International Seminar On Climate Variability, Change and Extreme Weather Events

1
Regional Climate Change over Southeast Asia Region
Mohan Kumar Sammathuria, Ling Leong Kwok Wan
Azli Wan Hassan Malaysian Meteorological
Department Ministry of Science, Technology
Innovation, Malaysia
International Seminar On Climate Variability,
Change and Extreme Weather Events 26-27 February
2008, Bangi, MALAYSIA
2
SCOPE

• Introduction
• Present Climate (1961-1990)
• Future Climate (2071-2100)
• Mean Temp (annual seasonal) Anomaly
• Mean Precip (annual seasonal) Anomaly
• Seasonal Mean Wind Anomaly
• Concluding Remarks

3
GCMs to Regional Adaptive Responses Modelling
Path
PRECIS 50 km
4
Projected climate change depend on illustrative
scenarios (storylines) of greenhouse gases
emissions Special Report on Emission Scenarios
(SRES)

• Based on different plausible pathways of future
• development of the world
• population growth and consumption patterns
• standards and life style of living
• energy consumption energy sources (e.g. fossil
  fuel usage)
• technology change
• land use change

5
Four Marker IPCCs SRES Future Emission Scenarios
A qualitative description of the SRES scenarios
6
The driving model HadCM3 has predict climate
change (global temperature rise) arising from
each of the four IPCCs SRES future emissions
scenarios
5.0oC 2.0oC IPCC AR4 B1 1.8oC
(1.1-2.9) B2 2.4oC (1.4-3.8) A2 3.4oC
(2.0-5.4) A1FI 4.0oC (2.4-6.4)
7
PRECIS

• Providing REgional Climates for Impact Studies
• High-resolution limited area model driven at its
  lateral and sea-surface boundaries by output from
  HadCM
• PRECIS runs on Linux PC (horizontal resolutions
  50 x 50 25 x 25 km).
• Needs data for the selected domain on lateral
  boundary conditions (LBC) from the driving GCM
  (e.g., HadCM3/ HadAM3) and the associated
  ancillary files (e.g., sea surface temp,
  vegetation, topography, etc).
• Hadley Centre, UK has been providing PRECIS as
  well as the driving data to several regional
  groups.
• Baseline (1961-90), A2 B2 scenarios
  (2071-2100). Reanalysis-driven runs provide
  comprehensive regional data sets representing
  current conditions, which can assist model
  evaluation as well as assessment of vulnerability
  to current climate variability.
• Ensembles to estimate model-related uncertainties.

8
Orography Resolution
PRECIS resolution 0.44 x 0.44
HadCM3 resolution 2.5 x 3.75
9
PRECIS Runs at MMD

• LBCs derived from HadAM3P. HadCM3 provided SST as
  boundary conditions for HadAM3P.
• A2 B2 scenarios runs of PRECIS performed
  consecutively on a PC.
• PRECIS runs on Linux PC (horizontal resolutions
  0.44 x 0.44)
• The LBCs have a length of 31 years, and are
  available for Baseline (1961-90), A2 B2
  scenarios (2071-2100), with the sulphur cycle.
• The basic parameters analyzed are the mean
  surface (1.5 m) temp and total precip.
• The precip temp obs data (CRU20, 1961-90) is
  used to validate model performance in simulating
  current climate.
• The analysis comprised of both annual mean and
  seasonal mean for DJF, MAM, JJA and SON.
• To detect possible atmospheric circulation change
  during monsoon periods (DJF JJA) in future
  climate, the seasonal mean 850 hPa wind for the
  lower emission scenario (B2) was analysed.

10
PRECIS captures important regional information on
summer monsoon rainfall missing in its parent GCM
simulations
11
PRECIS performs reasonably well too on winter
monsoon rainfall compared to its parent GCM
simulations
12
PRECIS Simulations of Present Climate (1961-1990)
Mean Annual Cycles of SEA Rainfall and Temperature
13
PRECIS Simulations of Future Climate (2071-2100)
Mean Annual Cycles of SEA Rainfall and Temperature
14
Mean Annual Temp Anomaly
Continental larger ve anomaly (A2, 3.0-4.5
C B2, 1.5-3.0 C)
Maritime smaller ve anomaly (A2, 2.0-3.5 C
B2, 0.5-1.5 C)
Larger anomaly over SCS vs western Pacific in A2
N-E P. Malaysia Smaller ve anomaly
c-S P. Malaysia, Sabah Sarawak Larger ve
anomaly
15
(A2-Baseline) Mean Seasonal Temperature Anomaly
MAM
DJF
SON
JJA
16
(B2-Baseline) Mean Seasonal Temperature Anomaly
DJF
MAM
SON
JJA
17
Mean Annual Precip Anomaly
-12
Precip deficit over maritime SEA
-7
Northern P. Malaysia (A2, 17 B2, 6)
Sarawak (A2, 5 B2, -8)
Southern P. Malaysia (A2, -3 B2, -20)
Sabah (A2, -15 B2, -18)
18
SEA Mean Seasonal Precip Anomaly
Larger deficit in B2
Deficit in most seasons
19
(A2-Baseline) Mean Precip ()
MAM -21
DJF -5
JJA -8
SON 1
20
(B2-Baseline) Mean Precip ()
DJF -21
MAM -24
JJA -13
SON -9
21
Mean Seasonal Precip Anomaly
Malaysia NEGATIVE ANOMALY mean precip in DJF
22
Mean Seasonal Precip Anomaly
Northern P. Malaysia POSITIVE ANOMALY mean
precip in JJA SON, deficit in DJF MAM
23
Mean Seasonal Precip Anomaly
Southern P. Malaysia deficit mean precip in
DJF, MAM JJA, ve anomaly in SON
24
Mean Seasonal Precip Anomaly
Sabah largest deficit in DJF MAM Sarawak
DJF only
25
Mean Seasonal 850 hPa Wind Anomaly (DJF)
Baseline
Anomaly
Weakening easterly (2.0-3.5 m/s)
Rainfall Anomaly
26
Mean Seasonal 850 hPa Wind Anomaly (JJA)
Baseline
Anomaly
Anomalous easterly comp. (1.5-2.5 m/s)
Rainfall Anomaly
27
Concluding Remarks

• PRECIS was found able to capture important
  regional information on seasonal rainfall which
  is missing in GCM simulation
• Both A2 B2 scenarios show an increase in the
  annual mean temp over SEA during 2071-2100, with
  A2 shows larger increase in temp
• The SEA land surface annual mean warming is in
  the range of 1.5-3.0 C with B2 and 3.0-4.5 C
  with A2
• The SEA maritime surface annual mean warming is
  0.5-1.5 C with B2 and 2.0-3.5 C with A2

28
Concluding Remarks (cont.)

• Both scenarios show a ve anomaly of mean annual
  precip over SEA continent while a -ve anomaly
  over maritime region
• SEA, at large will experience a deficit in mean
  annual precipitation for both A2 and B2
  scenarios, with B2 giving the larger deficit
• Weakening of the easterly during the winter
  months (DJF) over the western Pacific region in
  B2 scenarios indicates a weakening of the NE
  monsoon in SEA region
• In summer (JJA), the anomalous easterly component
  winds over the Indian Ocean will tend to enhance
  the ve IOD phenomenon

29
Note Multi-Model IPCC AR4 Uncertainty Ranges B1
1.8oC (1.1 - 2.9) A1T 2.4oC (1.4 -
3.8) B2 2.4oC (1.4 - 3.8) A1B 2.8oC (1.7
- 4.4) A2 3.4oC (2.0 - 5.4) A1F1 4.0oC (2.4
- 6.4)
Concluding Remarks (cont.) This is our
preliminary results. More works are needed to
obtain credible climate change scenarios with
better certainty..
IPCCs AR4 employed multi-model means of surface
warming for the SRES marker scenarios. Numbers
indicate the number of models which have been run
for a given scenario. The gray bars at right
indicate the best estimate (solid line within
each bar) and the likely range assessed for the
SRES marker scenarios. RCM (e.g. PRECIS), too,
should be driven by multi-model in order to know
the uncertainty range of climate change
30
Thank You