CMIP5 is meant to provide a framework for coordinated climate change experiments for the next five years and thus includes simulations for assessment in the AR5 as well as others that extend beyond the AR5. CMIP5 is not, however, meant to be comprehensive; it cannot possibly include all the different model intercomparison activities
0.25 degree atmosphere/ocean; 3-hour coupling; heat and momentum fluxes to the ocean; NEMO returns SST.
Climate Modelling for Ireland -Dr Ray McGrath, Met Eireann
Climate Modelling for Ireland Ray McGrath, Met Éireann EPA Climate Change Conference, 30 June 2010, Aviva Stadium, Dublin
Outline <ul><li>Brief history of climate modelling in Met Éireann. </li></ul><ul><li>Move from Regional to Global modelling </li></ul><ul><li>Current modelling work and future plans. </li></ul>
History <ul><li>Met É ireann monitors the Irish climate – maintains a network of observing stations. </li></ul><ul><li>Climate & Observations Division analyses historical data and regularly issues reports on case studies, trends in the climate, etc. </li></ul><ul><li>Involved in Numerical Weather Prediction (NWP) for many decades – models used to produce our operational weather forecasts. </li></ul><ul><li>Climate modelling, in spite of synergy with NWP, arrived late… </li></ul>
History <ul><li>Community Climate Change Consortium for Ireland (C4I) project launched in 2003. </li></ul><ul><li>Externally funded (EPA, SEAI, HEA/CosmoGrid) the joint Met Éireann /UCD project focused on Regional Climate Modelling . </li></ul><ul><li>Finished in 2007. </li></ul><ul><li>Provided future scenarios (changes in mean climate, extremes, flooding, waves, storm surge). </li></ul>
History <ul><li>Impact of C4I project </li></ul><ul><li>Climate modelling/research became a core activity for Met Éireann - National Climate Change Strategy (2007-2012). </li></ul><ul><li>Partner in ENSEMBLES project (2004-2009). </li></ul><ul><li>EPA Fellowships (2007- ): weather extremes (analysis of Irish precipitation); new coupled atmosphere/ocean Regional Climate Model developed; air quality modelling. </li></ul>
Average number of days per month with ground frost (1961-1990) EU ENSEMBLES Project – European Climate (2004 – 2009) ENSEMBLES : EU FP6-funded project involving 66 institutes (including Met Éireann) from 20 countries. Goal: Run ensembles of different climate models to sample uncertainties Assess reliability of models for historical periods Working towards a probabilistic framework for projections of climate change Focus: season, decadal, centennial; global, regional and local. Applications of societal relevance ENSEMBLES has produced ~25 regional climate simulation datasets for Europe; also 7 global simulations.
EU ENSEMBLES Project – European Climate (Sep 2004 – Dec 2009) C4I contribution: 2 centennial simulations over Europe
EU ENSEMBLES Project – European Climate (Sep 2004 – Dec 2009) Source: van der Linden P., and J.F.B. Mitchell (eds.) 2009: ENEMBLES: Climate Change and its Impacts: Summary of research and results from the ENSEMBLES project.
Flood or drought? Change in river Suck discharge: 2021-2060 relative to 1961-2000. Based on 13 ENSEMBLES simulations. EU ENSEMBLES Project: local applications
ENSEMBLES/C4I: expected change (%) in seasonal rainfall 2021-2050 compared with 1961-1990 (mean of 20 climate model simulations) Winter Spring Summer Autumn But large spread between the simulations… too much uncertainty
ENSEMBLES/C4I RCM simulations have highlighted the uncertainty in precipitation forecasts. Example: expected winter change (%) 2021-2050 compared with 1961-1990 20 member ensemble 12 member ensemble Spread reflects model differences. GHG emission uncertainty is not included (same in all simulations) .
ENSEMBLES: expected change (%) in seasonal rainfall 2071-2100 compared with 1961-1990 (mean of 12 climate model simulations) Winter Spring Summer Autumn
Lessons learned from C4I/ENSEMBLES <ul><li>Regional climate modelling introduces additional uncertainty – errors in GCMs compounded with errors in RCM models. </li></ul><ul><li>Local model cannot “correct” fundamental GCM errors. </li></ul><ul><li>Logistical difficulties in accessing GCM outputs (at required spatial/temporal resolution) for downscaling </li></ul><ul><li>Solution: move to a GCM </li></ul><ul><li>Still need RCM for local downscaling – not (yet!) feasible to run GCM at sufficient resolution to model local effects. </li></ul><ul><li>Running own GCM provides autonomy/flexibility in designing experiments. </li></ul>
In 2007 Met Éireann /ICHEC / UCD became partners in EC-Earth , a consortium of European Weather Services and Institutes (~23 in total) established to develop and use a common global Earth System Model: atmosphere, ocean/sea-ice, land, atmospheric chemistry, Carbon cycle – unified within a single package. EC-Earth modelling work will feed into the next IPCC Assessment Report Climate modelling in Met Éireann: moving to the global scale
<ul><li>Consortium influenced by concept of ‘seamless prediction’ in operational weather forecasting - same physical principles (but different processes acting on different scales) for weather and climate. </li></ul><ul><li>merging of capability in short-range, seasonal and climate (decadal/centennial) forecasting areas i.e. single system. </li></ul><ul><li>System based on the ECMWF seasonal prediction system – strong synergy with operational forecasting system in Met É ireann </li></ul>Background to EC-Earth
Joint EC-Earth and ECMWF seasonal forecast components Planned EC-Earth components For CMIP5: T159L62, 1 deg Ocean On 6 platforms OASIS Atmosphere GCM: IFS Ocean GCM: NEMO Sea-ice:LIM2/3 Land: IFS H-tessel Atmospheric Chemistry and aerosols: TM5 Vegetation: LPJ Marine ecosystem: PISCES New EC-Earth components
Timeline Testing/tuning –> Spin-up runs –> CMIP5 runs An ensemble of simulations planned – distributed among partners Test phase Historical run + Initialized Decadal hindcasts ~900 yrs 1850 2005 Pre-industrial Spin-up 2 independent simulations by Met É ireann / ICHEC and DMI/KNMI. RCP runs 2100
CMIP5 - Coupled Model Intercomparison Project Phase 5 <ul><li>Outputs from EC-Earth (~50 terabytes of core data) will be delivered to CMPI5 for assessment by IPCC </li></ul><ul><li>CMIP is a standard experimental protocol for studying the output of coupled ocean-atmosphere general circulation models (GCMs) … in support of climate model diagnosis, validation, intercomparison, documentation and data access. </li></ul><ul><li>The purpose… is to address outstanding scientific questions that arose as part of the IPCC/AR4 … and to provide estimates of future climate change. </li></ul>
CMIP5 - Coupled Model Intercomparison Project Phase 5 <ul><li>CMIP5 Goals: </li></ul><ul><li>Evaluate realism of the models. </li></ul><ul><li>Provide projections of future climate change on two time scales, “near term” (out to about 2035) and “long term” (out to 2100+) </li></ul><ul><li>Understand some of the factors responsible for differences in model projections, including quantifying some key feedbacks such as those involving clouds and the carbon cycle. </li></ul>“ Long-Term” (century & longer, 2100/2300) TIER 1 TIER 2 CORE “ realistic” diagnostic “ Near-Term” (decadal) (initialized ocean state) prediction & predictability CORE TIER 1
CMIP5 - Coupled Model Intercomparison Project Phase 5 <ul><li>CMIP5 Schedule: </li></ul><ul><li>July 2010: First model output is expected to be available for analysis. </li></ul><ul><li>July 31, 2012: By this date papers must be submitted for publication to be eligible for assessment by WG1. </li></ul><ul><li>March 15, 2013 : By this date papers cited by WG1 must be published or accepted. </li></ul><ul><li>The IPCC’s AR5 is scheduled to be published in September 2013 . </li></ul>
Irish partner ICHEC: Designated Earth System Grid ‘Data Node’ through e-INIS (the Irish National e-Infrastructure) PCMDI - Program for Climate Model Diagnosis and Intercomparison BADC - British Atmospheric Data Centre ICHEC – Irish Centre for High-End Computing ICHEC
Met É ireann / ICHEC schedule of EC-Earth simulations: CMIP5 Commitment <ul><li>900 year spin-up of coupled atmosphere/ocean model (pre-Industrial GHG) ~ completed. </li></ul><ul><li>Provides different launch points (700, 725… 825 years) for Industrial run (1850-2005) with prescribed GHG and aerosols for EC-Earth partners. </li></ul><ul><li>Future runs (2006-2100): </li></ul><ul><li>3 x simulations with prescribed GHG and aerosols. </li></ul><ul><li>Decadal simulations (atmosphere only with initialised ocean): </li></ul><ul><li>10/30-year simulations from 1960, 1965,…2005 launch points. </li></ul><ul><li>30-year simulations from 1960, 1980, 2005 launch points. </li></ul>
Met É ireann / ICHEC schedule of EC-Earth simulations for local use <ul><li>High-resolution decadal simulations (atmosphere only) from initialised ocean analysis: </li></ul><ul><li>1990-2010 </li></ul><ul><li>2010-2030 </li></ul><ul><li>Target horizontal resolution ~25km with either 1° or 0.25° ocean (depending on computer resources). </li></ul>How Ireland/UK elevations “appear” on the 25km global EC-Earth grid
EC-Earth spin-up event: change in the meridional overturning circulation (MOC) after ~500 years and recovery SST Anomaly during MOC decrease Cold spell lasted ~50 years, lagged MOC decline by ~20 years… An unforced event simulated by the model (no influx of fresh water into the N. Atlantic, etc.)
EC-Earth spin-up event: changes in the average SST (anomalies) over the Atlantic SST oscillations ~40 year cycle – similar to observed Atlantic Multidecadal Oscillation. Cold dip an echo of 8.2kyr event in Holocene? Relative year
Data source: http://www.ncdc.noaa.gov/paleo/pubs/alley2000/alley2000.htm 8.2kyr cold event Central Greenland reconstructed temperature 8.2kyr event linked with drainage of glacial lake into N. Atlantic… but may be simply ‘natural variability’ as seen in EC-Earth simulation
Regional models not neglected: coupled atmosphere-ocean regional climate model (RCA_NEMO) developed System evaluated by simulating climate over the above area 1961-1990 and comparing against ERA-40 reanalysis data.
<ul><li>First step in building an integrated system to complement the EC-EARTH GCM </li></ul><ul><li>Will upgrade to next generation Harmonie atmospheric model (non-hydrostatic model): </li></ul><ul><li>HARMONIE will be implemented for operational forecasting in Met Éireann in 2010. </li></ul><ul><li>Model has strong synergy with ECMWF IFS model (e.g. common physics modules with EC-EARTH). </li></ul><ul><li>Feedback between operational (day-to-day) regional forecasting and climate modelling will enhance development. </li></ul>Coupled atmosphere-ocean regional climate model
Greenland 5 km DEM, Ice Thickness, and Bedrock Elevation Grids <ul><li>EC-Earth used to investigate </li></ul><ul><li>Impact of declining Arctic sea ice cover on climate </li></ul><ul><li>Dynamical impact of declining Greenland ice sheet (reduced to bedrock). </li></ul><ul><li>Sensitivity of atmospheric ‘blocking’ to model horizontal resolution </li></ul>
Future Plans EC-Earth outputs (particularly the decadal simulations) will be used to investigate changes in weather extremes e.g. rainfall. Selected outputs downscaled to 2-5km grid. Outputs will be freely available to drive applications to study local impacts of climate change (flooding, storm surge, coastal erosion, renewable energy – wind/wave, agriculture, forestry,… Empirical Orthogonal Function (EOF) analysis of annual Irish rainfall (1941-2009). The time evolution of the Principle Component for the first mode is shown with a 5-year running trend line. Where is the climate heading?
Acknowledgements <ul><li>Tido Semmler (Met É ireann) </li></ul><ul><li>Emily Gleeson (Met É ireann) </li></ul><ul><li>Alastair McKinstry (ICHEC) </li></ul><ul><li>Shiyu Wang (ICHEC/Met É ireann– EPA Fellowship) </li></ul><ul><li>Seamus Walsh (Met É ireann) </li></ul>