An Overview of New and Future Developments with the NCEP Climate Forecast System

1
An Overview of New and Future Developments with
the NCEP Climate Forecast System

• SURANJANA SAHA and HUA-LU PAN
• Environmental Modeling Center
• NCEP/NWS/NOAA
• 21st Annual Climate Diagnostics and Prediction
  Workshop
• Boulder, CO
• 23 October 2006

2

• The NCEP Climate Forecast was made operational in
  August 2004
• Currently, two fully-coupled nine-month
  forecasts are made every day
• The present CFS operational system is frozen
• Development work is underway at EMC to improve
  the CFS
• We anticipate a new CFS implementation will take
  in January 2010

3
For a new CFS implementation Two main
components CFS Reanalysis (1979-2007)
T254L64 CFS Retrospective Forecasts (1981-2007)
T126L64
4

• For a new CFS implementation (contd)
• Analysis Systems GSI, GODAS, GLDAS
• Atmospheric Model GFS
• Ocean Model MOM4 (1/40 at the Equator,
  1/20 at globally beyond 100N and 100S
• SEA ICE MODEL

5
Proposed Time-Line for the next CFS
implementation in January 2010 October 2006 In
anticipation of the operational implementation of
the sigma-p version of the GSI in February 2007,
start testing this version for a few particular
years in the 1979-2006 period, where changes in
satellites will lead to new corrections in
radiance measurements. Run the GSI for a few
random years and save the fluxes. (Saha,
Treadon, van Delst, Derber, Kliest, Woollen, Pan,
Kistler, White)
6
Proposed Time-Line for the next CFS
implementation in January 2010 June 2007 In
anticipation of the ESMF version of the coupler
and MOM4, start testing the new GODAS (with MOM4
and sea ice) and GLDAS (with Noah Land model and
observed precipitation) with fluxes obtained from
atmospheric GSI analyses of the few random
years. (Behringer, Mitchell, Sheinin, Wang, Wu,
Nadiga, Stokes, Pan, Moorthi, Grumbine, Saha,
Meng, Wei)
7
Proposed Time-Line for the next CFS
implementation in January 2010 August 2007 In
anticipation of the operational implementation of
the next GFS, start pilot studies with the first
prototype of the fully coupled CFS reanalysis,
which will include the GSI (with new GFS), GODAS
and GLDAS. (Saha, Pan, Iredell, Moorthi, Derber,
Behringer, Mitchell, Meng, Yang, Treadon,
Sheinin, Woollen, Kistler)
8
Proposed Time-Line for the next CFS
implementation in January 2010 January 2008
Begin Production and Evaluation of the CFS
Reanalysis for the full period from 1979 to 2007
(29 years) Reanalysis Team Saha, Thiaw, Wang,
Nadiga, Wu, Lu. Evaluation Team EMC and CPC
Personnel
9
Proposed Time-Line for the next CFS
implementation in January 2010 July 2008 Begin
running CFS Retrospective Forecasts for 2 initial
months October and April, and evaluate the
monthly forecasts as well as the seasonal Lead-1
DJF and JJA forecasts. Hindcast Team Saha,
Thiaw, Wang, Nadiga, Wu, Lu Evaluation Team EMC
and CPC Personnel
10
Proposed Time-Line for the next CFS
implementation in January 2010 January 2009
Operational Implementation of the CFS
Reforecast Project (for the rest of the 10
calendar months) in the proposed slot of REFCST
in the NCEP production suite. (NCO personnel,
Saha, Moorthi, Pan and Thiaw)
11
Proposed Time-Line for the next CFS
implementation in January 2010 November 2009
Begin computing calibration statistics for CFS
daily, monthly and seasonal forecasts. Prepare
CFS Reanalysis and Retrospective Forecast data
for public distribution. (Saha, Thiaw, Pan and
CPC personnel) January 2010 Operational
implementation of the next CFS monthly and
seasonal forecast
12
CURRENT OPERATIONAL VERSION OF THE GFS (USED FOR
WEATHER PREDICTION) UPGRADES TO THE CFS VERSION

• NOAH Land Model 4 soil levels. Improved
  treatment of snow and frozen soil and glacial
  physics
• Sea Ice Model Prediction of ice concentration
  and ice fraction
• Sub grid scale mountain blocking
• Reduced vertical diffusion

13
CURRENT OPERATIONAL VERSION OF THE GFS (USED FOR
WEATHER PREDICTION) UPGRADES TO THE CFS VERSION
RRTM long wave radiation (clouds are maximum
random, which leads to reduced cloud cover) ESMF
Version NRL Based Ozone Climatology for
Production and destruction
14

• Ongoing GFS Developmental work
• Test the generalized vertical coordinate system
  with a combination of sigma, theta and pressure
  levels
• Test new convection scheme (RAS)
• Test upgrades to operational convection scheme
  (SAS)
• Test Moorthi-Ferrier microphysics package for
  large scale condensation
• Test improved boundary layer physics
• Test convectively forced gravity wave drag

15

• Other Ongoing CFS Developmental work
• Test the new Grid Point Statistical Interpolation
  Scheme (GSI)
• Test historical settings of CO2, aerosol and
  solar cycle in the CFS
• Test the new MPI level coupling of GFS to MOM4

16
GSI NCEPs Next Generation Analysis System

• John Derber1, Lidia Cucurull2, Daryl Kleist3, Xu
  Li3, Curtis Marshall3, Dave Parrish1, Manuel
  Pondeca3, Jim Purser3, Russ Treadon1, Paul
  vanDelst2, Wan-Shu Wu1
• 1 NOAA/NWS/NCEP/EMC, 2 UCAR, 3 SAIC

Courtesy Russ Treadon, EMC
17
Assimilated data types

• All data types currently assimilated by SSI may
  also be assimilated by GSI
• Sondes, ship reports, surface stations, aircraft
  data, profilers, etc
• Cloud drift and water vapor winds
• TOVS, ATOVS, AQUA, and GOES sounder brightness
  temperatures
• SBUV ozone profiles and total ozone
• SSM/I and QuikScat surface winds
• SSM/I and TMI rain rates

Courtesy Russ Treadon, EMC
18
GSI development Analysis variables

• SST analysis
• Physical retrieval from AVHRR Tb data
• Option to add / assimilate in-situ SST data

rms
Slight, but consistent reduction in rms and
bias fits to independent buoy SST data
bias
Courtesy Russ Treadon, EMC
19
GSI development New radiance data

• Aqua AIRS/AMSU-A
• Operational as of 12 UTC, 31 May 2005
• Future improvements
• Examine all FOVs to determine warmest spots
• Use MODIS data for cloud detection
• SSM/I
• Use of Tb data reduces model moisture bias
• Forward model for emissivity includes effects of
  surface winds
• Assimilation of SSM/I Tb data can affect surface
  winds
• Could (should) turn off assimilate of SSM/I wind
  product

Courtesy Russ Treadon, EMC
20
GSI development New radiance data

• NOAA N (18) Summer 2005
• HIRS, AMSU-A, MHS
• Code ready and waiting for data to evaluate
• SSM/IS
• QC and bias correction difficulties because FOVs
  not collocated
• AMSR-E (NASA)
• Beginning tests with radiative transfer model
• AVHRR and GOES imagery
• Testing underway

Courtesy Russ Treadon, EMC
21
GSI development GPS radio occultation

• Preparation for COSMIC well advanced
• Code for assimilating local refractivities done
  and being tested
• Adding code to handle local bending angle
  underway
• QC issues
• Tracking errors
• Caused by complicated refractivity structure in
  moist lower troposphere
• Super-refraction
• Occurs on sharp top of moist PBL

Courtesy Russ Treadon, EMC
22
GSI development Doppler Radar

• Code being developed to handle radar radial
  velocities
• Currently working on data processing, quality
  control, and superobs issues
• Longer term project is to make use of radar
  reflectivities
• Currently working on quality control issues
• Bird migration, ground clutter, anomalous
  propagation, etc

Courtesy Russ Treadon, EMC
23
GSI development CRTM development

• Proto-type CRTM with modular design
• Simplifies user interaction with code
• Permits easier evaluation of various algorithms
• Soon will include
• Algorithms to handle scattering and absorption
  from clouds for microwave channels

Courtesy Russ Treadon, EMC
24
Observation influence extends into mountains
indiscriminately
Anisotropic vs Isotropic Error Covariances
Error Correlations Plotted Over Utah Topography
Observation influence restricted to areas of
similar elevation
Courtesy Russ Treadon, EMC
25
INTRODUCTION TO POSTERS
26
Examples of Subseasonal Forecasting with the CFS
The Sahel
Augustin Vintzileos and Wassila Thiaw

• Forecast skill for Cumulative Precipitation over
  the Sahel

27
Examples of Subseasonal Forecasting with the CFS
The Sahel
Observed precipitation over the western Sahel for
2006 (Xie)
Real Time 30-day Forecasts good vs. bad
prediction
2006 Precipitation (blue)
Mean annual Precipitation (red)
Observed
Forecast
Observed
The 2006 monsoon period over the Sahel started
drier than average. Circa July 16th, there was a
sudden jump towards persistent wetter conditions
and finally the 2006 monsoon season was a good
one. We discuss the ability of the CFS to
forecast these conditions at subseasonal lead
times on the Poster Session P1.12 on Tuesday.
Forecast
28
CFS Retrospective Forecast Daily Climatology in
the EMC/NCEP NOMAD public server Åke Johansson,
Catherine Thiaw and Suranjana Saha, Environmental
Modeling Center, NCEP/NWS/NOAA
Methodology A Fourier series is fitted to the
raw 24-yr average data y(tj) which are available
at 180 irregularily spaced days in a 365 day year
T 365 days . The method of least squares is
used, i.e., minimization of gives the Fourier
Coefficients. The climatology is defined by
considering 4 harmonics, i.e., in accordance
with NCEP practice, Schemm et al. 1998. There
is a risk of Overfitting Noise included
in the climatology Underfitting
Not all of the true climatology is included in
the
calculated climatology
29
CFS Retrospective Forecast Daily Climatology in
the EMC/NCEP NOMAD public server Åke Johansson,
Catherine Thiaw and Suranjana Saha, Environmental
Modeling Center, NCEP/NWS/NOAA
MEAN
SD
2m Temperature Washington
1 month lead
Winter-time Warming of 2-3 ºC
Winter-time Variability reduced by 15
8 month lead
30
NCEP REGIONAL CLIMATE FORECAST BY NCEP RSM MODEL

• Regional climate models have been used in
  dynamic downscaling for the seasonal climate
  forecasts and regional climate change assessment
  in the past decade and demonstrated capability in
  regenerate seasonal mean climate and inter-annual
  variability.
• NCEP Regional Spectral Model (RSM) nested in
  Global Forecast System (GFS) has been implemented
  for CONUS regional climate prediction. A 20 year
  3 member ensemble hindcast using sst forecasted
  from CFS was conducted to explore the capability
  of NCEP RSM model on seasonal forecast.

Jun Wang and Henry Juang, P1.19 920-1050AM,
Tuesday, 10/24/2006
31
Jun Wang and Henry Juang, P1.19 920-1050AM,
Tuesday, 10/24/2006