Download Day3_CFM_Downscaling Techniques & Scenario

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts
no text concepts found
Transcript
Downscaling Techniques and Regional
Climate Modelling
PRECIS Workshop, MMD, KL, November 2012
© Crown copyright Met Office
Objectives of the session
• To review the different methods of obtaining finescale climate information from global climate
models (GCMs), with an emphasis on regional
climate models (RCMs).
© Crown copyright Met Office
Outline
1. Downscaling techniques
- Statistical methods
- Dynamical methods
2. Suitability of downscaling techniques
3. Use of regional climate models
© Crown copyright Met Office
What are downscaling
techniques?
© Crown copyright Met Office
Climate downscaling
• Techniques which allow fine scale information to be derived
from GCM output.
• Smaller scale climate results from an interaction between
global climate and local physiographic details
• Impact assessors need regional detail to assess
vulnerability and possible adaptation strategies
• AOGCM projections lack that regional detail due to coarse
spatial resolution
• Downscaling for climate change assessment differs from
downscaling of seasonal climate prediction
© Crown copyright Met Office
Going from global to local climate
© Crown copyright Met Office
Classification
• Statistical
• Transfer functions
• Weather generators
• Weather typing
• Dynamical
• High resolution and variable resolution AGCMs
• Regional Climate Models
© Crown copyright Met Office
Statistical or empirical techniques
1) Find a function F, derived from statistical relationships
between observations of fine-scale and large scale
variables:
fine scale value = F (large-scale variables)
2) Use F to find future fine-scale values from future large-scale
variables:
future fine-scale value = F (AOGCM large-scale variables)
© Crown copyright Met Office
Stretched grid AGCMs
The spatial resolution here is equivalent
to a grid mesh of approximately 30 km.
The spatial resolution is progressively
relaxed towards the antipode (near NewZealand).
© Crown copyright Met Office
Regional climate models
Courtesy of H. von Storch
Regional atmospheric modelling: nesting into a global state
© Crown copyright Met Office
Suitability of regionalisation
techniques
© Crown copyright Met Office
Suitability of regionalisation techniques
Method
Strengths
Weaknesses
Statistical
 High resolution
 Computationally
cheap
• Dependent on empirical relationship
Stretched grid
AGCMs
Regional
models
 High (very high)
resolution
 Can represent
extremes
 Physically based
 Many variables
 RCM: easily
relocatable
derived from present-day climate
• Dependent on long time-series and
good quality historical data
• Few variable available
• Not easily relocatable
• Dependent on surface boundary
conditions from coupled model
• Computationally expensive
• Have to parameterize across scales
• Dependent on driving model and
surface boundary conditions
• Possible lack of two-way nesting
• Have
to parameterize across scales
(
scales
© Crown copyright Met Office
)
Regional modelling vs. statistical
downscaling
• The major theoretical weakness of statistical downscaling
methods is that these empirically-based techniques cannot
account for possible systematic changes in regional forcing
conditions or feedback processes.
• The possibility of tailoring the statistical model to the
requested regional or local information is a distinct
advantage. However, it has the drawback that a systematic
assessment of the uncertainty of this type of technique, as
well as a comparison with other techniques, is difficult and
may need to be carried out on a case-by-case basis.
© Crown copyright Met Office
Boundary conditions
© Crown copyright Met Office
One way nesting methodology
• A RCM is a limited area model (LAM),
similar to those used in numerical weather
prediction (NWP), i.e. short term weather
forecasting
• LAMs are driven at the boundaries by GCM
or observed data
• Lateral (side) and bottom (sea surface)
• LAMs are highly dependent on their
boundary conditions and can not exist
without them
© Crown copyright Met Office
Lateral boundary conditions
• LBCs = Meteorological boundary conditions at the lateral (side)
boundaries of the RCM domain
• They constrain the RCM throughout its simulation
• Provide the information the RCM needs from outside its domain
• Data come from a GCM or observations
• Lateral boundary condition variables
• Temperature
• Water
• Pressure
• Aerosols
© Crown copyright Met Office
LBC variables
• Wind
LBC variables
LBC variables
Sea surface boundary conditions
• Two methods of supplying SST and sea ice:
• Using outputs from a coupled AOGCM
• Need good quality simulation of SST and sea ice in model
• Necessary for future simulations
• Using observed values
• Useful for the present-day simulation.
• For future climate need add changes in SST and ice from a
coupled GCM to the observed values – complicated
© Crown copyright Met Office
Added value of RCMs
© Crown copyright Met Office
RCMs simulate current climate more
realistically
Patterns of present-day mean winter precipitation over Great Britain
© Crown copyright Met Office
RCMs simulate
current climate
more realistically
Patterns of present-day
winter precipitation
over Great Britain
© Crown copyright Met Office
Represent smaller islands
Projected changes in summer surface air temperature between
present day and the end of the 21st century.
© Crown copyright Met Office
Predict climate change with more detail
Projected changes in winter precipitation between now and 2080s.
© Crown copyright Met Office
Simulate and predict changes in
extremes more realistically
Frequency of winter days over the Alps with different daily rainfall
thresholds.
© Crown copyright Met Office
Simulate cyclones and hurricanes
A tropical cyclone is evident in the RCM (right) but not in the GCM
© Crown copyright Met Office
Summary
• Downscaling techniques are used to add fine
scale details to a GCM projection
• Several methods are available with their own
strengths and weaknesses
• PRECIS is a physically-based and
computationally accessible regional climate
model for downscaling GCM projections
© Crown copyright Met Office
Contents
• What are Climate Scenarios?
• Types of Climate Scenarios
• Examples of Impacts studies that have
used PRECIS
• An impacts case study – why the method
for constructing a scenario is important.
© Crown copyright Met Office
Climate scenarios
• Climate scenario
“A scenario is a coherent, internally consistent and
plausible description of a possible future state of the
world. It is not a forecast; rather, each scenario is one
alternative image of how the future can unfold.”
Key point 1: Internal consistency
Socio-Economic scenario → Emissions scenario → Climate scenario
Key point 2: Scenarios are NOT the same as
‘predictions’: we can have many plausible scenarios.
© Crown copyright Met Office
Types of climate scenarios
• Incremental scenarios for sensitivity studies
• Analogue scenarios
• Scenarios based on outputs from Climate Models
© Crown copyright Met Office
1. Incremental scenarios
• Particular climatic elements are changed incrementally
by plausible though arbitrary amounts.
• Use for testing system sensitivity
• Use for identifying critical thresholds or discontinuities
in climate
• Potentially leads to unrealistic scenarios
• Not related to anthropogenic emissions
© Crown copyright Met Office
2. Analogue scenarios
• Identify recorded climate regimes which may resemble the
future climate in a given region.
• Spatial analogues
• Temporal analogues
• Palaeoclimatic
• Instrumental
• Not related to anthropogenic emissions
• Often physically implausible
© Crown copyright Met Office
3. Scenarios based on outputs from
Climate Models
• Coupled Atmosphere-Ocean Global Climate Models (AOGCMs)
• Coarse resolution, and often have large biases
• Based on physics
• Internally consistent
• Dynamically downscaled AOGCMs
• High resolution GCMS (e.g. PRECIS)
• Require large computer resources
• Can inherit biases from AOGCM
• Statistically downscaled AOGCMs
• Statistical methods are based on current climate and trained on short-term
variability
• Difficult to develop internally consistent climate variables
© Crown copyright Met Office
Stages required to develop climate change
scenarios
© Crown copyright Met Office
More adverse than beneficial impacts on
ecological and socioeconomic systems
are projected
© Crown copyright Met Office
Impacts Assessment
• Evaluation of the detrimental and beneficial
consequences of climate change on natural and
human systems.
• Impacts models require climate scenarios as inputs.
• The impact of the climate change is determined by
contrasting the effect of the observed/baseline
climate with that of the future climate (scenario) on
the exposure unit
© Crown copyright Met Office
Using climate model scenarios with
impact models
Climate change impact = System under future climate – system under current climate (baseline)
Baseline
Observations
Future
Observations and
climate model change
factor
Appropriate?
 If we have sufficient obs.
data to drive our impact
model, yes.
 If we see only systematic
biases in our model
simulations, yes.
Climate model baseline
Climate model future
 If our climate model
baseline is realistic, yes.
Observations
Climate model future
 Not normally. Only if
model baseline is very
similar to observed –
otherwise the result is a
combination of climate
model error and climate
change impact.
© Crown copyright Met Office
One approach to combining climate
observations and simulations
• If, as a result of systematic biases in the GCM/RCM simulations, the
impact baseline is unrealistic then a simple approach is to apply the
model change factor rather than the model output directly
• Model change factor = Model future – Model baseline: or
• Model change factor =(model future / model baseline) *100
• We can then add the change factor to an observed record to get a
future scenario with the bias seen in the baseline removed
• Future climate scenario = Observed + model change factor: or
• Future climate scenario = Observed * model change factor (%)
• This approach may provide impact results which are more reasonable
but the simple change factor applied does not account for changes in
variability and may result in inconsistent future climates
© Crown copyright Met Office
Example: Modelling Impacts of
climate change on agriculture
Use of PRECIS and the crop model CERES to simulate yield changes per
hectare of three grain crops (rice, wheat, maize) in China when applying
one future climate scenario and a representation of CO2 fertilization
© Crown copyright Met Office
Xiong et al, 2007, Climate change and critical thresholds in China’s food
security, Climatic Change, 81:205-221
Example: modelling climate
change impacts on Hydrology
2020s
• Change in water stress in the
Ganges-Brahmaputra-Meghna
Basin derived using the Global
Water Availability Assessment
(GWAVA) model
2050s
(CLASIC project – work with
CEGIS)
© Crown copyright Met Office
Example: Modelling Storm
Surge under climate scenarios
Simulated tropical cyclone and resulting storm surge.
Produced using PRECIS and POLCOM storm surge model
SLR projections from GCM
© Crown copyright Met Office
The PRECIS modelling system:
An impacts case-study
© Crown copyright Met Office
Climate change scenarios from a
recent climate model: estimating
change in runoff in southern Africa
• Nigel Arnell
• (Dept. of Geography, University of Southampton, U.K.)
• Debbie Hudson and Richard Jones
• (Hadley Centre for Climate Prediction and Research)
© Crown copyright Met Office
Methods
• Runoff: calculated from water balance
• runoff = precipitation – evaporation – absorption by
soil
• Two sets of models - a climate model and a runoff model
• Baseline climate → Run-off model → Baseline Run-off
• Future climate → Run-off model → Future Run-off
• Compares different methods of constructing future
climate scenario
© Crown copyright Met Office
Mean temperature and rainfall
Average annual rainfall is systematically overestimated by the model
© Crown copyright Met Office
Rainfall variability is accurately represented
by the model
© Crown copyright Met Office
Some different methods for CCS
construction
Constructing the baseline and future timeseries of
data required by the runoff model. For example:
BASELINE
FUTURE CLIMATE
SCENARIO
Mean
Variance
Mean
Variance
Observed Observed Observed + Observed
model
difference
Simulated Simulated Simulated Simulated
…these are just some of the possibilities
© Crown copyright Met Office
© Crown copyright Met Office
The best method CCS
construction in this case?
BASELINE
Mean
Observed
Simulated
© Crown copyright Met Office
FUTURE CLIMATE
SCENARIO
Variance
Mean
Variance
Observed Observed + Observed
model
difference
Simulated Simulated Simulated
Summary
• There are several techniques for producing future
climate information
• Only climate model based climate change predictions
can be used for providing climate scenarios which are
plausible and self consistent
• Even when using a single climate model (or family of
models) there are many different ways to provide
climate change information for impacts studies
• The method of climate scenario construction adds a
further uncertainty in assessing impacts of climate
change
© Crown copyright Met Office
Questions
© Crown copyright Met Office