Transcript A. Anthony Chen - Uwi.edu - University of the West Indies

CARIBBEAN CLIMATE MODELLING
and Adaptation
A. ANTHONY CHEN &
JAYAKA CAMPBELL
THE CLIMATE STUDIES GROUP MONA (CSGM)
UNIVERSITY OF THE WEST INDIES, MONA
OUTLINE

Before 2007
Climate Modelling
 Importance of modelling
 Early Works

2007 – IPCC 4th Assessment
 After 2007
 Some Results
 Current
• Future Work

2
CLIMATE

Defining key terms
 Climate
 “average weather” over period of ~ 30 years usually.
 Climate Change
 A significant change from one climatic condition to
another
 Due to natural or anthropogenic causes.
3
Climate Modelling is important

Atmosphere is too vast

Impossible to carry out large scale experiments
Climate simulation or modelling is the only way to study
large scale systems of the atmosphere
 Climate models are an important tool for scientists to
understand the complexities of Earth’s climate
 incorporate both theory and direct observations of the
past and present in order to project climate into the
future
 Lay the foundation for decision making concerning
climate change

Climate Models
 Introducing
key tools
 Dynamic Climate Model
Equations describing processes
Types
 General Circulation Models – GCM’s
 Regional Climate Models – RCM’s


Domain –Region over which RCM works
Statistical Downscaling Models
Reducing GCM to finer scale by use of
statistical relationships between atmospheric
parameters and climate variables
Require ~ 30 years of daily data
5
Dynamic Computer Models
Solves for/calculates and steps forward in time
• Equations of motion
• First law of thermodynamics
• Physics of water vapor and clouds
•Chemical processes in atmosphere
• Land - atmosphere interactions
Biological processes
• Land - ocean interactions
Distinction between GCM
and RCM
GCM Resolution are too coarse
to identify small islands
(Boxes too large)
RCMs have finer resolution
(Smaller boxes)
7
Jamaica as an example
GCM – 2 grid boxes
RCM – 10 grid boxes
Use Scenarios for Climate Change
Studies
 Plausible future climate based on perceived
greenhouse gas emission which depend on
developments in
 Demographics
 Technology
 Economics
 Special
Report on Emission Scenarios
(SRES)

SRES Scenarios can be classified into four
family groups further divided into scenario
groups.
9
SRES FAMILY TREE
• Roots
are the drivers
• Population
• Economy
• Technology
• Etc
• Branches depend on whether
development is
• globally intergated or
regional
• Driven by economics or
environmental considerations
Source : http://www.grida.no/climate/ipcc/emission/
10
SRES S CENARIOS
Source : http://www.grida.no/climate/ipcc/emission/spm-3.htm
11
CARIBBEAN CLIMATE MODELLING
INITIATIVE 2002

CSGM along with other Caribbean partners (INSMET –
Cuba, UWI Cavehill – Barbados and CCCCC – Belize) begun
Caribbean modelling initiative

Deliberate collaborative effort to produce Caribbean
climate projections at scale of Caribbean.

Premised on shared workload to get results out quickly.

Premised on building of capacity in the region.

Multiple components to the strategy, but concentrate on a
major one.
EARLY WORKS –
2002 CLIMATE MODELLING INITIATIVE
Introduction a Regional Climate Model
PRECIS - Providing REgional Climates for Impact Studies
• Hadley Centre, UK
•
•
•
•
•
•
•
Dynamical Downscaling Model
(RCM)
Can be used for any part of the
Globe
Has a resolution of up to 25km
Driven by full suite of physics
Multiple variables on multiple
levels in atmosphere.
Forced at its boundaries by other
GCMs - the HADAM3P GCM and
ECHAM.
Built by UK Hadley Centre but run
locally
PRECIS Modelling
•
Complex but
computationally less
expensive than a
GCM.
•
Requires a Desktop
Standard Desktop
Pentium 4 Processor
•
Could be run locally
EARLY WORKS – CLIMATE MODELLING INITIATIVE
DOMAIN
•
Big Domain including all Caribbean, Central America, southern USA and
northern South America. Run at 50 km
Two smaller domains:
• Western & Eastern Caribbean (yellow and red enclosure) at 25 km
EARLY WORKS – CLIMATE MODELLING
INITIATIVE
PRECIS Workload Shared
Cuba
Carib basin
(INSMET)
50 x 50 km
Jamaica –
UWI (Mona)
Carib Basin
Barbados –
UWI (Cave
Hill)
Eastern Caribbean
A2 (30 yrs) & B2 (30 yrs)
25 x 25 km
Baseline (30 yrs)
Belize - 5C’s
Caribbean and
Eastern Caribbean
Multiple runs
50 x 50 km
B1 (30 yrs) & A2 (30 yrs)
Baseline (30 yrs)
Reanalysis (15 yrs)
A2 (30 yrs) & B2 (30 yrs)
Baseline (30 yrs)
Baseline (1961 – 1990), Reanalysis (1979 – 1993), A2
& B2 (2071 – 2100)
E ARLY W ORKS – M ODEL V ALIDATION
Precipitation
(A) Rainfall
Early season April –
June
Late season August November
Temperature
(B) General temperature
Pattern Captured
16
E ARLY W ORKS –
M ODEL P ROJECTIONS
A2
General tendency
for drying (main
Caribbean basin)
by end of the
century.
Drying between 25%
and 30%
B2
Mean changes in the annual rainfall for
2071-2099 with respect to 1961-1990, as
simulated by PRECIS_ECH and
PRECIS_Had for SRESA2 and SRESB2.
Possibly wetter far
north Caribbean
NDJ and FMA.
Drying exceeds
natural variability
June-October – wet
season dryer!
E ARLY W ORKS –
M ODEL P ROJECTIONS
Precipitation Natural Variability
18
E ARLY W ORKS –
M ODEL P ROJECTIONS
Projected Precipitation Change
General tendency
for drying (main
Caribbean basin)
by end of the
century.
Drying between
25% and 30%
Possibly wetter
far north
Caribbean NDJ
and FMA.
Drying exceeds
natural variability
June-October – wet
season dryer!
19
E ARLY W ORKS –
M ODEL P ROJECTIONS
A2
Irrespective of
scenario the
Caribbean expected
to warm.
Warming between 1
and 5oC
B2
Mean changes in the annual mean
surface temperature for 2071-2099
with respect to 1961-1990, as
simulated by PRECIS_ECH &
PRECIS_Had for SRESA2 and SRESB2.
Warming greater
under A2 scenario.
Warming consistent
with projections for
other parts of globe.
Warming far exceeds
natural variability
E ARLY W ORKS –
M ODEL P ROJECTIONS
Temperature Natural Variability
21
E ARLY W ORKS –
M ODEL P ROJECTIONS
Projected Temperature Change
22
Not Reported in IPCC
(2007)
Unable to report results until 2007 due to
amount of work involved. Not reported in IPCC
2007 4th Assessment
IPCC 2007 4th Assessment mentioned:
a) GCM work done at UPR Mayaguez:
Angeles, M.E., J.E. Gonzalez, D.J. Erickson, and J.L. Hernández, 2007:
Predictions of future climate change in the Caribbean region using global
general circulation models Int. J. Climatol., 27, 555-569
b) Work
done on statistical downscaling at UWI, Mona:
http://www.aiaccproject.org
IPCC Fourth Assessment 2003 – 2006
Published in 2007
Working Group 1 Report: The
Physics Science Basis
(Data Analysis & Climate Modelling)
Working Group II Report:
Impacts, Adaptation and
Vulnerability
Working Group III Report
"Mitigation of Climate Change"
NOT ENOUGH REGIONAL MODELLING
DONE FOR THE CARIBBEAN
o Report
based on course General Circualtion models
Temperature Changes under A1B Scenario
1980-1999 to 2080-2099
Temperature Changes under A1B Scenario
1980-1999 to 2080-2099
Rainfall Changes under A1B Scenario
1980-1999 to 2080-2099
Rainfall Changes Changes under A1B Scenario 19801999 to 2080-2099
After 2007- A Few Publications –
Modelling Initiative
Taylor MA, A Centella, J Charlery, I
Borrajero , A Bezanilla, J Campbell, R Rivero,
TS Stephenson, F Whyte, R Watson, 2007.
Glimpses of the Future: A Briefing from the
PRECIS Caribbean Climate Change Project,
Caribbean Community Climate Change
Centre, Belmopan, Belize.
Campbell J, MA Taylor, TS Stephenson, RA
Watson, FS Whyte, 2010: Future Climate of
the Caribbean from a Regional Climate Model.
Int J Climatol DOI: 10.1002/joc.2200.
Other reports:
Cashman, A., L. Nurse, and J. Charlery. 2010. Climate change in
the Caribbean: The water management implications. The Journal
of Environment & Development 19(1): 42-67.
Charlery, J., and L. Nurse. 2010. Areal downscaling of global
climate models: an approach that avoids data remodelling. Clim
Res 43: 241–249. doi: 10.3354/cr00875.
After 2007 – USE OF DATA
Reporting Purposes
St. Lucia
RCM
RCM
Compiling projections for
use in 2nd National
Communications:
Antigua
St. Lucia
St. Vincent
Grenada
Jamaica
2070s
2070s
JAN
1.6 - 1.9
JAN
-46.91 --25.90
FEB
1.8 - 2.3
FEB
-78.37 - -50.04
MAR
1.9 - 2.5
MAR
-86.14 - -50.95
APR
1.9 - 2.8
APR
-81.80 - -57.79
MAY
2.2 - 2.7
MAY
-69.91 - -47.49
JUN
2.1 - 2.7
JUN
-77.10 - -47.26
JUL
1.9 - 2.5
JUL
-57.86 - -29.02
AUG
1.9 - 2.2
AUG
-36.29 - -21.95
SEP
2.0 - 2.2
SEP
-39.94 - -21.95
OCT
1.9 - 2.3
OCT
-33.13 - -3.76
NOV
1.8 - 2.1
NOV
-30.37 - +18.58
DEC
1.8 - 2.1
DEC
-48.69 - -8.32
ANNU
AL
1.9 - 2.4
ANNUAL
-57.21 - -27.94
Temps
Rainfall
MORE RECENT – PILOT PROJECT
FOR CLIMATE RESILIENCE (PPCR)
Climatology, Trends,
Projections and Impacts for
Jamaica
 Funded by adaptation and
mitigation funds following
Copenhagen and Cancun
 PPCR – the first program
developed and operational
under the Strategic Climate
Fund (SCF),
 SCF one of two funds within
the design of the Climate
Investment Funds (CIF)

31
CURRENT & Future –
MODELLING INITIATIVE
 Uncertainies
in model results
 Increase confidence limits by averaging the results
of many models
 Expand range of RCMs
 PRECIS
 WRF
 REMO
 CCSM
 Use
of high resolution GCM (Japanese 20 km
model)
C URRENT –

M ODELLING I NITIATIVE
Investigating Hurricane* like vortices present in climate
models.
Model
WRF - H
33
C URRENT –
M ODELLING I NITIATIVE

Addition of the Japanese MRI Earth Simulator model
results
 GCM
 Horizontal resolution 20km
 24 Levels of atmosphere
 26 Variables (16 surface, 10 3D)
 All variables are monthly averages

GCM model with highest resolution so far
34
J APANESE M ODEL R ESULTS T OPOGRAPHY
C URRENT –
M ODELLING I NITIATIVE
Diagram shows how accurately the 20 km
JMA/MRI Model represents topographical heights.
GTOPO30 - observed topographical heights.
20, 60, 120 and 180 km model simulations
Note : Above 60 Km even the mountains of larger
territory of Hispaniola aren’t simulated
35
CURRENT – MODELLING INITIATIVE
J APANESE M ODEL R ESULTS T OPOGRAPHY
Looking at Western Caribbean mountains (Belize)
from South to North, 20 km resolution better than
60, 120, 180
Diagram the comparison of Grey blocks (GTOPO30 - observed topographical heights) and
modeled 20, 60, 120 and 180 km simulations of cross-section for particular latitudes
36
F UTURE – M ODELLING

I NITIATIVE
Past & Current work offers a first look at the possibilities
for the future of Caribbean climate. However we need to:
 Expand the range of forcing GCMs of the PRECIS
model from HADCM3 and ECHAM5
 Incorporate more RCMs to add a greater degree of
confidence to generated projections.
 Move from atmosphere only models to ones with a
coupled ocean-atmosphere framework
 Run a full ensemble of climate modelling scenarios, not
just the extremes in each case.
 More Statistical downscaling
37
Future - Analysis of Model Results using Representative
Concentration Pathways (RCP) for IPCC 5th
Radiative
forcing:
measure of
difference
between
incoming and
outgoing
radiation in
the
atmosphere
Representative concentration pathways.
Image 2.6 emissions will restrict temperature rise to 2ºC
RH Moss et al. Nature 463, 747-756 (2010) doi:10.1038/nature08823
F UTURE – M ODELLING

&
OTHER SECTORS
Form closer relationship with the Impact and Adaptation
community.
 Climate Impacts our daily life; How do we adapt?
 The Caribbean is heavily dependent on agriculture
and tourism, so what is the significance of a
projected 1ºC – 5ºC rise in temperature to those
sectors?
 What do the projected changes mean for the onset
or spread of tropical diseases (examples .. dengue or
leptospirosis)?
 How or will the projected changes in climate impact
development and plannning?
39
F UTURE
More Partnerships
Too difficult, time consuming for any one country
or institution.
Collaboration heightens efficiency for producing
usable results.
Builds synergies/support groups across
institutions.
End-Note




Climate affects all aspects of life (Taylor, 2010)
Climate change is a threat to all life
Concern of all our government ministries, private
sector organizations and civil society
With no global accord on the reduction of GHGs,
 threat of climate change will become very
severe
 the cost of adapting, especially to sea level
rise, will become enormous
 future prosperity of the Caribbean community
will be at risk.
Thank You
Climate Studies Group Mona