Download SEACAMd

The South East Asia Climate Analysis
& Modelling project (SEACAM)
David Hein, Climate Information Capability for International Development, Met Office
© Crown copyright Met Office
Contents
• Background and project objectives
• SEACAM activities
• Literature review
• Assessing user needs
• Model configuration and experimental design
• SEACAM results
• Model verification - assessing historical simulations
• Assessment of Mid- and Long-term Climate Change
Projections
• Summary and next steps
© Crown copyright Met Office
Background and project objectives
© Crown copyright Met Office
Geography and climate of Southeast
Asia
• Region expected to experience
serious impacts of climate change
(growing and urbanising population,
reliance on climate-sensitive sectors
e.g. agriculture, fisheries, natural
resources).
• Already affected by extreme weather
events, particularly tropical cyclones,
droughts and floods are expected to
increase
• Impacts of climate change not be
evenly distributed - some regions
may experience harsher impacts
than others.
• Policy-makers require reliable
science and information to inform
response
• Limited studies and climate projects
available for the region
© Crown copyright Met Office
Map showing the continental section of Southeast Asia on the top-left and the maritime
section to its south and east.
The SEACAM experiments
• Primary goal was the production and
analysis of climate projections for
Southeast Asia
• Planning workshop (June 2012,) brought
together representatives from 8 of the 10
ASEAN member countries
• Discussed the need for a dedicated
project for Southeast Asia using a
regional climate model (RCM)
• RCM would be driven by several global
models. Attendees decided to share the
work load required to run the models
• Six 150-year (1949-2099) “PRECIS”
regional climate model experiments would
be run, nicknamed the “DURIAN”
experiments
Trainers and participants at the June 2012 workshop in
Singapore.
• SEACAM would jointly analyse the 900
years (150 years times six experiments) of
climate model output data from the DURIAN
experiments
Acknowledgements
Project partners
•
Project supported through a Memorandum of
Understanding between the National
Environment Agency of Singapore and the Met
Office
•
Memorandum of Understanding promotes intercountry cooperation for climate change
research for Southeast Asia
•
Co-funding provided by the UK Foreign and
Commonwealth Office - “Southeast Asia
Prosperity Fund”
•
19 project partners from Meteorological
institutions and academia across Southeast
Asia
© Crown copyright Met Office
Raizan Rahmat (report coordinator), Boonlert Archevarahuprok
Chai Ping Kang, Chung Jing Xiang, David Hein,
Dodo Gunawan, Erasmo Buonomo, Grace Redmond,
Haji Sidup , Haji Sirabaha, Han Swe, Itesh Dash,
Jacqueline Lim See Yee, Jose Rizal, Kornrawee
Sitthichivapak, Kumarenthiran Subramaniam, Lalit Dagar,
Liew Ju Neng, Ling Leong Kwok, Lyhon Ho, Mai Van Khiem,
Matt Palmer, Monichoth So Im, Muhammad Yunus
Ahmad Mazuki, Ngai Sheau Tieh, Ngo Duc Thanh, Saiful
Azmi bin Haji Husain, Shamsuddin Shahid, Simon Tucker,
Srivatsan V. , Suppakorn Chinvanno, Thelma Acebes
Cinco, Thuy Tran Thanh, Vo Van Hoa, and Wittyi Soe
1 Thailand Meteorological Department (TMD)
2 Universiti Teknologi Malaysia (UTM)
3 Universiti Kebangsaan Malaysia (UKM)
4 Met Office Hadley Centre (MOHC)
5 Badan Meteorologi, Klimatologi, dan Geofisika, Indonesia (BMKG)
6 Brunei Darussalam Meteorological Department
7 Department of Meteorology and Hydrology, Myanmar (DMH)
8 Regional Integrated Multi-Hazard Early Warning System (RIMES,
Thailand)
9 Centre for Climate Research, Singapore (CCRS)
10 Malaysian Meteorological Department (MMD)
11 University Brunei Darussalam (UBD)
12 Vietnam Institute of Meteorology, Hydrology and Environment (IMHEN)
13 Department of Meteorology, Cambodia
14 University Malaya (UM)
15 Hanoi University of Science
16 Tropical Marine Science Institute (TMSI), National University of Singapore
17 Southeast Asia START Regional Center, Thailand
18 Philippines Atmospheric, Geophysical & Astronomical Services
Administration
19 National Center for Hydro-Meteorological Forecasting (NCHMF), Vietnam
SEACAM activities
© Crown copyright Met Office
Project activities
Inception
(May 2011June 2012)
• Agree need for
RCM
ensemble
• Design RCM
experiments
• Run DURIAN
expeiments
Planning
of Analysis
and
research
• Literature
review
• End user
survey
• Workshop
planning
Review
workshop
(Cambodia
August
2013)
• 5 days
• Analysis of
results in 4 key
areas
• Agree further
analysis
Final
workshop
(Singapore
February
2014)
Photos from the first review workshop in August 2013 in Phnom Penh, Cambodia.
© Crown copyright Met Office
• Review outputs
• Produce
guidance
• Final report
(April 2014)
Literature review
© Crown copyright Met Office
SEACAM Literature review
• Identified a lack of regional downscaling
modelling experiments in Southeast Asia
• Determined that the IPCC AR5 contained the
most comprehensive assessment of Southeast
Asian climate to date:
• AR5 indicates an increase in median temperature
(0.8 – 3.2 degrees) and moderate increase in
precipitation (1-8%)
• AR5 provides various regional variations and low
confidence assessment of circulations features
and climate extremes
• AR5 recommended that ensemble simulations
using additional Regional Climate Models driven
by other General Circulation Models should be
carried out
© Crown copyright Met Office
Intergovernmental Panel on Climate Change:
5th Assessment
Report
• Warming is very likely to continue with
substantial sub-regional variations.
• Medium confidence in an moderate
increase in rainfall over continental
Southeast Asia.
• To the south there is generally a drying although this may not be significant
relative to the natural decadal variations
in this region.
Temperature change for Southeast Asia (land areas only) during JuneAugust, from CMIP5 global models for the four RCP greenhouse gas
concentration levels.
• Strong regional variations are expected
because of terrain.
• Extreme heavy rainfall events are
projected to increase across the whole
region.
© Crown copyright Met Office
Precipitation change for S.outheast Asia (land areasonly) during AprilSeptember, from CMIP5 global models for the four RCP greenhouse gas
concentration levels. Results for precipitation are not clearly
distinguishable among the RCPs.
IPPC AR5 - Percentage changes in
temperature and precipitation for
Southeast Asia
Percentage changes in temperature and precipitation for Southeast Asia according to
model in the CMIP5 ensemble using RCP 4.5 (medium level greenhouse gas
concentrations).
© Crown copyright Met Office
A little about Regional Climate
Models…….
© Crown copyright Met Office
Resolution is important (example)
© Crown copyright Met Office
Resolution is important (example)
© Crown copyright Met Office
What is a Regional Climate Model
(RCM)?
• Covers a limited area of the Earth’s surface instead of the entire Earth
• Like Global Circulation Models (GCMs), RCMs contains representations of
the atmosphere, land and surface, and generate weather (and therefore
climate)
• Global models typically having a horizontal resolution of between 250 and
600 km
© Crown copyright Met Office
What is a Regional Climate Model
(RCM)?
• Their main advantage is that they allow for higher resolution climate
modelling - . In most cases, higher resolution = more useful and higher
quality information
• For example, the HadRM3P used in Precis can be run a two resolutions
50km and 25km
© Crown copyright Met Office
RCM’s: about the input data …
To obtain the global
influence on regional
climate RCMs take
input at the boundaries
(the edges of the
region) from GCMs or
reanalysis experiments
© Crown copyright Met Office
Limited area
models are driven
at the boundaries
by GCMs or
observations.
Assessment of End User needs
© Crown copyright Met Office
Stakeholder Engagement
• A survey was conducted
to identify potential user
needs of regional climate
projections
• 41 respondents from 25
agencies and research
institutes
• 54% hydrology sector,
41% ocean/marine, 39%
agricultural
• Most commonly required
climate variables were:
 Surface air temperature,
 Large scale precipitation,
 Surface winds
 Convective precipitation
© Crown copyright Met Office
Respondents' area of research or work. Note that percentages do not add up to 100% as some
respondents were involved in multi-faceted work areas.
• “... The need for
greater access to
various types of
climate information
products is
immense;
however, capacity
to interpret and
correctly use the
information is low.”
--comment from a
survey participant
© Crown copyright Met Office
Qualitative responses....notable
responses included.........
““... assess the impact of climate change on its water resources in order to manage it in sustainable way...”
“... alarming change in the past decade with respect to climate and its effects on agriculture, fisheries, drought.
But proper training in using/running these models are missing. Appreciate if you can conduct a training in
Southeast Asia ...”
“... sea level rise, coastal erosion, (flash) flooding, landslides, groundwater salinisation, increasing
temperature and warm nights/days ...”
“ ... impact on rainforest, haze...”
“ ... The land-use change, increasing temperatures and erratic pattern of weather ...”
“... agricultural, fisheries influence of climate change, food security, renewable energy ...”
“... we need adaptation strategy ...”
“... users need to truly know the meaning of climate projections so as to avoid making wrong decisions ...”
“... variables must demonstrate sufficient validity (bias correction, ability to reasonably replicate seasonality, etc.).
Merely uploading variables without sufficient user education is dangerous...”
“... change in rainfall patterns and inter-annual variability...”
“ ... change in tropical cyclone numbers, ENSO, NE Monsoon...”
“ ... cold spells, hot spells ...”
© Crown copyright Met Office
Model configuration and experimental
design
© Crown copyright Met Office
SEACAM Experimental Design
• Key considerations for design
of the experiments included:
• Regional model type used
• Domain and Resolution
• Driving data and emission
scenarios
• Time periods for analysis
• Observational data used in
evaluation
• SEACAM uses the HadRM3P
regional climate model used in
PRECIS
© Crown copyright Met Office
The Components of PRECIS
• PC version of the Hadley Centre’s HadRM3P Regional Climate
Model
• Horizontal resolution of 50km or 25km
• runs on the free Linux operating system
• Easy to use Graphical User Interface to set
up RCM experiments
• Data processing and analysis software
• Boundary conditions (input data)
• Training workshop and materials
• Technical Support (forum.precisrcm.com)
• The PRECIS web site and email address:
http://www.metoffice.gov.uk/precis
[email protected]
© Crown copyright Met Office
The domain initially proposed. Too Big!
• Careful consideration of
size and location of
domain is required
• SEACAM originally hoped
to use a domain that
included all ASEAN
countries
• However at this scale the
model runs take too long
to simulate
© Crown copyright Met Office
The Final Domain
• It became necessary to reduce
eastern extent of the domain to
135 degrees
• West Papua region of
Indonesia excluded in final
domain
• Proposed domain sent to
participants to ensure all grid
Resolution:
0.22°islands or
boxes containing
coastal *cities
represented as
(~25km
25km)
land points and not ocean
208 by 200 grid points
West Papua is
used in DURIAN experiments. Blue indicates that the grid box is an ocean grid box. Green indicates that the grid
excluded Domain
box is land. Political boundaries are marked with red lines. The darker rim between the edge of the picture and the orange
line is where the lateral boundary conditions (LBCs) are applied - output in this region is not analysed.
© Crown copyright Met Office
Driving data and emission scenarios
© Crown copyright Met Office
Driving data and emission scenarios
• IPCC Special Report on Emissions
Scenarios
• Provides representations of future
levels of substances that influence
the total energy/heat in the
atmosphere (e.g. greenhouse
gases) or which can affect heatcontributing atmospheric
substances (e.g. sulphur dioxide,
which forms sulphate aerosols)
• Four families/groups of scenarios
(A1, A2, B1, B2) were identified
• Scenarios are based on a coherent
and internally consistent set of
driving forces such as demographic
and socio-economic developments
© Crown copyright Met Office
The IPCC SRES scenarios
Four families identified:
- A1: globalization, emphasis on
human wealth.
Globalized, intensive (market
forces)
- A2: regionalization, emphasis
on human wealth.
Regional, intensive (clash of
civilizations)
- B1: globalization, emphasis on
sustainability and equity.
Globalized, extensive
(sustainable development)
- B2: regionalization, emphasis
on sustainability and equity.
Regional, extensive (mixed
green bag)
© Crown copyright Met Office
IPCC scenario A1B selected
In SEACAM, the A1B scenario
is used for all experiments.
SRES A1B assumes:
• Rapid economic growth.
• A global population that reaches 9
billion in 2050 and then gradually
declines.
• The quick spread of new and
efficient technologies.
• A convergent world - income and
way of life converge between
regions. Extensive social and
cultural interactions worldwide.
• A balanced emphasis on all
energy sources (i.e. between
fossil
and non-fossil).
© Crown copyright Met Office
Total global cumulative CO2 emissions (GtC) from 1990 to 2100
Driving data choice
• Models must be able to represent processes that are smaller than the area
of the model grid box (e.g. Clouds).
• Many of these small scale processes are important for climate, but we are
not certain about their exact values. For example, we know snowflakes fall at
between 0.5 and 1.0 metres per second.
• In a Perturbed Physics Ensemble (PPE), a large number of model
simulations are run that sample the ranges of these small scale processes.
• Five GCMs from the HadCM3Q PPE were
downscaled: Q0, Q3, Q10, Q11 and Q13. These
were selected as these members were deemed
suitable to span the range of future possible
outcomes.
• The Max Planck Institute ECHAM5 was
also selected in order to provide some
element of a multi-model ensemble (models
from more than one institute).
© Crown copyright Met Office
Observational data used for evaluation
• To evaluate surface air temperature and precipitation, two different gridded datasets
(APHRODITE and CRU-TS3) were used
• CRU-TS3: monthly mean data at a spatial resolution of a 0.5° square grid.
• APHRODITE: daily mean
precipitation and temperature
on a 0.25° square grid
• Inclusion of daily mean data allows
evaluation of extreme weather
events
• APHRODITE has a denser network
of stations than CRU areas
© Crown copyright Met Office
Time periods and approach to analysis
Time periods
• Global models in SEACAM were
downscaled by HadRM3P from
1950-2100
Analysis
The evaluations were done in four
broad categories (also termed
“work packages”) as follow:
Two key periods selected
1. Annual cycle of temperature and
precipitation
• 2071-2100 chosen to provide
strong detectable signal for any
possible climate changes on a
longer term
2. Mean temperature and
precipitation
• 2031-2060 chosen to inform
future policy relevant for
adaptation planning
© Crown copyright Met Office
3. Circulation patterns during the
Northeast and Southwest
Monsoons, and
4. Extreme precipitation and
temperature.
SEACAM results - Model Validation
© Crown copyright Met Office
Annual
temperature
cycle
(surface air
temperature at
1.5m) in the
RCM
experiments
plotted against
data from
historical
observations
© Crown copyright Met Office
Annual
temperature
cycle
(surface air
temperature at
1.5m)
versus
Observed
© Crown copyright Met Office
Annual temperature cycle assessment
• In general, all the downscaled climate model simulations are able to simulate the
temperature cycle well by capturing the observed peaks and dips across the year
• This is especially true for countries in the mainland Southeast Asia region which includes
Cambodia, Laos, Myanmar, Thailand, and Vietnam, as well as the Philippines
• A general feature of the RCM simulations over these regions, with the exception of the
Philippines, is the over-prediction (warm bias) of the warm months and the underprediction (cool bias) of the cooler months
• Countries closer to the equator such as Brunei, Indonesia, Malaysia, Singapore and
East Timor, the 6 RCM simulations show larger variability in comparison
• These simulations under-predict for Brunei (with respect to APHRODITE; but not in
relation to CRU) and over-predict for Singapore throughout the year (with respect to both
APHRODITE and CRU).
• As for Malaysia and Indonesia, the biases are GCM-dependent
• Overall, simulations generate biases that range from 1.0 to 2.0°C.
© Crown copyright Met Office
Annual precipitation
cycle
precipitation in the
RCM experiments
plotted against
data from historical
observations
© Crown copyright Met Office
Annual precipitation
cycle
“...the precipitation
cycles, or lack
thereof, for Brunei,
Malaysia and
Singapore are
poorly captured in
the RCM.”
© Crown copyright Met Office
Annual precipitation cycle assessment
• Annual rainfall cycles of the simulations are evaluated in a similar manner to the
temperature cycle against APHRODITE and CRU for precipitation
• For precipitation cycle, Southeast Asia can generally be divided into 3 sub-regions;
• North where the Southwest Monsoon is dominant during the middle part of the year
• Middle near the equator (i.e. Brunei, Malaysia and Singapore)
• South (i.e. Indonesia and East Timor) where the wet and dry seasons are mirror
images of the counterparts in the north.
• Comparison of RCM simulations in these 3 regions shows the simulations’ ability in
capturing large scale seasonal (monsoonal) rainfall in the northern latitudes of the region
stands out compared to capturing thunderstorm development in the lower latitudes
• The results for regions in the north and south are encouraging with the RCM simulations
able to pick out the seasonal rainfall maxima during the June-September season (JJAS) for
Cambodia, Laos, Myanmar, Philippines, Thailand and Vietnam and during the NovemberApril season (NDJFMA) for Indonesia and East Timor.
• In contrast, the precipitation cycles, or lack thereof, for Brunei, Malaysia and Singapore are
poorly captured in the RCM
© Crown copyright Met Office
Evaluation of seasonal spatial
distributions
• Evaluation of seasonal spatial distribution of temperature (and also rainfall) across the
domain provides additional information over the evaluation for annual cycle, which has been
spatially averaged. E.g. details of the location of errors, unlike the annual cycle.
• Analysis was made for
seasonal mean
temperature,
seasonal minimum and
maximum temperature
and seasonal spatial
rainfall distribution
Difference between the spatial
distribution of simulated and
observed DJF seasonal
maximum temperature (°C). Red
shades show warm biases , blue
shades show cool biases .
© Crown copyright Met Office
Evaluation of seasonal spatial
distribution of mean temperature
• Generally, the mean seasonal temperature biases range between ±4°C with considerable
spatial and seasonal variations.
• Biases largely positive (warm biases) during the MAM season especially over the mainland
while during the SON season, these generally negative (cool biases)
• Broadly consistent with findings from the evaluation of the annual cycle in temperature
• Model produced cooler climate
over the NW edge of the domain
(mountainous areas)
• Generally patterns in RCM
closely resemble ERA-40
Difference between the spatial
distribution of simulated and observed
June July August seasonal maximum
temperature (°C).
© Crown copyright Met Office
Evaluation of seasonal minimum and
maximum temperature
• Generally, spatial bias structure of the seasonal maximum temperature is similar to the
seasonal mean temperature
• Maximum temperature are generally colder near the equatorial region with consistently
largest bias (~4°C) over the west-coast of Sumatra.
• Mainland Southeast Asia region, simulations are warmer than the observations, except at
the centre of the region near ~20°N.
• Large warm biases are ALSO noted
over Mainland especially during MAM
• All the HadCM3Q simulations
produce biases patterns which resemble
that of the ERA-40 simulations
(suggesting biases are largely sourced
from process-representations of RCM)
Difference between the spatial
distribution of simulated and observed
March April May seasonal maximum
temperature (°C).
© Crown copyright Met Office
Evaluation of seasonal spatial rainfall
distribution
• Generally, the bias patterns of seasonal rainfall are much noisier than temperature.
• Simulations produce moderate wet biases of about 20-40% through the years, except over
the western part of the mainland Southeast Asia where the biases are largely negative.
• Close to the northern boundary of the simulations, the model errors remain large,
suggesting influence of the boundary forcing over this region with steep and complex terrain.
• Generally, RCM simulations show consistent biases patterns to ERA-40 simulations.
• ECHAM5 simulations produced larger
wet biases over Equatorial Maritime
Continent (Sumatra and Borneo)
during JJA
Difference between the spatial
distribution of simulated (and observed
(MAM) seasonal rainfall in % of individual
grids’ observed climatology. Red shades
show dry biases of simulations, while
blue shades show wet biases of
simulations.
© Crown copyright Met Office
Evaluation of the Southwest Summer
Monsoon
• The RCM reasonably simulates the spatial
pattern of the summer monsoon circulation.
• However, there is a clear systematic
positive bias in wind speed over the region in
all model simulations.
Land-only precipitation time-latitude cross-section averaged over longitudes
90E to 135E for APHRODITE, ERA-40 simulations, and 6 RCM simulations
850 hPa mean wind speed in m/s and direction during July, averaged over the
1971-2000 period for ERA-40 reanalysis and 6 RCM simulations
• The precipitation distributions over time in
the model runs compare well to APHRODITE
observations during the southwest summer
monsoon.
• All the models are slightly wetter than
observations.
© Crown copyright Met Office
Evaluation of the Northeast Winter
Monsoon
• Monsoon circulation - Comparison (ERA40 reanalysis and models) show monthly
mean of meridional wind (1970-2000) are in
general agreement, however intensities and
the durations differ.
• The upper level flow which indicates the
returning branch of the Hadley Cell is captured
in all the simulations
Meridional wind component in m/s at 850 hPa averaged over the latitudes
13S to 30N over the 1971-2000 period for ERA-40 simulations and 6 RCM
simulations. Red shades indicate southerly winds, blue shades indicate
northerly winds.
• Precipitation extreme - Spatial patterns of
circulation In general the main precipitation
pattern for the Northeast Monsoon is well
captured in the projections. The heavy
rainfall during December to March is
indicated in all the RCM simulations.
95th percentile of DJF daily rainfall amounts for APHRODITE, ERA-40
simulations, and 6 RCM simulations.
© Crown copyright Met Office
Evaluation of extreme rainfall indices
• Annual maximum one day rainfall (Rx1day) - RCM simulations showed consistent
spatial patterns of biases with the ERA-40 simulations and with each other, which suggests
biases can be reasonably attributed to the RCM itself
• Annual maximum consecutive five days rainfall (Rx5day) - All the RCM simulations
reproduce the basic pattern of variation estimated from the APHRODITE dataset over the
whole area. However, the five HadCM3Q simulations show consistent bias patterns to that
produced by the ERA-40 simulations, except for ECHAM5 simulations which show a more
intense rainfall in the lower latitudes.
• Annual maximum of consecutive dry days
There is generally good agreement in the spatial
patterns of maximum CDD (or the longest dry
spell)in the ERA-40 simulations and 6 RCM
simulations, with APHRODITE
• Trends in extreme rainfall indices - Overall,
the RCM was able to reproduce the sign of the
trends (for Rx1day and Rx5day) and inter-annual
variability of the observed rainfall indices
Trends in APHRODITE and ERA-40 simulations of rainfall indices.
© Crown copyright Met Office
SEACAM results - Assessment of Midand Long-term Climate Change
Projections
© Crown copyright Met Office
Assessment of Mid- and Long-term
Climate Change Projections
• Projections provided for two 30-year periods,
“mid-term” 2031-2060 and “long term” 2071-2100.
• Baseline period
simulations)
is
1970-2000
(as
for
• Projections assessed for:
 Annual temperature cycle
 Annual precipitation cycle
 Seasonal mean temperature
 Seasonal minimum temperature
 Seasonal maximum temperature
 Diurnal temperature range
 Seasonal mean rainfall
 Southwest Summer Monsoon
 Northeast Winter Monsoon
 Extreme rainfall indices
 Extreme temperature indices
 Five-year return level for maximum
daily temperature and precipitation
• A summary of key findings is provided. Full
analysis available in final report
© Crown copyright Met Office
Key findings: Climate Change
Projections
Attribute
Analysis / findings
Annual temperature cycle
In general, surface air temperature is expected to rise by 2°C by midcentury and 4°C by end-century, with some countries projected to be
experiencing an increase by up to 5°C.
In contrast , the projections for precipitation show a lot more variations
across countries and seasons.
Annual precipitation cycle
Seasonal mean temperature Seasonal mean temperature shows an increase of 3-5°C by the endcentury. By mid-century, the estimated temperature is 2-4 °C warmer than
the present day.
Seasonal minimum
Generally, the warming patterns of the seasonal minimum temperature
temperature
closely resemble that of the mean temperature.
Seasonal maximum
temperature
Diurnal temperature range
© Crown copyright Met Office
Generally, the warming patterns of the seasonal maximum temperature
(like the seasonal minimum temperature) closely resemble that of the
mean temperature except that the warming rate is higher in the maximum
temperature towards the end-century.
Generally, the minimum temperature (night time) warms faster than the
maximum temperature (day time) during the northern hemisphere winter,
except over northern Borneo and Peninsular Malaysia.
Key findings: Climate Change
Projections
Attribute
Seasonal mean rainfall
Analysis / findings
Projections show drier climate over the sea and wetter climate over land.
The land-sea contrast is more obvious towards the end-century. In all of
the HadCM3Q projections, drier climate is projected over most areas
during northern hemisphere winter except central mainland Southeast
Asia. Wetter climate was projected south of the equator in ECHAM5.
Southwest Summer Monsoon It can be seen that during the summer monsoon (JJAS), generally more
rainfall is projected in the northern part of the region (approximately from
20°N northward), whereas drier conditions are projected for the Maritime
Continent.
Northeast Winter Monsoon The scale of projected precipitation changes (e.g. increases over land) for
extremes during DJF is not as significant as the changes seen in the SW
summer monsoon.
Extreme rainfall indices
In general, Rx1day and Rx5day for the end-century are projected to
increase in areas north of the 15°N latitude.
Extreme temperature indices In general, a 1-3°C change is projected for most land regions of Southeast
Asia across all RCM projections for the mid-century and 3-5°C change for
the end-century.
Selected examples/figures from main report on following slides
© Crown copyright Met Office
Projected Changes – Annual
temperature
Summary of estimated mid-term and long-term projections of annual cycle temperature changes.
© Crown copyright Met Office
Seasonal Mean temperature
© Crown copyright
Future changes in seasonal mean temperature (°C) for end-century (20712100) relative to the baseline period (1971-2000) in JJA for ECHAM5 and
Met Office
HadCM3Q0, 3, 10, 11, 13 for the A1B scenario.
Seasonal rainfall change
© Crown copyright
The median values (from the 6 simulations) of the changes of the late
century seasonal rainfall. The areas where all the 6 simulations agree on
the change sign are hatched.
Met Office
Southwest Summer Monsoon Changes
End-century changes of average 850 hPa wind (vectors) and rainfall (scalar) for June, July, August and
September (left to right columns) compared to the baseline period (1971-2000). From top to bottom are the
different model projections, HadCM3Q10, 11, 13, 0, 3 and ECHAM5. Purple (green) shades indicate
© Crown(decrease)
copyright Met in
Office
increase
rainfall intensity during that month.
Changes in Annual Mean Maximum
Temperature (Tmax)
Projected changes in annual mean TXx (annual maximum day time temperature in °C), from
HadCM3Q0, 3, 10, 11, 13 and ECHAM5 for mid-century (top row) and end-century (bottom row).
© Crown copyright Met Office
Key findings - Summary
• Annual cycle change in temperature for the mid-term projections ranged around 2°C, and
for the long-term around 4°C. These were statistically significant projections with fairly
consistent changes registered across the year.
• In contrast to temperature projections, the projections for precipitation showed a lot more
variations across countries and seasons which lead to difficulties in interpretation of the
annual cycle plots.
• Unlike for temperature, changes in rainfall projections show large spatial and seasonal
variations. Generally, the projections show drier climate over the sea and wetter climate
over land.
• The land-sea contrast is more obvious towards the end of the century. In all of the
HadCM3Q projections, drier climate is projected over most areas during boreal winter
except central mainland Southeast Asia.
• However, wetter climate is projected south of the equator in ECHAM5. Generally, intermodel agreement is high except during winter (DJF).
• For extreme rainfall indices, Rx1day and Rx5day for the end-century are projected to
increase in areas north of the 15°N latitude.
© Crown copyright Met Office
Key findings - Summary
• All projections show an increase in Consecutive Dry Days (i.e. longer dry spells) south of
15°N latitude in both time periods. For these projections, model agreement tends to be
good.
• For extreme temperature indices, a 1-3°C change is projected for most land regions of
Southeast Asia across all RCM projections for the mid-century and a 3-5°C change for
the end of the century.
• The magnitudes of change for these two time periods are comparable across all four
indices (TXx, TNx, TMx, and TMn) considered.
© Crown copyright Met Office
Summary and next steps
© Crown copyright Met Office
Key findings - Benefits
• This project allowed national scientists to develop detailed climate scenarios for ASEAN
countries, further enabling them to understand and communicate the implications of
climate change within their country.
• It has delivered scientifically sound information, relevant to raising awareness and
understanding of climate change, developing national adaptation and mitigation plans.
• Therefore enabling further building the capability of ASEAN countries to engage more
effectively in international climate change forums.
• The project has already started to strengthen the ability of the ASEAN countries to
develop national adaptation and mitigation plans based on sound science and to engage
with more confidence and a better information base in the UNFCCC, ASEAN and related
forums.
• This aims to increase the number of countries that reflect climate change mainstreaming
in budgets, plans, profiles and national policy. For example, the implications for national
policy on agriculture and food security, forestry, land use change, disaster risk reduction,
health and water resources.
© Crown copyright Met Office
Next steps
• We now have an established network of partners to develop capability going forward.
• The DURIAN experiments generally perform well when compared to historic rainfall and
temperature data.
• A number of information gaps and potential improvements have been identified and it is
proposed that further work is undertaken to address these.
• A resourced programme of activity should be identified and partners should work together
to support ongoing development, maintenance and validation.
• The PRECIS model has a number of potential applications and can be used to inform
risk and vulnerability assessments and subsequent adaptation strategies for a number of
areas (for example agriculture, health, water management, disaster risk reduction and
early warning systems, transport etc).
• The SEACAM model should be utilised to inform and underpin regional strategies across
these areas where appropriate
© Crown copyright Met Office
Further information
www.precisrcm.com/rcct
© Crown copyright Met Office
For further information please contact:
Jane Sattary, International Development Manager
Met Office
[email protected]
© Crown copyright Met Office