SCC Documentation: docs/faq/faq.rst@9391c86dee91 (annotated)

docs/faq/faq.rst@9391c86dee91 (annotated)

docs/faq/faq.rst

Mon, 05 Dec 2016 13:20:17 +0200

author: Ioannis Binietoglou <binietoglou@noa.gr>
date: Mon, 05 Dec 2016 13:20:17 +0200
changeset 88: 9391c86dee91
parent 83: 44e5ebae6813
child 127: 88d912df1073
permissions: -rw-r--r--

New netcdf structure pdf file.

 Frequently asked questions
 ==========================
 Using an ancillary file
 ~~~~~~~~~~~~~~~~~~~~~~~
 Q: I use the "Quick upload" to submit a measurement file together with an ancillary
 file. Will this ancillary file be used in the processing of my measurement.
 A: No. Which ancillary file is used when processing a measurement is defined in the
 measurement netcdf file. If your new ancillary file is not  mentioned in the measurement
 file, it will not be used.
 Reusing an ancillary file
 ~~~~~~~~~~~~~~~~~~~~~~~~~
 Q: I want to use one ancillary file (ex. an overlap file) in the processing of multiple
 measurements. Do I need to submit it multiple times?
 A: No. You just need to define the file name of the file to use in the measurements
 netcdf file.
 Deleting an ancillary file
 ~~~~~~~~~~~~~~~~~~~~~~~~~~
 Q: I want to delete an uploaded ancillary file but in the "Admin" interface I can't
 find a "Delete" button.
 A: Probably the ancillary file you are trying to delete is needed by some uploaded
 measurement, and for this reason you are not allowed to delete it. You will need
 to delete the corresponding measurements first, before deleting the ancillary file.
 Clouds in the data
 ~~~~~~~~~~~~~~~~~~
 Q: Is it necessary to provide only measurement periods with cloud free conditions and
 homogeneous atmosphere or is this part of ELDA or the pre-processing of the data?
 For example in cases with scattered low cumulus clouds.
 A: At moment you should provide cloud free data because the automatic cloud screening
 is not yet implemented in the SCC. We are working on this issue and hopefully the new
 module will be implemented at the end of ACTRIS project.
 Minimum number of analog files to submit
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 Q: Is it possible to submit a single analog profile for processing?
 A: It depepends on the type of product you ar trying to calculate.
 * If the product is a linear polarization calibration,  it is possible to submit a timeseries containing
   one single analog profile. This is allowed because some stations are performing depolarization calibration measurements using
   single profiles. If the errors are present they will be taken into account for the calculation of the error on calibration constant.
   If they are not the error on calibration constant is calculate as the standard deviation withing the calibration range.
 * If the product to calculate is NOT a linear polarization calibration, if the raw analog error have not been provided we need at
   least 3 raw profile in order to compute the statistical error
 High/low range channels
 ~~~~~~~~~~~~~~~~~~~~~~~
 Q: What is the definition of low range, ultra near range and high range channels?
 Are there any threshold values?
 A: There are no threshold values defined for the different range types. This information
 is used only in the gluing procedures just to identify which channel should be taken
 as low range and which one as far range. If no gluing is applied by the SCC the range
 id flags are not taken into account.
 Extra netcdf parameters
 ~~~~~~~~~~~~~~~~~~~~~~~
 Q: Is it possible for documentation purposes to put own parameters in the SCC-NetCDF file?
 For example who created the file.... Are there any reasons against this?
 A: Technically as long as you use not standard SCC variables for your own parameters there are
 no problems for the SCC. It will just ignore these not standard variables.
 Netcdf version
 ~~~~~~~~~~~~~~
 Q: Which NetCDF version is to use? NetCDF3, NetCDF3 Classic, NetCDF4, NetCDF4 Classic?
 A: The NetCDF libraries 4.1.3 are used in all the SCC modules. So all the NetCDF formats you
 have indicated should be compatible with the SCC (we have tested NetCDF3 and NetCDF4).
 Lidar ratio
 ~~~~~~~~~~~
 Q: What are the values for the lidar ratio used in the SCC_DB?
 A: The values of (fixed) lidar ratio used by the SCC in the elastic retrieval can be set by
 the user using the SCC web interface. In particular you can define a lidar ratio value for
 each elastic backscatter product: in the product page there is the section "Elastic Backscatter
 options" in which there is the field "Fixed lr". In case you want to use a lidar ratio profile
 you should set LR_Input accordingly and provide an external LR profile NetCDF file
 (see documentation on SCC file format).
 Calculation of Raman and elastic backscatter
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 Q: In cases of measurements where Raman channels are available, the SCC will calculate the Raman
 backscatter profile. If I want to retrieve Klett-retrievals for this channel, too (e.g 532nm)
 is it sufficient to set the value in LR_input(channels) to 1 or 0 plus a LR-profile to get
 both retrievals?
 A: No. In general, for each lidar configuration you can define a set of optical products to be calculated
 for that configuration using the SCC web interface. So suppose you have a system with 532nm and 607nm
 channels. In this case you have 2 options:
 . Raman backscatter and extinction in the e532 file and Raman backscatter (full resolution) in the b532
    file. In this case you should associate to the configuration a product of type "lidar ratio and  will
    produce the e532 file and a product of type "Raman backscatter" which will produce the b532 file
 . Raman backscatter and extinction in the e532 file and elastic backscatter in the b532 file.
    In this case you should associate to the configuration a product of type "lidar ratio and extinction"
    which will produce the e532 file and a product of type "elastic backscatter" which will produce the b532 file
 Note: you cannot calculate a b532 file containing the Raman and elastic backscatters at the same time.
 The reason is that the 2 products will produce an output file with the same name (according to the EARLINET rules).
 Moreover in general, it makes no sense to calculate the elastic backscatter when you can calculate the Raman
 one which usually is better.
 Filename conventions
 ~~~~~~~~~~~~~~~~~~~~
 Q: What are the conventions for the filenames for the various files that need to be uploaded?
 A: The following definitions apply:
 SCC raw lidar data file
 ^^^^^^^^^^^^^^^^^^^^^^^
 In the current version of the SCC there is not limit in the name of the raw
 data file. We suggest, however, that this file is named <measurement_id>.nc.
 For example, if your measurement had a measurement ID of 20130101cc00 the
 corresponding NetCDF file should be named 20130101cc00.nc
 Sounding file
 ^^^^^^^^^^^^^
 The file should be named as rs_measID.nc. Considering the above example the sounding file should be named rs_20130101cc00.nc
 In this case you should also set the global attribute Sounding_File_Name in the raw lidar data file as::
    Sounding_File_Name=rs_20130101cc00.nc
 Lidar ratio file
 ^^^^^^^^^^^^^^^^
 The file should be named as lr_measID.nc. Considering the above example the sounding file should be named lr_20130101cc00.nc
 In this case you should also set the global attribute LR_File_Name in the raw lidar data file as::
    LR_File_Name=lr_20130101cc00.nc
 Overlap file
 ^^^^^^^^^^^^
 The file should be named as ov_measID.nc. Considering the above example the sounding file should be named ov_20130101cc00.nc
 In this case you should also set the global attribute Overlap_File_Name in the raw lidar data file as::
    Overlap_File_Name=ov_20130101cc00.nc
 Photocounting values should be integers
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 Q: In one of my measurements I get an error concerning the photoncounting values::
    Pre processing (177): Found no integer values in photoncounting signal
 The Raw_Lidar_Data variable in the NetCDF-file is defined as double. So is it necessary for Photoncounting signals to only provide integer values?
 A: Two important considerations:
 . The Raw_Lidar_Data array should contain your *real* raw data.
    This means that *no corrections/operations* should be made on your signals
    before filling the Raw_Lidar_Data array. This is particularly important because *all*
    the operations and corrections should be applied by the SCC and not by the user before
    the submission. In this way we can keep track of all the operations made on the signals
    (for QA purposes) and moreover we are sure that all the corrections are applied in a
    correct order  (this is particularly important for non linear operations, think for example to the dead time correction).
 . The analog signals should be expressed in mV and the photoncounting signals in raw counts
 So if your photoncounting values are not integers they are not expressed in raw counts (which of course should be integers).
 So the point here is not how to convert them in integers but to submit the right quantity in the right units.
 So please check carefully your converter and be sure to really submit raw counts for photoncounting channels and raw mV for the analog ones.
 Preprocessing failed but no Exit code is provided
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 Q: The preprocessing of one of my measurements failed (I get a status -127). However, when I check the Exit codes
 to see the description of the problem, I get an empty value (-). What does this mean?
 A: This means that the preprocessor crushed unexpectedly! Sorry for that! Report the problem in the forum and it will be fixed
 soon.

Mercurial > public > sccdocs / annotate

docs/faq/faq.rst@9391c86dee91 (annotated)

docs/faq/faq.rst