Input

• --input.files Text files with data in columns separated by ";"
  • lat, lon, elev, value (any order is accepted). Note that the order of the columns in the output file is the same as it is in this input file.

in the case of more than one provider of observations, then use one file for each of them. By default, the providers are labeled according to the order of the input files. It is possible to specify the labels the providers will get through the option --prid

Note that in order to undergo the quality control, an observation must have all the metadata. Besides, the elevation must be in a predefined range (--zmin,--zmax).

Customization

It is possible to customize the IO by using optional command line arguments

• --varname.lat name for the latitude variable (in/out)
• --varname.lon name for the longitude variable (in/out)
• --varname.elev name for the elevation variable (in/out)
• --varname.value name for the temperature variable (in/out)
• --varname.opt additional optional variables to be written on the output (out)
• --varname.prid name for the provider identifier (out)
• --varname.dqc name for the data quality control flag (out)

Note the order of the columns in the output file is the same as the order of the columns in the first input file. The columns: prid, dqc are added after the other columns in the output file. In case one of the --varname.opt columns is not available in the input file, then it appears after the other columns in the output file.

The separator used in the input file can be specified with --separator , while the separator to use in the output file can be specified by --separator.out .

The number of decimal digits for several parameters in the output file can be specified by using the command-line parameters:

• --latlon.dig.out for the latitude and longitude.
• --elev.dig.out for the elevation
• --value.dig.out for the observations

Adjust the input data: add offsets and/or rescale

The observations can be adjusted by the user so to match the units of the first-guess fields. This might be useful if relative humidity is observed in the range 0-100% while the first-guess fields are in the range 0-1, for example.

The command line arguments to use are: --input.offset , --input.negoffset (offset sign) , --input.cfact (correction factor) and --input.negcfact (factor sign) . These arguments are numeric vectors as a different value can be specified for each observation provider.

The adjustment is:

adjusted_observation = observation * correction_factor* (-1)^sign(correction_factor) + offset * (-1)^sign(offset)

Abbreviations

• prid. identifier of the observation provider
• lat. latitude
• lon. longitude
• elev. elevation (m amsl)
• value. temperature value (degC)
• dqc. data quality control code