Challenges and Limitations of regional climate model
Download
Report
Transcript Challenges and Limitations of regional climate model
Challenges and Limitations of regional
climate model simulations in West Africa
for Present and Future studies
Gregory S. Jenkins
Penn State University
Present and Future Climate issues
in West Africa
• Present
• Future
• Climate Variability
• Land-Use Change
• Climate Variability
• Land-Use Change
• Climate change
Issues in Regional Modeling for
West Africa
• Model integration
• Regional Model parameterizations
• The state of knowledge in regional climate
systems and observations for verification.
• Science topics
• Extension to other disciplines
Model integration
• Preprocessing
• Initial and Boundary conditions
(observations and models)
• Grid Spacing
• Computing Power
• Storage
• Post-processing
Regional Model
parameterizations
• Land Surface-Atmosphere transfer
parameterization
• PBL parameterization
• Cloud parameterization
• Precipitation schemes
• Radiation schemes
The state of knowledge in regional
climate systems and observations for
verification.
• West Africa
–
–
–
–
–
–
Precipitation Processes (trigger, characteristics)
Low level southwesterly winds
600-700 hPa African Easterly Jet
African Easterly Waves (AEWs)
200 hPa Tropical Easterly Jet
Meridional surface temperature gradients.
MCSs links to the Water Cycle
• MCSs (squall lines, MCCs, non-squall clusters)
are responsible for most of West Africa’s rain.
• MCSs are found primarily in the Sahel region—
downstream of mountains.
• A significant fraction of the MCSs are long-lived
(> 24 hrs) –Mathon and Laurent, (2001).
MCSs links to the Water Cycle
• MCSs reach their maximum coverage
between 1800-0000 UTC.
• The highest frequency of MCCs are over
the Sahel.
Nesbitt et al. 2000 for Aug-Oct.
1998
• MCSs are a small
fraction of
precipitation features
in Africa (< 3%) but
account for more than
50% of rain.
MCSs are associated with high flash
rates (Nesbitt et al. 2000)
Conve ctive Rain Fr action
0.8
0.75
Fract ion
0.7
Conv. Fr ac ( Guinea)
0.65
Conv. Fr ac ( Sahel)
0.6
0.55
0.5
May
June
July
1998- 2000
Aug ust
Sept.
Str atifor m Rain Fr actions
0.45
0.4
Fract ion
0.35
Strat. Fr ac ( Guinea)
Strat. Fr ac ( Sahel)
0.3
0.25
0.2
May
June
July
1998- 2000
Aug ust
Sept.
Model Simulations
• CCM3.6 forced by Observed SSTs (AMIP)
• CCSM (A1B scenario) – SSTs computed
from fully dynamic ocean
• RegCM 60km and 90 km runs forced by
CCSM.
700 hPa AEJ simulations
CCM/CCSM/Observational Air
Temps.
Sea Level Pressure comparison
ECMWF/CCM3 u,v spectrum
ECMWF/CCSM spectrum of u,v
CCM3/CCSM u,v spectrum
Regional Climate model driven
by CCSM
Evidence of AEWs from
Meridional winds at .688s
Evidence of AEWs from Rain
rates
Evidence of AEWs from Water
vapor at .668s
Science topics
•
•
•
•
Climate Variability
Land-use Change
Climate change
Sensitivity Studies
West African Precipitation
Anomalies
“Africa is Highly Vulnerable to
Climate Change.”
IPCC 2001
A1B scenario
21st century temp./prec. diff
Extension to other disciplines
•
•
•
•
•
Agriculture
Water Resources
Health
Economic Development
Policy
Limitations of Regional climate
modeling for West Africa
• Resource to complete science (computers, storage,
boundary conditions)--especially for climate
change integrations.
• Recognition that regional modeling results are
very dependent on input of boundary conditions
(poor results can be anticipated !).
• Critical mass of scientists to address various
science questions.
• Observations to compare to model results.
• Forum to discuss/publish results and issues
Challenges of Regional climate
modeling for West Africa
• Formation of working groups to pool resources for
examining scientific issues (climate variability,
land-use, climate change, sensitivity tests,
improvements in model parameterization).
• Recognition that regional climate models may
need to undergoe significant modification to be a
useful tool in a particular region.
• Develop tools that will ease the use of climate
modeling (pre-processing, integration, postprocessing) on multiple platforms (PCs,
workstations, supercomputers).