ruc: Rapid Update Cycle Model Data

General Data Description

The Rapid Update Cycle (RUC) is an operational atmospheric prediction system comprised primarily of a numerical forecast model and an analysis system to initialize that model. Mesoscale Analysis and Prediction System (MAPS) is the research counterpart to the RUC. The RUC has been developed to serve users needing short-range weather forecasts. RUC runs operationally at the National Centers for Environmental Prediction (NCEP).

The key features of RUC/MAPS include:

  • high-frequency (every 1h) 3-d objective analyses over the contiguous United States
  • high-frequency (every 1h) short-range weather model forecasts (out to 12 h) in support of aviation and other mesoscale weather forecast users
  • assimilation of data from
    • commercial aircraft (relayed through ACARS - Aircraft Communications, Addressing, and Reporting System)
    • wind profilers (404 and boundary-layer 915 MHz)
    • rawinsondes and special dropwinsondes
    • surface reporting stations and buoys
    • RASS (Radio Acoustic Sounding System) - experimental
    • VAD (velocity-azimuth display) winds from NWS WSR-88D radars
    • GOES total precipitable water estimates
    • SSM/I total precipitable water estimates
    • GPS total precipitable water estimates
    • GOES high-density visible and IR cloud drift winds

The RUC uses a hybrid isentropic-sigma vertical coordinate in which most of the atmosphere is resolved on isentropic surfaces (defined by constant virtual potential temperature) except for layers near the ground where terrain-following (sigma) coordinates are used.

Over the period of time that ARM has been collecting RUC data, the resolution has improved from 60 to 40 to 20 to 13 kilometeters

ARM archives only the hourly analysis data not the forecasts.

RUC data were discontinued on May 1st, 2012. Rapid Refresh (RAP) data have replaced the RUC data as of 05/01/2012. Please see the information on the change in the section of Data User Notes below for the details.

Measurement Description

RUC Measurements

Temporal Coverage

Data Stream NameStart Date End Date
sgpfslruc60X1.c0 10/24/1994 11/16/1994
sgpallruc60X1.c0 05/08/1996 04/23/1998
sgpruc60X1.c1 05/08/1996 04/23/1998
sgpallruc40hybrX1.c1 12/04/1998 4/16/2002 ***
sgpallruc40isobX1.c1 04/20/1998 4/16/2002***
sgpallruc20hybrX1.c1& 4/18/2002 10/16/2002**
sgpallruc20isobX1.c1& 4/18/2002 10/16/2002**
pyeruc20isobX1.c1 & 3/1/2005 04/02/2006*
sgpruc20hybrX1.c1 & 4/18/2002 05/01/2012*
sgpruc20isobX1.c1 & 4/18/2002 05/01/2012*
sgpsynruc20isobX1.c1 & 4/18/2002 05/01/2012*
RUC GRIB Input files from FSL
sgpallruc40hybrX1.00 12/04/1998 4/16/2002 ***
sgpallruc40isobX1.00 04/20/1998 4/16/2002 ***
sgpallruc20hybrX1.00 4/18/2002 05/01/2012
sgpallruc20isobX1.00 & 4/18/2002 05/01/2012
sgpfnnallruc20isobX1.00 & 4/18/200205/01/2012
sgpallruc13hybrX1.00 02/21/2008 05/01/2012
sgpallruc13isobX1.00 & 02/01/2008 05/01/2012

*Current RUC 20 data are available at the ARM Archive approximately 4 days after they are generated.

& Beginning on 6/28/2005, the RUC20 GRIB and resulting cdf files are "look-alike" files generated from the RUC13 files.

** Although the sgpallruc20* netCDF files are no longer at the ARM Archive, they are still generated at the XDC and are kept for one month after generation for special needs. Contact Richard Wagener if these files are needed.

*** For RUC40 data from 19991209 - 20020416, hour 23Z is usually missing. The reason for these gaps is unknown.

Area Covered

The RUC covers the lower 48 United States and adjacent areas of Canada, Mexico, and oceanic areas. The grid is a subset of the AWIPS Lambert Conformal Grid (AWIPS/GRIB grid 215).
The 60-km grid has 81 by 62 points.
The 40-km grid (AWIPS grid ID 236) has 151 by 113 points.
The 20-km grid (AWIPS grid ID 252) is 301 by 225.
The 13-km grid (AWIPS grid ID 130) is 451 by 337.

RUC-20 and RUC-13 have 50 computational levels compared to 40 levels for the RUC-40.

The files named sgpallruc20* cover the entire RUC domain.
The files named pyeruc20isobX1.c1 cover latitudes 34.300 to 42.539 and longitudes -127.607 to -120.862.(see map)
The files named sgpruc20* are "cropped" to cover latitudes from 34.21 to 38.93 and longitudes from -100 to -95. (see map)
The files named sgpsynruc20isobX1.c1 have reduced resolution (grid cells are averaged to 40 km resolution) and cover latitudes from 25 to 45 and longitudes from -115 to -85. (see map)

Data Stream Inputs










Shaocheng Xie
Lawrence Livermore National Laboratory

Richard Wagener
Brookhaven National Laboratory

Laurie Gregory
Brookhaven National Laboratory

Data Source

Forecast Systems Laboratory (FSL) and
National Centers for Environmental Prediction (NCEP)

Reprocessing History

Reprocessing Date

In February, 2003 data was reprocessed spanning dates 19980421 - 19980930.

Start End Reprocessing

19980421 - 19980930

Reason Reprocessing

19980421 - 19980930 reprocessing:
Reprocessing completed: Data from 19980421 - 19980930 were reprocessed and corrected. Corrected data were sent to the Archive on 2003-02-07.
Reason data was reprocessed:The sgpallruc40isobX1.c1 data from 1998-04-21 to 1998-09-30 originally archived was incorrect. The data contained all missing values (-9999). The error was due to unexpected GRIB header information in the GRIB data files. From 1998-04-21 to 1999-02-16, the grid number was given as 255 (undefined) when the ingest expected it to be 236. Starting 1999-02-16, cycle 13Z, FSL changed the grid number to 236. Since the 1998 GRIB data are correct and there was a simple work-around to the ingest, the data were reprocessed to correct the output data.


Q: Can I get the forecast data?
A: The forecast data are only collected when the analysis data are unavailable for some reason. This is a very rare occurrence and for all intents and purposes the forecast data should be considered unavailable.
Q: What is the difference between the sgpallruc20isobX1.00 and sgpallruc20hybrX1.00 datastreams?
A: The sgpallruc20isobX1.00 are isobaric files. The sgpallruc20hybrX1.00 are hybrid. The hybrid data uses a combination of two types of vertical coordinates - isentropic and sigma. Hybrid is short for "hybrid-b isentropic-sigma coordinate".

Additional information is available at:
Technical Procedures Bulletin :
Q: I need RUC data outside of the SGP CART domain. How can I get it?
A: The full RUC domain is archived only in GRIB format and is available by emailing the desired datastream name and time period to the archive
For research relevant to ARM and for large requests or requests requiring data in netCDF format, please fill out a "Special Data Set Request".
Q: How can I create subsetted RUC 40 data similar to the RUC 20 data currently available at the Archive?
A:Currently, the RUC 40 netCDF data available at the archive are the full area, hourly files (sgpallruc40isobX1.c1 and sgpallruc40hybrX1.c1). Using the netCDF Operator (NCO) utilities, it is easy to produce subsetted, daily files similar to the RUC 20 subsets available (sgpruc20hybrX1.c1 and sgpruc20isobX1.c1).
The NCO utilities can be found at
Using the ncks (NCO "kitchen sink") command, you can subset the ruc data as follows:
ncks -d nx,, -d ny,,  
Note: xmin,xmax,ymin,ymax are index ranges.
For example to subset around the SGP site use the following index values for latitudes 34.209 to 38.928 and longitudes -99.798 to -95. xmin = 71 xmax = 82 ymin = 40 ymax = 53

ncks -d nx,71,82 -d ny,40,53 sgpallruc40isobX1.c1.20000314.130000.cdf sgpruc40isobX1.c1.subset.cdf

Here is a sample C-Shell script for subsetting the ruc40 and using the NCO command ncrcat to concatenate the hourly subsetted files into a ruc 40 subsetted daily file similar to the sgpruc20isobX1.c1 files.

# This script subsets the RUC 40 isobaric files
# and then concatenates hourly files into 1 daily file
# usage: subset_ruc40.csh 
#   where yyyymmdd is the date of the files to subset
#   and concatenate. Ex. 20000314 (yyyy=2000,mm=03, dd=14)
# Example: subset_ruc40.csh 20000314
# xmin,xmax,ymin,ymax are indices for
# lat = 34.209, 38.928, lon = -99.798, -95
set xmin = 71
set xmax = 82
set ymin = 40
set ymax = 53
set date = $1
foreach file (sgpallruc40isob*$date*)
  ncks -d nx,${xmin},${xmax} -d ny,${ymin},${ymax} $file $file.sub
ncrcat -h sgpallruc40isob*$date*sub sgpruc40isobX1.c1.$date.000000.cdf
rm *.sub

For more information on the NCO commands ncks and ncrcat see:

Also other commands for subsetting and analysis of netCDF data can be found at

Q: Why does the preciprate (or PRATE) parameter contain all zeros or missing values?
A: The following two answers are from Stan Benjamin from the ESRL/GSD RUC forum:

Question 1: from

Jim submitted the following question on 2004-01-28:
The F00 (analysis) set of GRIB products for the RUC (at NCEP and FSL) contain precip grids which are valid over some interesting time periods. One is an accumulation over two times which are the same (an instantaneous accumulation?) and the other looks like it is the one-hr accumulation ending at the current time, as it has t0 = 23 and t1 = 0. For the instantaneous accumulation, the large-scale precip field values are all 0, and the convective precip fields are all -99999. For what looks like the 1-hr accumulation (t0=23, t1=0) the values are 99999 for both precip types. Is there any chance that real values can be sent for these grids, especially the ones that look like the 1-hr accumulation anding at the analysis time? Thanks!

Answer: Stan answered the question on 2004-01-29:
Jim and everyone, The 0h RUC precip fields are basically carried over from the previous 1h forecast used as a background for the current analysis. They should not be regarded as a precipitation analysis. There is some information on convection based on GOES passed through the 1h convective precip to help trigger the RUC convective parameterization the next hour, but again, it shouldn't be used as a precip analysis. Good question.
Stan Benjamin

Question 2: from

Posted by Mike on 29 Aug 02, 8:19 MT:
As I am not a meteorologist I was hoping some one could explain the PRATE parameter? I am looking for precipation information that will enable me to display up to six levels of weather. This is the weather that is commonly seen on a weather radar. If PRATE can't do this is there a parameter or combination of parameters in RUC that can?

Answer: Posted by Stan Benjamin on 9 Sep 02, 12:04 MT, in response to PRATE explanation, posted by Mike on 29 Aug 02, 8:19 MT:
The PRATE field from the RUC is an instantaneous precipitation rate. The 1-h precipitation would be another candidate field for your application. It is possible to convert 3-d hydrometeor mixing ratios for rain, snow, and graupel into estimated reflectivities, but that probably goes beyond what you need.
I hope this helps.

Data User Notes

netCDF format files can be ordered directly through the regular ARM channels, e.g. the order data "Add to Cart" button on this page.

If you have determined you need to order the GRIB format data, send e-mail to including datastream names and time period that you require.

Note that on 2/1/2008, the format for the GRIB files changed from GRIB1 to GRIB2. Information on reading GRIB1 and GRIB2 formatted files can be found at: Reading netCDF, HDF, and GRIB Files

As of 2012-05-01, sgpallruc20isobX1.00 and sgpallruc20hybrX1.00 were discontinued. The next-generation model called the Rapid Refresh (RAP) data have replaced the RUC data. The ARM External Data Center (XDC) began collecting the following hourly RAP output products in grib2 format starting on 2012-05-01:

sgprap13plevX1.00: Grid 130 at 13-km resolution covering the CONUS domain, pressure level data. This is a replacement of sgpallruc20isobX1.00.

sgprap13nlevX1.00: Grid 130 at 13-km resolution covering the CONUS domain, model native level data. This is a replacement of sgpallruc20hybrX1.00.

In addition, XDC also collects the following hourly RAP data:

nacrap32X1.00: Grid 221 at 32-km resolution covering the full AWIP32 domain, pressure and model native level data.

nsarap11X1.00: Grid 242 at 11-km resolution covering Alaska, pressure and model native level data.

All four RAP datastreams (sgprap13plevX1.00, sgprap13nlevX1.00, nacrap32X1.00, nsarap11X1.00) are currently only available in their native GRIB2 format by special email request to

The RAP data differs from the RUC data in resolution, domain size, grid type and variables. More detailed information on the RAP data can be found at

As of 2012-06-07, the ARM XDC is in the process of updating the ingest process and will start producing daily netCDF files for SGP and NSA in a format similar to the current RUC files in the near future."


FSL Forecast Systems Laboratory
GRIBGRIdded Binary
MAPS Mesoscale Analysis and Prediction System
NCEPNational Centers for Environmental Prediction
PYEPoint Reyes
RUCRapid Update Cycle
SGPSouthern Great Plains