Objective Analysis for TOGA-COARE IOP
by Baode Chen (1996)
* About the Figure: GMS-4 Infrared Image
(by courtesy of Dr. T. Nakazawa, MRI) and 200 HPa Divergent Wind
of UCLA TOGA-COARE Analysis. 00UTC, Nov. 11, 1992
Using the NCEP/NCAR
reanalysis as the first guess, we have performed objective
analysis (successive correction method) of the final validated,
high resolution (~5 hPa) sounding data (see the Priority
Sounding Stations) for the Intensive Observation Period (IOP)
(November 1992 - February 1993) of TOGA
COARE. The product has two versions: (1) Large-domain analysis;
the domain ranges from 90E to 170W and 30S to 30N on a 2.5 X 2.5
grid mesh, with a vertical resolution of 50 hPa from 1000 hPa
to 50 hPa twice daily at 00 UTC and 12 UTC. (2) Small-domain analysis;
the region is from 140E to 180E and 10S to 10N on a 1 X 1 grid
mesh and the vertical resolution is 25 hPa with total of 39 levels
from 1000 hPa to 50 hPa daily at 00, 06, 12, 18 UTC. Judged with
OLR data and GMS brightness temperature, and compared with the
grid data from the analyses of NCEP/NCAR, JMA(Japan Meteorological
Agency) and ECMWF, which did not include all PSS data in their
analyses, UCLA analysis provides a better representation of the
atmospheric kinematic and thermodynamic characteristics during
the COARE-IOP. However, the moisture field and the moisture budget
residual Q2 appear to have some problems. The differences among
the four different analyses become clearer for the derived quantities
such as horizontal divergence.
- Priority Sounding Station
- IOP-Mean Vertical Distributions of Q1/Cp and Q2/Cp
averaged over the OSA (see the map of PSS stations).
- The number of available observations at 850hPa
used to calculate RMS errors for (a)temperature,
(b)relative humidity, (c)zonal
wind, and (d)meridional wind.
- 850 hPa root mean square error between the
station value and the value interpolated at the station from
UCLA analysis ( -1.0 means that RMS errors are not computed because
of missing data): (a)temperature,
(b)relative humidity, (c)zonal
wind, and (d)meridional wind.
- 850 hPa root mean square error between the
station value and the value interpolated at the station from
the NCEP/NCAR reanalysis ( -1.0 means that RMS errors are not
computed because of missing data): (a)temperature,
(b)relative humidity, (c)zonal
wind, and(d)meridional wind.
- 850 hPa root mean square difference between
NCEP/NCAR reanalysis and UCLA analysis: (a)zonal
wind, (b)meridional wind, (c)temperature,
and(d)relative humidity.
- Monthly-averaged wind: Nov.1992,
Dec.1992,
Jan.1993,Feb.1993.
- Monthly-averaged divergent wind: Nov.1992,
Dec.1992,
Jan.1993,Feb.1993.
- Monthly-averaged OLR: Nov.and
Dec.1992, Jan.and
Feb.1993
- Four-Month IOP Averaged Streamlines for UCLA,
NCEP/NCAR,
JMA,ECMWF.
- 850 hPa time-longtitude section of zonal
wind averaged between 5S and 5N, westerly shaded.
(a)UCLA objective analysis; (b)NCEP/NCAR; (c)JMA; (d)ECMWF .
- 850 hPa time-longtitude section of meridional
wind averaged between 5S and 5N, southerly shaded.
(a)UCLA objective analysis; (b)NCEP/NCAR; (c)JMA; (d)ECMWF .
- Time-height section of zonal wind averaged
over IFA , westerly shaded.
(a)UCLA objective analysis; (b)NCEP/NCAR; (c)JMA; (d)ECMWF .
- Time-height section of meridional wind averaged
over IFA , southerly shaded.
(a)UCLA objective analysis; (b)NCEP/NCAR; (c)JMA; (d)ECMWF .
- 850 hPa time-longtitude section of vorticity
averaged between 5S and 5N, Negative values shaded, interval
1*10E-5.
(a)UCLA objective analysis; (b)NCEP/NCAR; (c)JMA; (d)ECMWF .
- 200 hPa time-longtitude section of vorticity
averaged between 5S and 5N, Negative values shaded, interval
1*10E-5.
(a)UCLA objective analysis; (b)NCEP/NCAR; (c)JMA; (d)ECMWF .
- 850 hPa time-longtitude section of divergence
averaged between 5S and 5N, Negative values shaded, interval
2*10E-6.
(a)UCLA objective analysis; (b)NCEP/NCAR; (c)JMA; (d)ECMWF .
- 200 hPa time-longtitude section of divergence
averaged between 5S and 5N, Negative values shaded, interval
4*10E-6.
(a)UCLA objective analysis; (b)NCEP/NCAR; (c)JMA; (d)ECMWF .
Acknowledgements:
All figures displayed are produced with
the Grid Analysis and Display System (GrADS), which is developed
at the Center for Ocean-Land-Atmosphere Interaction, University
of Maryland.
Please send your comments, suggestions, and
questions to Baode Chen <baode@candor.gsfc.nasa.gov>
