October 1822, 2004

1
Impact of Precipitation Observations onRegional
Climate Simulations
Ana Nunes, John Roads, Masao Kanamitsu Scripps
Experimental Climate Prediction Center (ECPC)La
Jolla, CA and Phil Arkin Earth System Science
Interdisciplinary Center, University of
Maryland, College Park, MD anunes_at_ucsd.edu
2
Summary
Currently available global reanalyses (NCEP/NCAR
Reanalysis, NCEP/DOE Reanalysis, ERA-15, ERA-40
and others) provide reasonably accurate analysis
of large-scale atmospheric states, the weakest
component of those reanalyses is the
model-produced precipitation, which has very
large errors compared to observations. For this
reason, to develop downscaled analysis suitable
for regional forecast initial conditions and for
consistent energy budget research became a
nowadays topic. In this study, we use a
regional climate model to assimilate different
precipitation data sets (a) the .25 deg.
National Oceanic and Atmospheric Administration's
Climate Prediction Center (NOAA/CPC) daily
precipitation analyses (b) and the new .25 deg
NOAA/CPC MORPHed precipitation (CMORPH). To
study the sensitivity of the precipitation
assimilation method to these data sets, we chose
a large domain, which includes North and Central
America. To evaluate the performance of the
regional spectral model results, we compared them
to the North America Regional Reanalysis (NARR)
fields.
3
Model
The Scripps ECPC RSM, described previously by
Juang and Kanamitsu (1994) Anderson et al.
(2001) and Roads (2003), used for these
experiments had 50- and 60-km resolutions and 28
vertical levels. A Mercator projection was used
for the projection of the regional grid. The RSM
is a primitive equation model, with similar
physics as the driving NCEP-DOE reanalysis II
(R-2) Global Spectral Model as reported in
Kanamitsu et al. (2002). This study employed
Simplified and Relaxed Arakawa-Schubert cumulus
convection schemes (SAS and RAS). .
4
Data Sets

• Base and boundary conditions
• RSM initial and boundary conditions were
  obtained from the coarser scale R-2 reanalysis
  (1.875 resolution) and 28 vertical levels.
• SST (1 degree resolution) was taken from
  the Project to Intercompare Regional Climate
  Simulations (PIRCS) data set.
• (b) Precipitation data sets
• Daily rain rates were provided by the CPC
  precipitation analysis (see Higgins et al., 2000)
  over the U. S. domain. R-2 precipitation fields
  were used for the rest of the model domain,
  including Mexico.
• The 3-hourly and daily CMORPH
  precipitation analysis was provided on a regular
  grid of 0.25º. The CPC morphing (CMORPH)
  technique (Joyce et al, 2004) combines the low
  earth orbiting satellite passive microwave sensor
  (PMW) retrievals and the infrared channel of the
  geostationary satellite, which is used to
  spatially and temporally transport the rainfall
  features.

5
Physical Initialization (PI)
This scheme basically adjusts the humidity
profile using the difference between the
observed and predicted rain rates as factor of
this adjustment. In order to provide consistent
temperature profiles, the cumulus and large-scale
parameterizations are then requested. This
methodology differs from the used by the FSU
Nested Regional Spectral Model (Nunes and Cocke,
2003), where a modified Kuo parameterization is
the convection scheme, however the general PI
procedure follows the same structure as shown in
Fig. 1.
OBSERVED RAIN RATES TIME STEP ASSIMILATED
PI-ANALYSIS
PHYSICAL INITIALIZATION SCHEME
FORECAST
DAY -1 ANALYSIS
DAY 0 ANALYSIS
Fig. 1 - General overview of the PI procedure
considering a continuous data assimilation
system.
6
RSM 50- and 60-km Experiments
(1) The North and Central America experiment
using 60-km resolution started at July 1st, 1986
at 0 UTC, where RAS was the cumulus convection
scheme. July-August-September (JAS) 1988 will be
shown. The CPC daily rain rates were used by the
assimilation technique. (2) The North America
experiment was performed with 50-km model
resolution, starting at May 1st, 2003 at 0 UTC,
using SAS. June-July-August (JJA) will be shown.
The 3-hourly as well as daily CMORPHED
precipitation analyses were used. The Control
simulations do not assimilate precipitation. In
the PI simulations, the rain rates were updated
every 24-h (1 and 2) and 3-h (2), and the
moisture adjustment took place every time-step,
which was 2 min. The boundary conditions were
updated every 6 hours.
7
RSM 60-km (RAS) JAS 1988Precipitation (mm/d)
HigginsR-2
Control
PI
Area 1
Area 2
8
RSM-60km (RAS) x HigginsR2JAS 1988
9
RSM 60-km (RAS)Equitable Threat Score (ETS)
10
RSM 60-km (RAS)BIAS
11
NCEP North American Regional Reanalysis (NARR)
NARR is based on the Eta 32-km/45-layer
resolution (see Mesinger et al, 2002). NARR
assimilates observational data sets, which
include temperature, wind, and moisture. However,
the major component of the NARR is the
assimilation of precipitation. The
precipitation data set used by NARR comes from
different sources, including the CPC Merged
Analysis of Precipitation (CMAP), a merged
combination of satellite and gauge precipitation.
http//wwwt.emc.ncep.noaa.gov/mmb/rreanl
12
NCEP North American Regional Reanalysis (NARR)
The plot is courtesy of Matt Pyle of EMC.
http//wwwt.emc.ncep.noaa.gov/mmb/rreanl/eta_rean_
3245.gif
13
RSM 60-km JAS 1988Specific Humidity (g/kg)
PI
Control
NARR
14
RSM 60-km JAS 1988Temperature (K)
Control
NARR
PI
15
RSM 60-km JAS 1988Horizontal wind (m/s)
Control
PI
NARR
16
(No Transcript)
17
JAS 1988 Precipitation (mm/d)
Control
PI
HigginsR-2
NARR
18
(No Transcript)
19
RSM 50-km (SAS) JJA 2003Specific Humidity (g/kg)
3-h PI
Control
NARR
24-h PI
925-hPa
300-hPa
20
RSM 50-km (SAS) JJA 2003Temperature (K)
NARR
3-h PI
24-h PI
Control
925-hPa
300-hPa
21
RSM 50-km (SAS) JJA 2003Horizontal Wind (m/s)
NARR
3-h PI
24-h PI
Control
925-hPa
300-hPa
22
(No Transcript)
23
50-km JJA 2003Precipitation (mm/d)
Daily PI RSM
3-hourly PI RSM
Control RSM
3-h CMORPH
24-h CMORPH
NARR
24
(No Transcript)
25
Concluding Remarks
Precipitation assimilation has been used by the
ECPC-RSM to improve short- and long-term regional
precipitation simulations as well as simulations
of prognostic variables, and preliminary results
using different sets of precipitation data
produced model precipitation fields quite similar
to the assimilated precipitation analyses,
especially during warmer seasons, which was
reported by Mesinger et al. (2003) about the NARR
simulations as well. The ECPC merged
precipitation analysis (CPC daily R-2)
assimilations were able to bring the prognostic
variables closer to the NARR analysis. However,
the specific humidity fields at the high
troposphere had increased values. This could be
relate to the R-2 precipitation higher values
found at the same area. Daily and 3-hourly
CMORPH precipitation analyses had slightly
different responses, and increased specific
humidity values were not found during any of the
CMORPH assimilations.