Applications of Regional Climate Modeling for WeatherClimate Interface Research

1
Applications of Regional Climate Modeling for
Weather-Climate Interface Research

• James Done1, L. Ruby Leung2 and Bill Kuo1
• 1 NCAR / MMM
• 2 Pacific Northwest National Laboratory

2
Why Regional Climate Modeling?

• Current GCMs lack the spatial resolution to
  represent regional scale processes and feedbacks.
• High resolution ? more precise description of
  regional topographic forcings due to orography,
  land-sea contrasts and land surface
  characteristics.
• Applications
• Downscaling of climate variability and change at
  the regional scale (e.g., climate change effects
  on water resources, ecosystem, extreme weather
  hurricane frequency storm track distribution of
  MCS and warm season precipitation use of
  seasonal forecasts for water management)

3
Why Regional Climate Modeling?

• Applications continued. . . .
• Process studies (e.g., Amazon biomass burning and
  aerosol effects orographic effects
  land-atmosphere interactions ocean-atmosphere
  interactions sea ice cloud-radiation feedbacks)
• Upscaling of regional phenomena with global
  consequences (e.g., subtropical and tropical
  eastern boundary upwelling regimes subgrid-scale
  clouds organized convection gravity wave drag)
• - relevant for weather-climate interface
  research.

4
RCM Development Using WRF

• Since Oct 2003, NCAR project to develop regional
  climate modeling capability with the Weather
  Research and Forecasting (WRF) model
• Model tested for
• - simulating cold season orographic
  precipitation in the western U.S. at 6 km and 30
  km resolution.
• - simulating warm season precipitation in the
  Midwest at 30 km resolution in the 1993 flood
  case.
• - 10 year simulation driven by PCM.
• Compatible physics with CCSM implemented CAM3
  radiation and the Community Land Model (CLM3)
• RCM capability will be released in standard
  version later in 2005

5
Monthly Precipitation (July 1993)
WRF Forecasts
WRF Climate
Obs (1/24, 4km)
MM5 Climate
6
MM5 and WRF driven by PCM (2-year average)
MM5
WRF
MAM
Observations
JJA
7
Workshop, March 22-23 2005

• Research Needs and Directions of Regional Climate
  Modeling Using WRF and CCSM
• Organizing committee L. Ruby Leung, Bill Kuo,
  Joe Tribbia, Phil Merilees
• 60 US and international participants

8
- Challenges

• Sensitivity to domain size/location
• What is the value of interior spectral nudging?
• Physics compatibility between global and regional
  models
• How to quantify added value of RCM
• Relative impacts of dynamical core vs physics in
  RCMs
• Implications of model internal variability for
  seasonal climate forecasts, climate change
  experiments and sensitivity studies

9
- Science Issues

• Interactions between tropical convection and
  mid-latitude dynamics
• Large scale and regional scale control on diurnal
  cycle of rainfall.
• Land-atmosphere and ocean-atmosphere interactions
  (e.g. initiation and propagation of the
  Madden-Julian Oscillation)
• Climatic controls on organized convection
• Convection-cloud-radiation-chemistry interactions

10
Recommendation Towards a Regional Earth System
Model

• WRF/ROMS (regional ocean modeling system) nested
  within CCSM with WRF interacting with ROMS and
  CAM, and ROMS interacting with WRF and POP
  (global ocean model)
• Necessary to answer many science questions (e.g.
  role of ocean in initiation and propagation of
  MJO).

11
Upscaling Research
Example motivating question Does regional
convection affect climatological large-scale
circulation? Need to represent scale
interactions. - currently, not achievable with
GCMs for long term simulation because of
computational constraints and/or limitations of
the hydrostatic formulation. Most of the
intriguing scale interaction issues involve
feedbacks between different earth system
components. - further motivation for coupling
capability.
12
Upscaling Research

• Method Two-way coupling of non-hydrostatic
  regional models capable of resolving dynamical
  processes at the 5 30km spatial scale with
  global climate models over hot spots (e.g.,
  Maritime Continent, Monsoon regions, and
  Subtropical Eastern Boundaries).
• Expectation Improve global model results without
  increasing resolution globally.

13
Example hot spot I Subtropical Eastern
Boundaries
CCSM2 SST Bias
14
Large-Scale Effects of DSST lt 0 off South America
and South Africa
15
Example hot spot II the Western Pacific
regional Warm Pool
A 10 year simulation using 2-way nesting.
(Lorenz and Jacob, 2005)
Model Orography Global model Regional
model
16
Zonal Mean Temperature Difference
DJF
JJA
ECHAM4 ERA15
ECHAM4(2-way) - ECHAM4

• Reduced systematic temperature errors

17
Summary

• RCM has demonstrated skill in simulating
  atmospheric water cycle and its variability
• Transition to regional earth system model opens
  opportunities to investigate a wide range of
  climate questions (including air-sea
  interactions) and interdisciplinary studies
• Two-way GCM-RCM coupling facilitates research on
  scale interactions and teleconnections, and is a
  useful tool for Weather-Climate interface
  research.

18

• EXTRA SLIDES

19
Current status of WRF

• Atmospheric chemistry model ready
• CCSM radiation (CAM3) and land surface model
  (CLM) ready
• Ocean model development (John Michalakes)
• NCAR hydrological model development (Dave Gochis)
• Merge to ESMF format in parallel with CCSM merger

20
North American Regional Climate Change Assessment
Program (NARCCAP)

• Explore multiple uncertainties in regional and
  global climate model projections at regional
  scales
• Develop multiple high-resolution regional climate
  scenarios for use in impact models
• Further evaluate regional model performance over
  North America.
• Explore remaining uncertainties