Attribution of Climaterelated Events Nikos Christidis Peter A

Attribution of Climate-related Events Nikos Christidis, Peter A Stott, Nikos Christidis, Andrew Ciavarella, Fraser Scaife, Alberto EUCLEIA 1 Lott, General. Adam Assembly, ETH Zurich, 20 January. Arribas 2014 st © Crown copyright Met Office

Outline • Why? • Attribution of Climate-related Events • BAMS annual reports • EUCLEIA • Met Office Operational Attribution System • Communication © Crown copyright Met Office

January, 2014 © Crown copyright Met Office

This winter : the wettest winter in England Wales since 1766 © Crown copyright Met Office

“Human influence on the climate system is clear. ” AR 5 WGI SPM © Crown copyright Met Office

“It is virtually certain there will be more frequent hot extremes. ” “Extreme precipitation events over most of the mid latitudes and wet tropical regions will very likely become more intense and more frequent. ” © Crown copyright Met Office

Are all extreme weather and climate events due to humaninduced greenhouse gas emissions? Do we need to adapt to a greater frequency of such events in future or not? Misattribution, eg by attributing every extreme weather event to anthropogenic climate change could lead to poor adaptation decisions. Pakistan flooding, 2010 Queensland floods, Jan 2011 © Crown copyright Met Office Russian heatwave, 2010 Cold winters, UK, 2009, 2010 East African drought, 2011

Outline • • • © Crown copyright Met Office Why? Attribution of Climate-related Events BAMS annual reports EUCLEIA Met Office Operational Attribution System Communication

Attribution statements can be made about individual climate events by examining the changed probability European heatwave, 2003 Very likely more than doubled the probability (Stott et al) © Crown copyright Met Office Australian heatwave, 2013 Very likely increased the probability by more than 2. 5 times (Lewis et al)

Attribution of an individual climate event : application of probabilistic event attribution to the very hot European summer of 2003 Human influence very likely (>90% confidence level) at least doubled the probability of European summer temperatures as hot as 2003. Stott et al, Nature, 2004. © Crown copyright Met Office

Outline • • • © Crown copyright Met Office Motivation Attribution of Climate-related Events BAMS annual reports EUCLEIA Met Office Operational Attribution System Communication

Explaining extreme climate and weather events of the previous year from a climate perspective Tom Peterson, Martin Hoerling, Peter Stott, Stephanie Herring. • Report ground breaking in applying attribution science to recent extreme weather events. • Climate change has made some events more likely, some less likely • We do not see evidence for a strong human influence in all weather extremes. Natural variability also plays an important role • Inaugural report quickly became “most read” article in BAMS © Crown copyright Met Office

Explaining extreme climate and weather events of the previous year from a climate perspective Tom Peterson, Martin Hoerling, Peter Stott, Stephanie Herring. • Increase from 6 contributions last year to 19 this year • 18 different research groups, 12 extreme events • Some events have multiple different groups looking at them • About half the analyses found some evidence that anthropogenic climate change was a contributing factor • Natural climate variability a factor in all events © Crown copyright Met Office

Iberian winter drought Trigo et al, 2013 (BAMS report) A tendency towards a drier Mediterranean driven by anthropogenic emissions of greenhouse gases and aerosols, although modulated by the NAO phase. Expect 3 winter droughts per century in current decade compared to expectation of 2. 4 winter droughts per century in 1960 s © Crown copyright Met Office

Extreme rainfall, New Zealand Dean et al, 2013 (BAMS report) An atmospheric river brought very high levels of moisture, extreme rainfall and landslides to the mid-latitude location of Golden Bay, New Zealand on 14 December, 2011. • Total moisture available for precipitation in this event has increased by 1% to 5% as a result of the emission of greenhouse gases. • Models show an increase in the frequency of such events of between 8 and 32%. © Crown copyright Met Office • Predominantly due to a thermodynamic response.

Outline • Why? • Attribution of Climate-related Events • BAMS annual reports • EUCLEIA • Met Office Operational Attribution System • Communication © Crown copyright Met Office

EUropean CLimate and weather Events : Interpretation and Attribution • Heatwaves • Cold spells • Floods • Droughts • Storm surges • UK Met Office • ETH Zurich • CNRS • University of Edinburgh • IC 3 • DMI • KNMI • University of Reading • University of Oxford • HZG • Université de Versailles © Crown copyright Met Office

Main limitations at present • Event attribution studies attempted only for a small number of specific cases • There can be apparently conflicting results for the same event • Attribution results rely heavily on models whose limitations may not be fully understood • Biases and poor sampling in observational datasets • The requirements for such attribution studies are not well understood © Crown copyright Met Office

Project Objectives • Derive the requirements of targeted user groups for attribution products and demonstrate value to these users of attribution products developed under EUCLEIA (WP 4) • Develop experimental designs and clear ways of framing attribution studies (WP 5) • Develop methodology for representing level of confidence in attribution results (WP 6) • Demonstrate utility of attribution system on a set of test cases (WP 7) • Produce traceable and consistent attribution assessments (WP 8) • Fast-track in immediate aftermath • Seasonal basis to stakeholder groups • Annual basis to BAMS attribution report • Coordination between the FP 7 Copernicus climate change projects (WP 9) © Crown copyright Met Office

5 FP 7 -SPACE projects prepare for a Copernicus Climate Change Service: 1. 2. 3. 4. 5. ERA-CLIM 2 (global reanalysis, Dick Dee, ECMWF) UERRA (regional reanalysis, Per Unden, SMHI) QA 4 ECV (air quality ECVs, Folkert Boersma, KNMI) CLIPC (climate impact platform, Martin Juckes, STFC) EUCLEIA (event attribution, Peter Stott, Met Office) The overarching WP will co-ordinate information exchange between these projects and the outside world, including relevant Commission DGs and joint stakeholders

Outline • Why? • Attribution of Climate-related Events • BAMS annual reports • EUCLEIA • Met Office Operational Attribution System • Communication © Crown copyright Met Office

Coupled model assessments Annual mean T distributions in Giorgi regions Red: with human influences Green: without human influences Vertical black lines: Max temperature in CRUTEM 3 © Crown copyright Met Office

Attribution of Climate-related Events (ACE) Development of the Hadley Centre near-real time attribution system ACE approach: • Generate large ensembles (perturbing physics parameters), running our model with observed SSTs and external forcings. • Generate a second ensemble without the human influence. An estimate of the anthropogenic change in the SSTs is subtracted from the observations. Only natural forcings are included. Change in the likelihood of the event given certain modes of internal variability Hadley Centre near-real time attribution system Had. GEM 3 -A, N 96 L 38. (Currently upgrading to N 216 L 85) © Crown copyright Met Office

Patterns of the change in the SST January Had. GEM 1 Had. CM 3 Had. GEM 2 -ES © Crown copyright Met Office April July October

Validation Dec-Jan UK Temperature 1960 -2010 Timeseries Distributions Power Spectrum Red Lines: NCEP/NCAR Reanalysis Black Lines: 5 Model runs with ALL forcings and observed SSTs & SI © Crown copyright Met Office

Validation July Rainfall in Pakistan 1960 -2010 Timeseries Distributions Power Spectrum Red Lines: NCEP/NCAR Reanalysis Black Lines: 5 Model runs with ALL forcings and observed SSTs & SI © Crown copyright Met Office

Greater Horn Short Rains (SON) Temperature and Rainfall Above median temperature Below median rainfall © Crown copyright Met Office

Greater Horn Long Rains (MAM) Temperature and Rainfall Above median temperature Below median rainfall © Crown copyright Met Office

Next steps 1 st Year: • Finalise the model upgrade • Produce long simulations (1960 -present) • Development of coupled model response techniques • Update pre-computed FAR tables 2 nd Year: • Generate ensembles for targeted test cases • Investigate extremes for the BAMS report • Development and application of fast-track response techniques 3 rd Year: • Delivery of an operational event attribution system • Delivery of assessments for the BAMS report • Reports on fast-track response assessments and the new attribution service © Crown copyright Met Office

Human influence on the climate system is clear • Better understanding of the changing risks of extreme weather will help people cope with the effects of anthropogenic climate change. • Extreme weather and seasons result from the interplay of natural climate variability and anthropogenic climate change. • At the Met Office we are developing an “operational attribution” system to assess the risks of such extremes on a regular basis. • A new annual report provides puts extreme weather from last year in different regions of the world into the context of natural climate variability and anthropogenic climate change. © Crown copyright Met Office

Any questions? © Crown copyright Met Office

Australia January 2013 © Crown copyright Met Office

Hobart, Tasmania, 4 th January 2013 © Crown copyright Met Office

Dunalley, 4 th January 2013 © Crown copyright Met Office

AR 5 WGI SPM approved line-by-line in Stockholm by 110 governments © Crown copyright

Extreme rainfall, New Zealand Dean et al, 2013 (BAMS report) An atmospheric river brought very high levels of moisture, extreme rainfall and landslides to the mid-latitude location of Golden Bay, New Zealand on 14 December, 2011. • Total moisture available for precipitation in this event has increased by 1% to 5% as a result of the emission of greenhouse gases. • Models show an increase in the frequency of such events of between 8 and 32%. © Crown copyright Met Office • Predominantly due to a thermodynamic response.