3DVAR Tutorial

a. Downloading Test Data

(06/09/2003)

In the 3DVAR tutorial, you will be running the 3DVAR observation preprocessor (3DVAR_OBSPROC) as well as 3DVAR itself. This page describes how to access test datasets to run both 3DVAR_OBSPROC and 3DVAR. If you don't know anything about 3DVAR, you should first read the 3DVAR Overview.

1. Required Data

The 3DVAR_OBSPROC code currently inputs observations in “LITTLE_R” format (so-called because it is the format of the previously used LITTLE_R objective analysis package). In addition, 3DVAR_OBSPROC requires an MM5 format file (the first-guess forecast) from which header information describing the domain is retrieved in order to reject observations that are outside the area of interest. The 3DVAR_OBSPROC algorithm performs gross quality control and writes an observation file in “3DVAR” format ready for ingest into 3DVAR. The 3DVAR_OBSPROC system is described in further detail later in the tutorial.

The 3DVAR system requires three input files to run. In addition to the first guess and observation files used by 3DVAR_OBSPROC, a background error statistics file (containing background error covariances currently calculated via the “NMC-method). The use of these three data files in 3DVAR is described later in the tutorial.

The following table summarizes the above info:

Filename

Created By

Used In

ob.little_r

User’s Decoder

3DVAR_OBSPROC

fg

INTERPF or MM5

3DVAR_OBSPROC, 3DVAR

ob

3DVAR_OBSPROC

3DVAR

be

NMC_STATS

3DVAR

In this tutorial, example ob.little_r, fg, and be files are given. In your own work after the tutorial, you will need to create the input data sets yourselves.

2. Downloading Test Data

In this tutorial, we will store data in a directory defined by the environment variable $DAT_DIR and run 3DVAR in a directory defined by the environment variable $RUN_DIR. It is OK to set them equal if you wish. These directories can be in any location that can be read and written to from your current machine and which have at least 200MB of memory available (for this test case, may be more for other cases).  To set these variables, type

setenv DAT_DIR dat_dir

setenv RUN_DIR run_dir

where dat_dir and run_dir are your chosen directories. Create these directories (if they don’t already exist) and type

cd $DAT_DIR

Download the tutorial test data for our standard case from the NCAR ftp site. Once you have downloaded the 3dvar_test_data.tar.gz file to $DAT_DIR, then extract it by typing

gunzip 3dvar_test_data.tar.gz

tar -xvf 3dvar_test_data.tar

You should then see the be, fg, ob.little_r (and maybe a few other) data files in directory $DAT_DIR. If so, then to save space type

rm 3dvar_test_data.tar

3. What next?

OK, now you’ve downloaded the necessary data, you’re ready to do some observation preprocessing. Go to

b. Observation Preprocessing


Trouble Shooting:

1. Having trouble, ask mesouser.


 

Return to 3DVAR Tutorial Page