The program functions of LITTLE_R is similar to but not as complete as those in RAWINS. The major advantage of LITTLE_R is that it makes observation input easier to deal with for users who don't have access to NCAR's archived ADP data. Like RAWINS, this program reads in output from program REGRID (REGRID_DOMAINx), and observations (radiosonde and surface reports), performs an objective analysis which blends first guess with observations, and outputs data on pressure levels again (just like it is from REGRID). You should not be here if you have not successfully run the REGRID program.

To learn more about how the LITTLE_R program works, please read it on MM5 page and Appendix C of the Tutorial Notes.

Where to Start?

First, go back to directory MM5V3, and download the LITTLE_R program tar file from NCAR's ftp site. Once you have it on your machine, type the following to unzip and untar it:

gunzip LITTLE_R.TAR.gz
tar -xvf LITTLE_R.TAR

After this, you should see a program directory LITTLE_R/ created in your working directory. cd to this directory:

cd LITTLE_R

and you should see the following files and subdirectories listed:

ls -l

-rwxr-xr-x   1 mesouser users       2941 Dec 19 14:35 CHANGES
drwxr-xr-x   2 mesouser users       8192 Feb 25 15:03 Diff
drwxr-xr-x   2 mesouser users       8192 Feb 25 15:03 Doc
-rwxr-xr--   1 mesouser users       9157 Dec 14 15:03 Makefile
-rwxr-xr--   1 mesouser users       3078 Dec  4 12:19 README
-rwxr-xr--   1 mesouser users        419 Jun  7  2000 configure.rules
-rwxr-xr--   1 mesouser users         92 Jun  7  2000 macros_little_r
-rwxr-xr--   1 mesouser users       2987 Dec 19 14:35 namelist.input
drwxr-xr-x   2 mesouser users       8192 Feb 25 15:03 src
drwxr-xr-x   2 mesouser users       8192 Feb 25 15:03 util

Read the README file in the directory.

Now download the test data input2little_r.tar from the ftp site. After you untar it, you should see a new directory Test_data/:

ls -l Test_data

-rw-r--r--   1 mesouser users    7815812 Feb 14  2000 REGRID_DOMAIN1.gz
-rw-r--r--   1 mesouser users     145692 Feb  2  1999 obs-13_00.gz
-rw-r--r--   1 mesouser users      49773 Feb  2  1999 obs-13_06.gz
-rw-r--r--   1 mesouser users     153401 Feb  2  1999 obs-13_12.gz
-rw-r--r--   1 mesouser users      62591 Feb  2  1999 obs-13_18.gz
-rw-r--r--   1 mesouser users     141449 Feb  2  1999 obs-14_00.gz
-rw-r--r--   1 mesouser users      64216 Dec  6 15:16 surface_obs_r:1993-03-13_00.gz
-rw-r--r--   1 mesouser users      23726 Dec  6 15:16 surface_obs_r:1993-03-13_03.gz
-rw-r--r--   1 mesouser users      53761 Dec  6 15:16 surface_obs_r:1993-03-13_06.gz
-rw-r--r--   1 mesouser users      19994 Dec  6 15:16 surface_obs_r:1993-03-13_09.gz
-rw-r--r--   1 mesouser users      63674 Dec  6 15:16 surface_obs_r:1993-03-13_12.gz
-rw-r--r--   1 mesouser users      25532 Dec  6 15:16 surface_obs_r:1993-03-13_15.gz
-rw-r--r--   1 mesouser users      63674 Dec  6 15:16 surface_obs_r:1993-03-13_18.gz
-rw-r--r--   1 mesouser users      18258 Dec  6 15:16 surface_obs_r:1993-03-13_21.gz
-rw-r--r--   1 mesouser users      64219 Dec  6 15:16 surface_obs_r:1993-03-14_00.gz

cd to Test_data directory, and gunzip all files by typing gunzip *, and you'll have

-rw-r--r--   1 mesouser users   10061640 Feb 14 13:55 REGRID_DOMAIN1
-rw-r--r--   1 mesouser users    3119729 Feb  2  1999 obs-13_00
-rw-r--r--   1 mesouser users    1136846 Feb  2  1999 obs-13_06
-rw-r--r--   1 mesouser users    3250114 Feb  2  1999 obs-13_12
-rw-r--r--   1 mesouser users    1426484 Feb  2  1999 obs-13_18
-rw-r--r--   1 mesouser users    3025161 Feb  2  1999 obs-14_00
-rw-r--r--   1 mesouser users      64216 Dec  6 15:16 surface_obs_r:1993-03-13_00
-rw-r--r--   1 mesouser users      23726 Dec  6 15:16 surface_obs_r:1993-03-13_03
-rw-r--r--   1 mesouser users      53761 Dec  6 15:16 surface_obs_r:1993-03-13_06
-rw-r--r--   1 mesouser users      19994 Dec  6 15:16 surface_obs_r:1993-03-13_09
-rw-r--r--   1 mesouser users      63674 Dec  6 15:16 surface_obs_r:1993-03-13_12
-rw-r--r--   1 mesouser users      25532 Dec  6 15:16 surface_obs_r:1993-03-13_15
-rw-r--r--   1 mesouser users      63674 Dec  6 15:16 surface_obs_r:1993-03-13_18
-rw-r--r--   1 mesouser users      18258 Dec  6 15:16 surface_obs_r:1993-03-13_21
-rw-r--r--   1 mesouser users      64219 Dec  6 15:16 surface_obs_r:1993-03-14_00

Here REGRID_DOMAIN1 is the output file from REGRID. You can use this one, or the one you created if you have run REGRID successfully. Files with prefix 'obs' and a time stamp are pre-processed observations that contains both upperair and surface observations for that time (the source of the data is NCAR's ADP). Files with prefix 'surface' and a time stamp are pre-processed SURFACE observations for that time. These surface observations are the same are the surface observations that are included in the 'obs' files. They are exclusively used for the new FDDA feature build into LITTLE_R v3.5. If you do not plan on using FDDA (variable "f4d", record7 of namelist), you do not need these files.
 
 

Now you are ready to work with LITTLE_R.

How to Run LITTLE_R?

1. Compile LITTLE_R:

LITTLE_R is a Fortran 90 program, and its memory allocation is done at the run time stage. Like program REGRID, you only need to compile it once unless you modify the code. Similar to what you do in program REGRID, type

make >& make.out &

This compiles the code on a your machine, and writes output from "make" to file make.out. If it is successful, you should create an executable called little_r in the directory.  If you list the files you will see that little_r is a symbolic link into the src directory.

ls -l little_r

    ./little_r -> src/little_r

Since v3.5, two plotting routines with which you can view the observational data, have been added (plot_level  &  plot_soundings).  These routines make use of NCAR Graphics to build the plots.  If you do not have NCAR Graphics, set  "PROGS = $(I_DONT_HAVE_NCARG)",  in the Makefile,  and change the line:

        "LOCAL_LIBRARIES   =       -L$(NCARG_ROOT)/lib -lncarg -lncarg_gks -lncarg_c -lX11 -lm"

TO:

        "LOCAL_LIBRARIES   =       "
 

If you have NCAR Graphics on you machine, and the plotting routines compiled successfully, you will have 2 more executables in the LITTLE_R directory:
    ./plot_level -> src/plot_level
    ./plot_soundings -> src/plot_soundings
 

2. Edit the namelist.input file. It is already set up to run the SOC case, except for the input file locations for REGRID output and observations. Other highlighted namelist variables are the basic ones you need to modify when you work on a different case. For a complete description of the namelist variables, please see section C.2.1 in Appendix C of the Tutorial Notes. Once you've changed it, type the following to run:

little_r >& log &

If you are successful, you should find these additional files generated in the directory:

-rw-r--r-- 1 mesouser users    3854912 Feb 25 16:09 LITTLE_R_DOMAIN1
-rw-r--r-- 1 mesouser users     273233 Feb 25 16:02 discard_out_1993-03-13_00:00:00.0000
-rw-r--r-- 1 mesouser users      11285 Feb 25 16:10 log
-rw-r--r-- 1 mesouser users    1220793 Feb 25 16:04 obs_used_for_oa_out_1993-03-13_00:00:00.0000
-rw-r--r-- 1 mesouser users    1246804 Feb 25 16:04 plotobs_out_1993-03-13_00:00:00.0000
-rw-r--r-- 1 mesouser users    2009671 Feb 25 16:03 qc_out_1993-03-13_00:00:00.0000
-rw-r--r-- 1 mesouser users    2413247 Feb 25 16:02 result_out_1993-03-13_00:00:00.0000
-rw-r--r-- 1 mesouser users    2045829 Feb 25 16:02 useful_out_1993-03-13_00:00:00.0000

where log is a log file of all print statements generated by LITTLE_R, and this is the file you should take a closer look if a LITTLE_R job fails. Near the end of the log file, you should see the following line if the job is finished correctly:

STOP 99999

LITTLE_R_DOMAIN1 is the binary output file from LITTLE_R, and will be the input file for INTERPF. It contains 3-D pressure level analyses of winds, temperature, RH, geopotential height, and sea level pressure, and it is similar to the binary output file from RAWINS.  The other files are ASCII text files which detail the actions taken on observations through a time cycle of the program.
 

If you choose to do FDDA, you should find these additional files generated in the directory:

-rw-r--r-- 1 mesouser users     543968 Feb 25 16:10 SFCFDDA_DOMAIN1
-rw-r--r-- 1 mesouser users      91149 Feb 25 16:04 discard_out_sfc_fdda_1993-03-13_00:00:00.0000
-rw-r--r-- 1 mesouser users     347333 Feb 25 16:04 obs_used_for_oa_out_sfc_fdda_1993-03-13_00:00:00.0000
-rw-r--r-- 1 mesouser users     354678 Feb 25 16:04 plotobs_out_sfc_fdda_1993-03-13_00:00:00.0000
-rw-r--r-- 1 mesouser users     833325 Feb 25 16:04 qc_out_sfc_fdda_1993-03-13_00:00:00.0000
-rw-r--r-- 1 mesouser users    1030179 Feb 25 16:04 result_out_sfc_fdda_1993-03-13_00:00:00.0000
-rw-r--r-- 1 mesouser users     851775 Feb 25 16:04 useful_out_sfc_fdda_1993-03-13_00:00:00.0000

The log file in this case will include statements to the generation of the SFCFDDA_DOMAIN1 file.  SFCFDDA_DOMAIN1 contains surface analyses and a count of observations within 250 km of each grid point.  The ASCII files in this case is similar to the case files generated for the run without FDDA.
 

Since version 3.5, LITTLE_R provides two utility programs for plotting observations.  These programs are called "plot_soundings" and "plot_level".  These optional programs use NCAR Graphics, and are built automatically if the PROGS option in the top-level Makefile is set to $(I_HAVE_NCARG).  Both programs prompt the user for additional information.

    plot_soundings : This program generates soundings from either the quality-controlled ("qc_out_yyyy-mm-dd_hh:mm:ss:ffff") or the non-quality-controlled ("useful_out_yyyy-mm-dd_hh:mm:ss:ffff") upper-air data.  Only the data that are on the requested analysis levels are processed.  The program asks the user for an input filename, and creates the file "sounding.cgm".

    plot_level : This program creates station plots for each analysis level. These plots contain both observations that have passed all QC tests and observations that have failed the QC tests. Observations that failed the QC tests are plotted in various colors according to which test was failed.  The program prompts the user for a date in the form yyyymmddhh, and creates the file "levels.cgm".
 

To check whether your analysis is OK, you can run program GRAPH/RIP if you have NCAR Graphics.

Miscellanies:

1. If you change other print namelist variables to be true, you'll get various output files.

2. If you have an NCAR computing account, you can access ADP-format observations readily. You can use the fetch.deck in the LITTLE_R/util directory on NCAR's Cray (or fetch.deck.ibm, on NCAR's IBM's) for this purpose.  You may run this fetch job interactively or in batch mode on NCAR's computer. You may then ftp or rcp the data back to your local computer to run LITTLE_R.

3. A "deck" is provided for running on NCAR's IBM computers. You will find little_r.deck.ibm.batch under the LITTLE_R/util directory.  This deck can be run interactively or in batch mode.  This deck combines the "fetch.deck" (see 2 above) with compiling and running of little_r.  You need to have an account on one of NCAR's IBM computer to use this deck.

4. If you need to convert data you have access to, to the format accepted by LITTLE_R, please read the complete LITTLE_R document in the Tutorial Notes, and refer to the utility program upa.f residing in the LITTLER/util directory.

5. For a brief  introduction to the LITTLE_R program, you can also view the bi-annual tutorial slides.

Trouble Shooting:

1. The little_r program produce extensive printout if the user sets the print switches in the namelist to TRUE.  Most of the time, this is enough to capture the problem.  The first place to look is towards the end of the log file.

2. To find how many observation reports are available to the program, search for "Number" in the log file:
       Number of observations successfully ingested:           1191.
    Number of empty observations discarded:                    4.
    Number of observations discarded outside of domain:      534.

3. The little_r program must have the namelist.input file in the working directory, the gridded input file must exist, the 3d obs files must exist (though
they may be empty).  The FDDA obs need to exist if the f4d flag is TRUE (again, though they may be empty).  The requested dates in the namelist file must exist in the first-guess data set.

4. Frequent troubles arise from trying to write observations into the format required by the little_r program.  If you are writing data in this format, refer to the utility program upa.f residing in the LITTLER/util directory.

5. A large disparity in the data distribution for the observations (such as would happen if half of the domain has lots more observations than the other half of the domain) increases the change for a poor analysis if the multiquadratic scheme is chosen.  The multiquadratic scheme is two-dimensional, but the effective radius of influence is the entire analysis domain.

If you have run LITTLE_R successfully, move to the next program: INTERPF.

If you have questions, ask mesouser.