
    ::
         Dead_Time = _, 10, 10, 10 ;
 Dead_Time_Corr_Type(channels
    This optional array defines which kind of dead time correction has
    to be applied on each photoncounting channel. The SCC will correct
    the data supposing a not-paralyzable channel if a value of 0 is found
    while a paralyzable channel is supposed if a value of 1 is found. Of
    course for analog signals no dead time correction will be applied and
    so the corresponding values have to be set to undefined value. This
    information can be also taken from SCC\_DB. In our example the 1064
    nm channel is acquired in analog mode so the corresponding has to be
    undefined. If we want to consider all the photoncounting signals as
    not-paralyzable ones: we have to set:
    ::
         Dead_Time_Corr_Type = _, 0, 0, 0 ;
 Trigger_Delay(channels)
    This optional array defines the delay (in ns) of the middle of the
    first rangebin with respect to the output laser pulse for each lidar
    channel. The SCC will use the values given by this array to correct
    for trigger delay. This information can be also taken from SCC\_DB.
    Let’s suppose that in our example all the photoncounting channels are
    not affected by this delay and only the analog channel at 1064nm is
    acquired with a delay of 50ns. In this case we have to set:
    ::
        Trigger_Delay = 50, 0, 0, 0 ;
 Background_Mode(channels
    This optional array defines how the atmospheric background has to be
    subtracted from the lidar channel. Two options are available for the
    calculation of atmospheric background:
    #. Average in the far field of lidar channel. In this case the value
       of this variable has to be 1
    #. Average within pre-trigger bins. In this case the value of this
       variable has to be 0
    This information can be also taken from SCC\_DB. Let’s suppose in our
    example we use the pre-trigger for the 1064nm channel and the far
    field for all other channels. In this case we have to set:
    ::
        Background_Mode = 0, 1, 1, 1 ;
 Background_Low(channels)
    This mandatory array defines the minimum altitude (in meters) to
    consider in calculating the atmospheric background for each channel.
    In case pre-trigger mode is used the corresponding value has to be
    set to the rangebin to be used as lower limit (within pre-trigger
    region) for background calculation. In our example, if we want to
    calculate the background between 30000 and 50000 meters for all
    photoncounting channels and we want to use the first 500 pre-trigger
    bins for the background calculation for the 1064nm channel we have to
    set:
    ::
         Background_Low= 0, 30000, 30000, 30000 ;
 Background_High(channels)
    This mandatory array defines the maximum altitude (in meters) to
    consider in calculating the atmospheric background for each channel.
    In case pre-trigger mode is used the corresponding value has to be
    set to the rangebin to be used as upper limit (within pre-trigger
    region) for background calculation. In our example, if we want to
    calculate the background between 30000 and 50000 meters for all
    photoncounting channels and we want to use the first 500 pre-trigger
    bins for the background calculation for the 1064nm channel we have to
    set:
    ::
        Background_High = 500, 50000, 50000, 50000 ;
 Molecular_Calc
    This mandatory variable defines the way used by SCC to calculate the
    molecular density profile. At the moment two options are available:
    #. US Standard Atmosphere 1976. In this case the value of this
       variable has to be 0
    #. Radiosounding. In this case the value of this variable has to be 1
    If we decide to use the option 1. we have to provide also the
    measured pressure and temperature at lidar station level. Indeed if
    we decide to use the option 2. a radiosounding file has to be
    submitted separately in NetCDF format (the structure of this file is
    summarized in table tab:sounding). Let’s suppose we want to use the
    option 1. so:
    ::
        Molecular_Calc = 0 ;
 Pressure_at_Lidar_Station
    Because we have chosen the US Standard Atmosphere for calculation of
    the molecular density profile we have to give the pressure in hPa at
    lidar station level:
    ::
        Pressure_at_Lidar_Station = 1010 ;
 Temperature_at_Lidar_Station
    Because we have chosen the US Standard Atmosphere for calculation of
    the molecular density profile we have to give the temperature in C at
    lidar station level:
    ::
        Temperature_at_Lidar_Station = 19.8 ;
 Depolarization_Factor(channels)
    This array is required only for lidar systems that use the two
    depolarization channels for the backscatter retrieval. It represents
    the factor :math:`f` to calculate the total backscatter signal
    :math:`S_t` combining its cross :math:`S_c` and parallel
    :math:`S_p` components: :math:`S_t=S_p+fS_c`. This factor is
    mandatory only for systems acquiring :math:`S_c` and :math:`S_p`
    and not :math:`S_t`. For systems acquiring :math:`S_c`,
    :math:`S_p` and :math:`S_t` this factor is optional and it will
    be used only for depolarizaton ratio calculation. Moreover only the
    values of the array corresponding to cross polarization channels will
    be considered; all other values will be not taken into account and
    should be set to undefined value. In our example for the wavelength
 nm we have only the cross and the parallel components and not the
    total one. So we have to give the value of this factor only in
    correspondence of the 532nm cross polarization channel that
    corresponds to the channel index 1. Suppose that this factor is 0.88.
    Moreover, because we don’t have any other depolarization channels we
    have also to set all other values of the array to undefined value.
    ::
        Depolarization_Factor = _,0.88,_,_ ;
 LR_Input(channels)
    This array is required only for lidar channels for which elastic
    backscatter retrieval has to be performed. It defines the lidar ratio
    to be used within this retrieval. Two options are available:
    #. The user can submit a lidar ratio profile. In this case the value
       of this variable has to be 0.
    #. A fixed value of lidar ratio can be used. In this case the value
       of this variable has to be 1.
    If we decide to use the option 1. a lidar ratio file has to be
    submitted separately in NetCDF format (the structure of this file is
    summarized in table tab:lr). If we decide to use the option 2. the
    fixed value of lidar ratio will be taken from SCC\_DB. In our example
    we have to give a value of this array only for the 1064nm lidar
    channel because for the 532nm we will be able to retrieve a Raman
    backscatter coefficient. In case we want to use the fixed value
    stored in SCC\_DB we have to set:
    ::
        LR_Input = 1,_,_,_ ;
 DAQ_Range(channels)
    This array is required only if one or more lidar signals are
    acquired in analog mode. It gives the analog scale in mV used to
    acquire the analog signals. In our example we have only the 1064nm
    channel acquired in analog mode. If we have used a 100mV analog scale
    to acquire this channel we have to set:
    ::
        DAQ_Range = 100,_,_,_ ;
 Global attributes
 ~~~~~~~~~~~~~~~~~
 Measurement_ID
    This mandatory global attribute defines the measurement ID
    corresponding to the actual lidar measurement. It is a string
    composed by 12 characters. The first 8 characters give the start date
    of measurement in the format YYYYMMDD. The next 2 characters give the
    Earlinet call-sign of the station. The last 2 characters are used to
    distinguish between different time-series within the same date. In
    our example we have to set:
    ::
         Measurement_ID= "20090130cc00" ;
 RawData_Start_Date
    This mandatory global attribute defines the start date of lidar
    measurements in the format YYYYMMDD. In our case we have:
    ::
         RawData_Start_Date = "20090130" ;
 RawData_Start_Time_UT
    This mandatory global attribute defines the UT start time of lidar
    measurements in the format HHMMSS. In our case we have:
    ::
         RawData_Start_Time_UT = "000001" ;
 RawData_Stop_Time_UT``
    This mandatory global attribute defines the UT stop time of lidar
    measurements in the format HHMMSS. In our case we have:
    ::
         RawData_Stop_Time_UT = "000501" ;
 RawBck_Start_Date
    This optional global attribute defines the start date of dark
    measurements in the format YYYYMMDD. In our case we have:
    ::
         RawBck_Start_Date = "20090129" ;
 RawBck_Start_Time_UT
    This optional global attribute defines the UT start time of dark
    measurements in the format HHMMSS. In our case we have:
    ::
         RawBck_Start_Time_UT = "235001" ;
 RawBck_Stop_Time_UT
    This optional global attribute defines the UT stop time of dark
    measurements in the format HHMMSS. In our case we have:
    ::
         RawBck_Stop_Time_UT = "235301" ;
 Example of file (CDL format)
 ----------------------------
 To summarize we have the following NetCDF  Raw Lidar Data file (in CDL
 format):
 ::
     dimensions:
             points = 5000 ;
             channels = 4 ;
             time = UNLIMITED ; // (10 currently)
             nb_of_time_scales = 2 ;
             scan_angles = 1 ;
             time_bck = 6 ;
     variables:
             int channel_ID(channels) ;
             int Laser_Repetition_Rate(channels) ;
             double Laser_Pointing_Angle(scan_angles) ;
             int ID_Range(channels) ;
             int Scattering_Mechanism(channels) ;
             double Emitted_Wavelength(channels) ;
             double Detected_Wavelength(channels) ;
             double Raw_Data_Range_Resolution(channels) ;
             int Background_Mode(channels) ;
             double Background_Low(channels) ;
             double Background_High(channels) ;
             int Molecular_Calc ;
             double Pressure_at_Lidar_Station ;
             double Temperature_at_Lidar_Station ;
             int id_timescale(channels) ;
             double Dead_Time(channels) ;
             int Dead_Time_Corr_Type(channels) ;
             int Acquisition_Mode(channels) ;
             double Trigger_Delay(channels) ;
             int LR_Input(channels) ;
             int Laser_Pointing_Angle_of_Profiles(time, nb_of_time_scales) ;
             int Raw_Data_Start_Time(time, nb_of_time_scales) ;
             int Raw_Data_Stop_Time(time, nb_of_time_scales) ;
             int Raw_Bck_Start_Time(time_bck, nb_of_time_scales) ;
             int Raw_Bck_Stop_Time(time_bck, nb_of_time_scales) ;
             int Laser_Shots(time, channels) ;
             double Raw_Lidar_Data(time, channels, points) ;
             double Background_Profile(time_bck, channels, points) ;
             double DAQ_Range(channels) ;
     // global attributes:
                     :Measurement_ID = "20090130cc00" ;
                     :RawData_Start_Date = "20090130" ;
                     :RawData_Start_Time_UT = "000001" ;
                     :RawData_Stop_Time_UT = "000501" ;
                     :RawBck_Start_Date = "20090129" ;
                     :RawBck_Start_Time_UT = "235001" ;
                     :RawBck_Stop_Time_UT = "235301" ;
     data:
      channel_ID = 7, 5, 6, 8 ;
      Laser_Repetition_Rate = 50, 50, 50, 50 ;
      Laser_Pointing_Angle = 5 ;
      ID_Range = 1, 1, 1, 1 ;
      Scattering_Mechanism = 0, 2, 3, 1 ;
      Emitted_Wavelength = 1064, 532, 532, 532 ;
      Detected_Wavelength = 1064, 532, 532, 607 ;
      Raw_Data_Range_Resolution = 7.5, 15, 15, 15 ;
      Background_Mode = 0, 1, 1, 1 ;
      Background_Low = 0, 30000, 30000, 30000 ;
      Background_High = 500, 50000, 50000, 50000 ;
      Molecular_Calc = 0 ;
      Pressure_at_Lidar_Station = 1010 ;
      Temperature_at_Lidar_Station = 19.8 ;
      id_timescale = 1, 0, 0, 0 ;
      Dead_Time = _, 10, 10, 10 ;
      Dead_Time_Corr_Type = _, 0, 0, 0 ;
      Acquisition_Mode = 0, 1, 1, 1 ;
      Trigger_Delay = 50, 0, 0, 0 ;
      LR_Input = 1,_,_,_ ;
      DAQ_Range = 100,_,_,_ ;
      Laser_Pointing_Angle_of_Profiles =
 , 0,
 , 0,
 , 0,
 , 0,
 , 0,
       _, 0,
       _, 0,
       _, 0,
       _, 0,
       _, 0 ;
      Raw_Data_Start_Time =
 , 0,
 , 30,
 , 60,
 , 90,
 , 120,
       _, 150,
       _, 180,
       _, 210,
       _, 240,
       _, 270 ;
      Raw_Data_Stop_Time =
 , 30,
 , 60,
 , 90,
 , 120,
 , 150,
       _, 180,
       _, 210,
       _, 240,
       _, 270,
       _, 300 ;
      Raw_Bck_Start_Time =
 , 0,
 , 30,
 , 60,
       _, 90,
       _, 120,
       _, 150;
     Raw_Bck_Stop_Time =
 , 30,
 , 60,
 , 90,
       _, 120,
       _, 150,
       _, 180 ;
      Laser_Shots =
 , 3000, 3000, 3000,
 , 3000, 3000, 3000,
 , 3000, 3000, 3000,
 , 3000, 3000, 3000,
 , 3000, 3000, 3000,
 , _, _, _,
 , _, _, _,
 , _, _, _,
 , _, _, _,
 , _, _, _ ;
      Raw_Lidar_Data = ...
      Background_Profile = ...
 Please keep in mind that in case you submit a file like the previous one
 all the parameters present in it will be used by the SCC even if you
 have different values for the same parameters within the SCC\_DB. If you
 want to use the values already stored in SCC\_DB (this should be the
 usual way to use SCC) the  Raw Lidar Data input file has to be
 modified as follows:
 ::
     dimensions:
             points = 5000 ;
             channels = 4 ;
             time = UNLIMITED ; // (10 currently)
             nb_of_time_scales = 2 ;
             scan_angles = 1 ;
             time_bck = 6 ;
     variables:
             int channel_ID(channels) ;
             double Laser_Pointing_Angle(scan_angles) ;
             double Background_Low(channels) ;
             double Background_High(channels) ;
             int Molecular_Calc ;
             double Pressure_at_Lidar_Station ;
             double Temperature_at_Lidar_Station ;
             int id_timescale(channels) ;
             int Laser_Pointing_Angle_of_Profiles(time, nb_of_time_scales) ;
             int Raw_Data_Start_Time(time, nb_of_time_scales) ;
             int Raw_Data_Stop_Time(time, nb_of_time_scales) ;
             int Raw_Bck_Start_Time(time_bck, nb_of_time_scales) ;
             int Raw_Bck_Stop_Time(time_bck, nb_of_time_scales) ;
             int LR_Input(channels) ;
             int Laser_Shots(time, channels) ;
             double Raw_Lidar_Data(time, channels, points) ;
             double Background_Profile(time_bck, channels, points) ;
             double DAQ_Range(channels) ;
     // global attributes:
                     :Measurement_ID = "20090130cc00" ;
                     :RawData_Start_Date = "20090130" ;
                     :RawData_Start_Time_UT = "000001" ;
                     :RawData_Stop_Time_UT = "000501" ;
                     :RawBck_Start_Date = "20090129" ;
                     :RawBck_Start_Time_UT = "235001" ;
                     :RawBck_Stop_Time_UT = "235301" ;
     data:
      channel_ID = 7, 5, 6, 8 ;
      Laser_Pointing_Angle = 5 ;
      Background_Low = 0, 30000, 30000, 30000 ;
      Background_High = 500, 50000, 50000, 50000 ;
      Molecular_Calc = 0 ;
      Pressure_at_Lidar_Station = 1010 ;
      Temperature_at_Lidar_Station = 19.8 ;
      id_timescale = 1, 0, 0, 0 ;
      LR_Input = 1,_,_,_ ;
      DAQ_Range = 100,_,_,_ ;
      Laser_Pointing_Angle_of_Profiles =
 , 0,
 , 0,
 , 0,
 , 0,
 , 0,
       _, 0,
       _, 0,
       _, 0,
       _, 0,
       _, 0 ;
      Raw_Data_Start_Time =
 , 0,
 , 30,
 , 60,
 , 90,
 , 120,
       _, 150,
       _, 180,
       _, 210,
       _, 240,
       _, 270 ;
      Raw_Data_Stop_Time =
 , 30,
 , 60,
 , 90,
 , 120,
 , 150,
       _, 180,
       _, 210,
       _, 240,
       _, 270,
       _, 300 ;
      Raw_Bck_Start_Time =
 , 0,
 , 30,
 , 60,
       _, 90,
       _, 120,
       _, 150;
      Raw_Bck_Stop_Time =
 , 30,
 , 60,
 , 90,
       _, 120,
       _, 150,
       _, 180 ;
      Laser_Shots =
 , 3000, 3000, 3000,
 , 3000, 3000, 3000,
 , 3000, 3000, 3000,
 , 3000, 3000, 3000,
 , 3000, 3000, 3000,
 , _, _, _,
 , _, _, _,
 , _, _, _,
 , _, _, _,
 , _, _, _ ;
      Raw_Lidar_Data = ...
      Background_Profile = ...
 This example file contains the minimum collection of mandatory
 information that has to be found within the  Raw Lidar Data input
 file. If it is really necessary, the user can decide to add to these
 mandatory parameters any number of additional parameters considered in
 the previous example.
 Finally, suppose we want to make the following changes with respect to
 the previous example:
 #. use a sounding file for molecular density calculation instead of “US
    Standar Atmosphere 1976”
 #. supply a lidar ratio profile to use in elastic backscatter retrieval
    instead of a fixed value
 #. provide a overlap function for overlap correction
 In this case we have to generate the following NetCDF additional files:
 rs_20090130cc00.nc
     The name of  Sounding Data file has to be computed as follows:
     ``"rs_"``+``Measurement_ID``
     The structure of this file is summarized in table tab:sounding.
 ov_20090130cc00.nc
     The name of  Overlap file has to be computed as follows:
     ``"ov_"``+``Measurement_ID``
     The structure of this file is summarized in table tab:overlap.
 lr_20090130cc00.nc
     The name of  Lidar Ratio file has to be computed as follows:
     ``"lr_"``+``Measurement_ID``
     The structure of this file is summarized in table tab:lr.
 Moreover we need to apply the following changes to the  Raw Lidar Data
 input file:
 . Change the value of the variable ``Molecular_Calc`` as follows:
    ::
        Molecular_Calc = 1 ;
    Of course the variables ``Pressure_at_Lidar_Station`` and
    ``Temperature_at_Lidar_Station`` are not necessary anymore.
 . Change the values of the array ``LR_Input`` as follows:
    ::
        LR_Input = 0,_,_,_ ;
 . Add the global attribute ``Sounding_File_Name``
    ::
        Sounding_File_Name = "rs_20090130cc00.nc" ;
 . Add the global attribute ``LR_File_Name``
    ::
        LR_File_Name = "lr_20090130cc00.nc" ;
 . Add the global attribute ``Overlap_File_Name``
    ::
        Overlap_File_Name = "ov_20090130cc00.nc" ;

Mercurial > public > sccdocs / annotate

netcdf_file.rst@58ac47f4bbea (annotated)

netcdf_file.rst