Jump to DART Documentation Main Index
The MADIS (Meteorological Assimilation Data Ingest System) service provides access to real-time and archived data of a variety of types, with added Quality Control (QC) and integration of data from a variety of sources.
To convert a series of MADIS data files (where different types of observations are distributed in separate files), one high level view of the workflow is:
There are two satellite wind converter programs; the one in this directory and one in the SSEC directory. The observations distributed here come from NESDIS. The SSEC observations are processed by SSEC itself and will differ from the observations converted here.
The programs in the DART/observations/MADIS/ directory extract data from the distribution files and create DART observation sequence (obs_seq) files. Build them in the work directory by running the ./quickbuild.csh script. In addition to the converters, the advance_time and obs_sequence_tool utilities will be built.
There are currently converters for these data types:
|ACARS aircraft T,U,V,Q data||convert_madis_acars|
|Marine surface data||convert_madis_marine|
|Mesonet surface data||convert_madis_mesonet|
|Wind Profiler data||convert_madis_profiler|
|Satellite Wind data||convert_madis_satwnd|
Example data files are in the data directory. Example scripts for converting batches of these files are in the shell_scripts directory. These are NOT intended to be turnkey scripts; they will certainly need to be customized for your use. There are comments at the top of the scripts saying what options they include, and should be commented enough to indicate where changes will be likely to need to be made.
Several converters have compile-time choices for outputting various types of moist variables. Check the source code for more details. Some converters also read multiple T/F strings from the console (standard input) to control at run-time what types of observations to convert. Again, check the source code for more details.
Each converter has hard-coded input and output filenames:
The expected usage pattern is that a script will copy, rename, or make a symbolic link from the actual input file (which often contains a timestamp in the name) to the fixed input name before conversion, and move the output file to an appropriate filename before the next invocation of the converter. If an existing observation sequence file of the same output name is found when the converter is run again, it will open that file and append the next set of observations to it.
DART software - Copyright 2004 - 2013 UCAR.
This open source software is provided by UCAR, "as is",
|Revision:||$Revision: 9188 $|
|Source:||$URL: https://svn-dares-dart.cgd.ucar.edu/DART/releases/classic/observations/MADIS/MADIS.html $|
|Change Date:||$Date: 2015-12-07 14:43:45 -0700 (Mon, 07 Dec 2015) $|
|Change history:||try "svn log" or "svn diff"|