NCDELTA.DAT

Basic Information

File-Type

ncdelta.dat

File-Form

FORMATTED

Version

April 2023

Description-Date

October 2022

Significance of the File

contains general steering data for the program NCDELTA.

File-Contents (in Catchwords)

steering data

• Block Program_Parameters: general steering data
  1. Max_Read_Storage: number of data items in byte read from input file during each analysis cycle. This value gives a good approximation of NCDELTA's mamximum memory requirements for each variable.
  2. Max_Distance: maximum distance in [m] up to which points with different location are compared.
  3. (optional) Print_Modus_Classify: parameter to steer informative printout messages in program section classification:
     1. 0 = no extensive messages printed;
     2. 1 = more extensive messages about classified variables will be printed.
  4. (optional) Print_Modus_Pairs: parameter to steer informative printout messages in program section primary partner variables:
     1. 0 = no extensive messages printed;
     2. 1 = more extensive messages about detected variable pairs will be printed;
     3. 2 = also informations about rejected variable pairs will be printed.
  5. (optional) Print_Modus_Define: parameter to steer informative printout messages in program section definition of computational results:
     1. 0 = no extensive messages printed;
     2. 1 = more extensive informations with respect to metadata of computational results will be printed.
  6. (optional) Print_Modus_Compute: parameter to steer informative printout messages in program section computation of results:
     1. 0 = no extensive messages printed;
     2. 1 = more extensive messages about computational result variables will be printed.
  7. (optional) NC_Deflate_Level: deflate level (0,1,2,...,9) for online compression of data during write. 0 corresponds to no compression and 9 to maximum compression
  8. (optional) NC_Cmode_Id: creation mode flag used for CF NetCDF method NF90_CREATE. Typical values used in this application are:
     • 4 = NF90_64BIT_OFFSET (NetCDF 64Bit file, no HDF - no online compression);
     • 5 = NF90_CLASSIC_MODEL (HDF file without use of extended HDF features - with online compression).
• Block Data_Files: data files
  1. Reference_File: file (NETCDF) with reference data (filetype cf-netcdf.nc);
  2. Variant_File: file (NETCDF) with variant data (filetype cf-netcdf.nc);
  3. Result_File: file (NETCDF) for result data (filetype cf-netcdf.nc);
  4. (optional) Reference_Period: start and end time of period of comparison for reference data (for synoptic data only);
  5. (optional) Variant_Period: start and end time of period of comparison for variant data (for synoptic data only).
• (optional) Block Data_Operations: data operations
  1. (optional) With_Ordinary_Differences: compute (.true.) common differences (if possible) or do not compute (.false.);
  2. (optional) With_Taylor_Diagram_Data: compute (.true.) data for Taylor diagram (if possible) or do not compute (.false.); including skill scores according to Taylor (2001) equations 4 and 5;
  3. (optional) With_Median: compute (.true.) median (>= 32 events) or do not compute (.false.);
  4. (optional) With_Percentiles_05_95: compute (.true.) percentiles Q05, Q95 (>= 32 events) or do not compute (.false.);
  5. (optional) With_Percentiles_01_99: compute (.true.) percentiles Q01, Q99 (>= 51 events) or do not compute (.false.);
  6. (optional) With_Murphy_Skill_1988: compute (.true.) skill according to Murphy (1988) equation 4 or do not compute (.false.);
  7. (optional) With_Skill_Willmott_1981: compute (.true.) skill according to Willmott (1981) Index of agreement (d);
  8. (optional) With_Original_Data: copy (.true.) or not (.false.) input data from which ordinary differences are computed.
• (optional) Block Exclude_Variable_From_NCDELTA: list of variables to be excluded from computation.
  1. Variant_Variable: variable from Variant-File. Key can be prescribed several times. With exception of the variables listed here results will be computed for all primary variable pairs.
• (optional) Block Include_Only_Variable_for_NCDELTA: list of variables for which results ar exclusively computed.
  1. Variant_Variable: variable from Variant-File. Key can be prescribed several times. Results will be solely computed for the variables listed here.
• (optional) Block No_Normal_Vector_Variable: The list of variables which are normal vectors is (mostly) automatically determined with NCDELTA. So far, unfortunately, not all of them can be safely identified based on CF metadata. With this list variables can be explicitely excluded from the class of normal vector variables.
  1. Variable: variable from Variant-File or Reference_File. Key can be prescribed several times.
• (optional) Block Normal_Vector_Variable: The list of variables which are normal vectors is (mostly) automatically determined with NCDELTA. So far, unfortunately, not all of them can be safely identified based on CF metadata. With this list variables can be explicitely included in the class of normal vector variables.
  1. Variable: variable from Variant-File or Reference_File. Key can be prescribed several times.
• (optional) Block Label_Relation: For intercomparison of fractioned variables, for which the relationship between individual fractions (sediment, tidal constituent, etc.) cannot be automatically determined. In such situations the corresponding names must be prescribed by the user. Block can be specified several times.
  1. Variant_Label_KV: name of a valid label coordinate variable from Variant_File or NONE.
  2. Reference_Label_KV: name of a valid label coordinate variable from Reference_File or NONE.
  3. Relation: name of a fraction in Variant_File : name of a fraction in Reference_File. Key can be specified several times. NONE must be specified in case there exists no corresponding label coordinate variable. Colon is used as separator.

General Remarks

1. Dictionary files used from directory $PROGHOME/dic/:
   • ncdelta_dico.dat.
2. Use Cases:
   • Case 1: Reference_File = simulation 1, Variant_File = simulation 2;
   • Case 2: Reference_File = measurement, Variant_File = simulation;
   • Case 3: Reference_File = measurement 1, Variant_File = measurement 2;

   Synoptic data or data analysis results from program NCANALYSE can be used.

3. General requirements:
   • Synoptic data:
     • length of time period must be identical (periods may differ), and
     • time step must be identical.
   • Analysis results (NCANALYSE):
     • length of period of data analysis must be comparable, and
     • number of events, e.g. tidal high water, must also be identical.
   • Z coordinate variables:

   z coordinate variables, if present, must be consistent with respect to

   • CF attribute positive all with value "down", or
   • CF attribute positive all with value "up"

   (see CF metadata conventions).

4. Special requirements:
   • ref. case 1:
     • same vertical structure (number of layers), and
     • identical (tracer-) fractions (if present).
   • ref. case 2:
     • identical (tracer-) fractions (if present), and
     • measured data may contain for the same (geographical) location (x,y) several measured data in different depths (z). Furthermore
     • for comparisons with measured current velocity attribute "name_id" of variable "current_velocity" must be set to 836. This requirement should be checked before execution (e.g. using ncdump). In case this condition is not fulfilled NCO tool NCATTED
       ncatted -a name_id,current_velocity,m,i,836 -O reference_file.nc
       can be used to redefine "name_id" accordingly.
   • ref. case 3:
     • each (geographical) location is allowed to be present only once; locations with several stacked data positions are not allowed currently with this use case.

Programs using this Type of File

NCDELTA

Example-File

$PROGHOME/examples/ncdelta/ncdelta.dat

---

back to: File Descriptions

---

Overview