Updates to the BAMS Reanalysis Article

The March 1996 issue of the Bulletin of the American Meteorological Society contains an article and CD-ROM describing the NCEP/NCAR Reanalysis. Normally the authors of a paper quietly hope that published mistakes go unnoticed. Failing that, time eventually heals the embarrassment. Unfortunately we made the grave error of publishing a useful article with a very wide distribution. Since the data on the CD-ROM may be used for years, we have decided to have a web page with all known errors in the Reanalysis article and CD-ROM.

Almost Right

The BAMS article states that "Absolute Vorticity" is available on the pressure levels. Truth is relative, and so is the vorticity on the pressure levels.

In Table 3, replace "virtual temperature" by "temperature" and "specific humidity" by "relative humidity".

On page 467, the Climate Prediction Center (CPC) was formerly known as the Climate Analysis Center (CAC) and not as CDC.

On page 468, "RUNOFF Runoff (kg m^-2) per 6 hour analysis step."

On page 468, SOILW10 is the "Volumetric soil moisture content for the 0 to 10 cm layer (fraction)."

On page 468, SOILW200 is the "Volumetric soil moisture content for the 10 to 200 cm layer (fraction)."

Appendix A. The isentropic data is on a 2.5x2.5 degree grid and on 11 levels.

On page 464, the units of potential vorticity is m*m/s/kg not m*m*kg/s.

The skin temperature should be considered a Class B variable over land and ice-covered points (see Appendix A). Over open water, the skin temperature is obtained from an SST analysis with a correction for surface elevation.


A problem has been discovered with the "DEMO" program on the CD-ROM. The anomaly and climatological fields are plotted incorrectly when the following sequence of menu options are selected.

1. Menu option "m" is selected ("interactive display/plot monthly fields")


2. A pressure level variable is selected (U, V, Z, etc.)


3. Menu option "e" is selected


4. The option of drawing the mean monthly map ("field" option) is set to "no".

If this specific sequence of options are chosen, the anomaly and climatology maps are not correct.

To avoid this problem, EITHER:

a) ALWAYS enable (set to "yes") the "field" option in menu item "e" or

b) Install the new DEMO program which also has some new map projection options. Click here here for the installation page.


The high, middle and low cloud fractions can be found in the monthly means and climatologies. When there are no convective clouds present, these cloud fractions are as described. However, when convective clouds are present, they quantities are a non-linear function of the convective and large-scale clouds. The resultant "cloud fraction" can be as much as 70% less than the cloud fraction as seen by the model. Please do not use HCLDCOV, MCLDCOV and LCLDCOV. The TCLDCOV (total cloud cover) does not suffer from the problems.

Tape Problems: Oct. and Nov. 1985

Some problems with the tape drives corrupted tapes read/written for the months of October and November 1985. For more details, click here .

TOGA COARE November 1992-February 1993

The data over the TOGA COARE region (western equatorial Pacific) during the TOGA COARE observing period (11-92 to 2-93) was affected by a problem with the data processing. The temperatures for the TOGA COARE observations (special dataset from NCAR) were not converted to virtual temperatures as required by the model. As a result, the analyses are too cold over the TOGA COARE region. This cold bias is most pronounced over the surface where the humidities are the highest. This temperature problem affected the upper tropospheric heights through the hydrostatic equation. The four months in question have been rerun. There is talk about reruning the TOGA-COARE period with a higher resolution model.

Observed Rainfall: Part I

The merged precipitation dataset (MRGDRAIN) on the BAMS CD-ROM by Jae Schemm has been modified because of changes in the satellite precipitation estimates. An updated (latest?) version is on the annual Reanalaysis CD-ROMs.

Observed Rainfall: Part II

An error was discovered in the GPCP precipitation data set that appeared on the NCEP/NCAR CD-ROM that was distributed with the March 1996 issue of the Bulletin of the American Meteorological Society.

The error involved the method in which pentad (5-day) accumulations of SSM/I emission precipitation estimates were aggregated into psuedo-calendar months prior to the original authors' (Huffman et. al., 1995) merging of these data with other sources (raingages, IR estimates, etc.). Since pentads do not generally fall on calendar month boundaries, the data must be aggregated to simulate calendar months. Unfortunately, the data from the various sources were not normalized to calendar monthly amounts in the same fashion.

The end result is that time series of global precipitation from the original data set shows erroneous maxima each August that are attributable to the way that the pentad data were aggregated into monthly amounts. The new data set, whose structure is detailed below, remedies this problem.

REFERENCE: Huffman, G. F. (et. al.) 1995, J. Climate, vol. 8, 1284-1295.

Digital Alphas

The BAMS cd-rom is not directly usable on Digital Alpha workstations. The cdrom would also be incompatible with the CRAY except that CRAYs don't have cd-rom drives. The problem is that GrADS uses different index files for Alphas and Cray computers. You need to get your own copy of the control file and remake the index files for the appropriate machine. For example, to read the prfl climatology (assuming cdrom is mounted at /cdrom),

make a new copy of control file in the current directory
cp /cdrom/climate/prflcl.ctl prflcl.ctl
make the control file writable
chmod 644 prflcl.ctl
edit the new control file
vi prflcl.ctl
change the first line from
dset ^prflcl%m2.grb
dset /cdrom/climate/prflcl%m2.grb
make the index files
gribmap -0 -i prflcl.ctl [that's an zero not an "oh"]
The "dset" line is the location of the GRIB files with a caret being GrADS notation for "./" (current directory).

PDP-8, PDP-11, PDP-20, 360, 370, 3090, Apple I, Apple ][, Sinclair, Honeywell, Univac, Ohio Scientific, BBC, Acorn, VAX, Sol, Atari, Sphere, Timex, RCA, Kim, Kaypro, Big Board, Sinclair, Northstar, Sperry, CDC, Transputer, Altos, Heathkit, MITS, Eagle, Sanyo, ENIAC, Z100, Prime, Wang, Apollo, Data General (mini), Xerox, Convex, ELXSI, Thinking Computer, ATT B2, PIC, Tandy CoCo, TRS-80, and IMSAI probably won't work with the BAMS cd-rom.

contact: Wesley.Ebisuzaki@noaa.gov or Wei.Shi@noaa.gov

Privacy Policy