TOGA COARE was a multidisciplinary, international research effort that investigated the scientific phenomena associated with the interaction ... between the atmosphere and the ocean in the warm pool region of the western Pacific. The field experiment phase of the program took place from 1 November 1992 through 28 February 1993 and involved the deployment of oceanographic ships and buoys, several ship and land based Doppler radars, multiple low and high level aircraft equipped with Doppler radar and other airborne sensors, as well as a variety of surface based instruments for in situ observations.
DATASET INFORMATION
The 4-D Assimilated Dataset is a 4-month, global, gridded atmospheric dataset for use in large-scale atmospheric research. The Laboratory for Atmospheres Data Assimilation Office (DAO) at Goddard Space Flight Center (GSFC) produced this dataset by assimilating rawinsonde reports, satellite retrievals of geopotential thickness, cloud-motion winds, aircraft, ship, and rocketsonde reports with model forecasts employing version 1 of the Goddard Earth Observing System (GEOS-1) atmospheric general circulation model (GCM). At the lower boundary, the assimilating GCM is constrained by the monthly mean observed sea surface temperature and climatological soil moisture derived from monthly mean observed surface air temperature and precipitation fields.
The data are well suited for process studies and intraseasonal variability research since they are produced by a fixed assimilation system designed to minimize spinup in the hydrological cycle. By using a nonvarying system the variability due to algorithm change is eliminated and geophysical variability can be more confidently isolated.
The data archived at the Goddard Distributed Active Archive Center (DAAC) are a subset of the full dataset produced by the DAO. The data are organized chronologically in a time series format to facilitate the computation of statistics spanning long time periods. Data are available for the time period November 1992 through February 1993. Data cover the geographical area 26N-26S and 40E-100W. These data will be online. An expanded subset is available covering the geographical area 50S-50N and 180W-180E.
In calculating variables for this dataset, the DAO makes a distinction between PROGNOSTIC and DIAGNOSTIC variables. A prognostic parameter is an atmospheric variable that the model forecasts. During the assimilation these are the parameters most directly influenced by the observations.
Diagnostic parameters are generally not measured, but are calculated by the model in a manner consistent with the observations. Examples of prognostic parameters include wind, temperature, and humidity. Surface fields include albedo, pressure, wetness, temperature, and ice/water flags. Both upper air and surface prognostic parameters are sampled every 6 hours, though it must be kept in mind that in the current version of the GCM (GEOS-1) some of the surface fields (e.g., albedo, sea surface temperature) are specified from temporally interpolated monthly mean values, while the soil moisture is specified using climatology; the fields are saved more frequently, simply for convenience.
There are also many time averaged diagnostic parameters (e.g., heating rates, precipitation, surface heat fluxes, radiative fluxes at the top and bottom of the atmosphere, and upper air temperature and moisture tendencies) provided either 8 or 4 times daily, depending upon whether the parameter is a surface, single level or upper air quantity.
DATA SOURCE INFORMATION
The assimilated data were a synthesis of observations and short-term model forecasts. Data were collected from globally deployed in situ and remote observations throughout the assimilation period. The platforms used to collect observations were:
Sources 1, 4, 5, and 6 are used in the upper air analyses of height and wind, while the moisture analysis uses only rawinsonde reports. Sources 2 and 3 are used in the determination of sea level pressure and near-surface wind analysis over the oceans. It should be noted that for this DAO TOGA COARE analysis all wind and sea level pressure data collected during COARE (that was received over the GTS) were assimilated. TOGA COARE upper air moisture and temperature data were withheld.
The analysis scheme is carried out at a horizontal resolution of 2 degrees latitude by 2.5 degrees longitude at 14 upper-air pressure levels (20, 30, 50, 70, 100, 150, 200, 250, 300, 400, 500, 700, 850, 1000 mb) and at sea level. Note, however, that the assimilated data are provided at 18 pressure levels with higher vertical resolution in the lower troposphere.
The analysis increments are computed every 6 hours using observations from a +/-3 hour data window centered on the analysis times (00, 06, 12, and 18 UTC). The model computes a 3-hour forecast from an initial reference time. This forecast is compared to its observational counterpart (which was collected during a 6 hour interval centered on the forecast time). An estimate of the forecast error is computed and divided into incremental updates. The model is then re-run from the initial reference time for 6 hours with the incremental updates applied to the simulation. Another three-hour forecast is produced and the cycle is repeated. It is the model integration forced by the incremental updates that makes up the assimilated data provided here. For more detailed information on the assimilation system, please refer to the appropriate references in section 7 of this document.
DATASET ORGANIZATION
There are 22 different files for each month in the assimilation period, for a total of 88 data files over the 4-month run. Associated with every data file is a short ASCII "table" file containing descriptive information to allow data interpretation by the Grid Analysis and Display System (GrADS) visualization package developed at the University of Maryland. The table file includes the name of the associated data file, the x, y, and z coordinate scales, the value used for undefined data, the number of time periods, the date, and a list of the parameters contained in the data file. Since the data files contain only binary science data, all spatial, temporal and parameter information for these files must be inferred from the file name or the contents of the accompanying table file. The file naming conventions are as follows:
Data file : edvl049.prs.xxxxxxx_ss.byymmdd.eyymmdd Table file : edvl049.tabl.xxxxxxx_ss.byymmdd.eyymmdd
where byymmdd, eyymmdd are the beginning and ending year, month, and day, respectively (spanning 1 month, e.g., b921101.e921130), and xxxxxxx denotes the parameters contained in that file. These parameters are drawn from the designations given in the third column of the list below. The first part of the file names, edvl049.prs and edvl049.tabl, remain the same for all files. Each file contains parameters for all days of the specified month.
The parameters are listed below along with their corresponding units and filename designator:
Surface diagnostic fields (group 2) Units File Location ----------------------------------- ----- ------------- SURFACE PRESSURE - PTOP (PTOP= 10mb) (mb) diag2 U-MOMENTUM SURFACE STRESS (N/m^2) diag2 V-MOMENTUM SURFACE STRESS (N/m^2) diag2 SURFACE FLUX OF SENSIBLE HEAT (W/m^2) diag2 SURFACE DRAG COEF. FOR T AND Q - diag2 SURFACE DRAG COEF. FOR U AND V - diag2 SURFACE WIND SPEED (m/s) diag2 FRICTION VELOCITY USTAR (m/s) diag2 SURFACE ROUGHNESS Z0 (m) diag2 PBL DEPTH (mb) diag2
Surface diagnostic fields (group 3) Units File Location ----------------------------------- ----- ------------- SURFACE PRESSURE - PTOP (PTOP= 10mb) (mb) diag3 NET UPWARD LW RADIATION AT GROUND (W/m^2) diag3 NET DOWNWARD SW RADIATION AT GROUND (W/m^2) diag3 OUTGOING LONGWAVE RADIATION (W/m^2) diag3 OUTGOING LONGWAVE RADIATION CLEAR SKY (W/m^2) diag3 SURFACE LONGWAVE FLUX CLEAR SKY (W/m^2) diag3 INCIDENT SW RADIATION AT TOP OF ATMOS (W/m^2) diag3 OUTGOING SHORTWAVE RADIATION (W/m^2) diag3 OUTGOING SHORTWAVE RADIATION CLEAR SKY (W/m^2) diag3 SURFACE SHORTWAVE FLUX CLEAR SKY (W/m^2) diag3 2-DIMENSIONAL TOTAL CLOUD FRACTION (0-1) diag3
Surface diagnostic fields (group 4) Units File Location ----------------------------------- ----- ------------- SURFACE PRESSURE - PTOP (PTOP= 10mb) (mb) diag4 GROUND TEMPERATURE (K) diag4 SURFACE AIR TEMPERATURE (K) diag4 SATURATION SURFACE SPECIFIC HUMIDITY (g/kg) diag4 SURFACE PRESSURE TENDENCY (mb/day) diag4 U AT 2 METERS (m/s) diag4 V AT 2 METERS (m/s) diag4 T AT 2 METERS (K) diag4 Q AT 2 METERS (kg/kg) diag4 U AT 10 METERS (m/s) diag4 V AT 10 METERS (m/s) diag4 T AT 10 METERS (K) diag4 Q AT 10 METERS (kg/kg) diag4
Upper air diagnostic fields (18 levels) Units File Location --------------------------------------- ----- ------------- U-MOMENTUM CHANGES DUE TO TURBULENCE (m/s/day) turbu V-MOMENTUM CHANGES DUE TO TURBULENCE (m/s/day) turbv MOISTURE CHANGES DUE TO TURBULENCE (g/kg/day) turbq TEMPERATURE CHANGES DUE TO TURBULENCE (K/day) turbt TEMPERATURE CHANGES DUE TO MOIST PROCESSES (K/day) moistt MOISTURE CHANGES DUE TO MOIST PROCESSES (g/kg/day) moistq TEMPERATURE CHANGES DUE TO LW RADIATION (K/day) radlw TEMPERATURE CHANGES DUE TO SW RADIATION (K/day) radsw VERTICAL VELOCITY (mb/day) omega FILLING OF NEGATIVE SPECIFIC HUMIDITY (g/kg/day) qfill
Thus, surface diagnostics are contained in 4 files, surface prognostics in 1 file, upper air prognostics in 7 files, and upper air diagnostics in 10 files. Each upper air prognostic file or each upper air diagnostic file will contain a single parameter for 18 pressure levels. These levels are located at 1000, 950, 900, 850, 800, 700, 600, 500, 400, 300, 250, 200, 150, 100, 70, 50, 30, and 20 mb. On the other hand, each surface prognostic file or each surface diagnostic file will contain all the parameters in that category evaluated at the surface or the vertical average through the atmosphere. A summary of the individual file characteristics is shown in the following table:
All prognostic quantities (in files 1-8 above) represent instantaneous values at the designated synoptic time, i.e., at 00Z, 06Z, 12Z or 18Z. The surface diagnostics (in files 9-12), which are reported 8 times daily at 00Z, 03Z, 06Z, etc., are averages over the 3 hour period immediately prior to the designated time tag. For example, a tag of 18Z denotes an average from 15Z-18Z. The upper air diagnostics (in files 13-22) refer to averages over a 6 hour period centered on the designated time tag (i.e., a tag of 18Z denotes an average from 15Z-21Z).
All fields are written in a time-series format, i.e., all global parameter arrays are written out for each daily time period in sequence and then for each day of the month. As an example, for surface prognostics for November 1992, the data and table file names will be:
with the data file having the following organization:
November 1 00Z field 1 ----> field 9 (geopotential height) (vert. int. barotropic vwind)
November 1 06Z field 1 ----> field 9 (geopotential height) (vert. int. barotropic vwind)
November 1 12Z field 1 ----> field 9 (geopotential height) (vert. int. barotropic vwind)
November 1 18Z field 1 ----> field 9 (geopotential height) (vert. int. barotropic vwind)
followed by the same geophysical fields for November 2, November 3,..., November 30.
In the case of the upper air prognostic variable u-wind, the files would be called: edvl049.prs.uwnd_ss.b921101.e921130 edvl049.tabl.uwnd_ss.b921101.e921130
with the data file having the following organization:
November 1 00Z field 1 ----> field 18 (1000mb U-wind) (20 mb U-wind)
November 1 06Z field 1 ----> field 18 (1000mb U-wind) (20mb U-wind)
November 1 12Z field 1 ----> field 18 (1000mb U-wind) (20mb U-wind)
November 1 18Z field 1 ----> field 18 (1000mb U-wind) (20mb U-wind)
followed by the same geophysical parameters for November 2, November 3,..., November 30.
The logical organization of an upper air diagnostic file will be similar to the upper air prognostic file shown above, while the surface diagnostic files (diag1, diag2, diag3, diag4) will be similar to the surface prognostic files except that there will be twice the number of time periods (3 hour increments) and a different number of fields (9 for diag1, 10 for diag2, 11 for diag3, 13 for diag4 (see above table listings)).
For each field in a file, there are 89 grid points in longitude with the first grid point at 40E and with a grid spacing of 2.5 degrees. There are 27 grid points in latitude with the first grid point at 26S and with a grid spacing of 2.0 degrees. The data are stored such that all values along the first latitude (26S) are written first, followed by data from the next northernmost latitude (24S), and continuing on to the last latitude at 26N. For longitude, the first data point is 40E followed by 42.5E and so on to the 89th grid point at 100W. All data are written as IEEE 32-bit floating point numbers.
DATA
The data are IEEE 32-bit floating point, written sequentially. There are no header or trailer records, and the data are distributed uncompressed.
The expanded set is available on 4 mm (DAT), high or low density 8 mm (Exabyte), and 6250 bpi 9-track tapes.
Reading the 4D Assimilation Data Files: Two FORTRAN programs are available from the Goddard DAAC to read the upper air data products. These programs are called the UPPER AIR SAMPLE READ PROGRAM and the SINGLE LEVEL SAMPLE READ PROGRAM.
For each DATA FILE there is an associated TABLE FILE as mentioned above. Some DATA FILES contain a number of different fields at a single level; the TABLE FILE is used to determine how many fields are in the dataset.
>From the TABLE FILE NTIMES is provided by TDEF (number of time steps) NXX is provided by ZDEF for upper air files (vertical levels) NXX is provided by VARS for single level files (parameters) IM is provided by XDEF (longitude) JNP is provided by YDEF (latitude)
The 4-D Assimilated Dataset is compatible with the Grid Analysis and Display System (GrADS) data analysis and visualization software package. This package was developed by Brian Doty with support from Jim Kinter at the Center for Ocean-Land-Atmosphere Studies (COLA) at the University of Maryland. It has gained widespread acceptance by the scientific community as a valuable data analysis tool. The system handles a variety of gridded datasets and observational data. Both the data and table files are required as input in the GrADS package. The software is distributed and supported by the authors via anonymous FTP at COLA.
