Based on data to 2000, 20 years of additional data could halve uncertainty in future warming © Crown copyright Met Office Stott and Kettleborough, 2002

HadCRUT3 Global Temperature Record

© Crown copyright Met Office Test β=(β ant, β nat ) β≠0 (detection) β=1 (consistent) Total least squares regression in reduced dimension space Best Linear Unbiased Estimator Evaluate goodness of fit Climate model simulations Anthropogenic Natural

© Crown copyright Met Office Robust quantification of contributors to past temperature change enables quantification of likely future rates of warming (ASK) Global mean NH - SH Land - oceanMTG Stott et al, 2006

© Crown copyright Met Office Robust quantification of contributors to past temperature change enables quantification of likely future rates of warming (ASK) Global mean NH - SH Land - oceanMTG

© Crown copyright Met Office Robust quantification of contributors to past temperature change enables quantification of likely future rates of warming (ASK) Global mean NH - SH Land - oceanMTG Frame et al, 2006

© Crown copyright Met Office Robust quantification of contributors to past temperature change enables quantification of likely future rates of warming (ASK) Global mean NH - SH Land - oceanMTG

© Crown copyright Met Office The two extremes proposed by Andreae et al (2005) as consistent with strong and weak present day aerosol cooling are very unlikely Andreae et al, Nature, Stott and Forest, 2007 Phil. Trans. Roy. Soc.; Stott et al, Tellus, 2007.

© Crown copyright Met Office Stott and Kettleborough, 2002 Perfect model post 2000 Perfect data Perfect world post 2000

How has the “very likely” range of attributable ghg warming and of transient climate response changed with 10 more years of data ? What is the effect of a greater wealth of evidence on our confidence in that assessment ?

© Crown copyright Met Office

Global temperatures are evolving as predicted in response to human influence Global temperature response to greenhouse gases and aerosols Solid: climate model simulation (HadCM2) Dashed: recalibrated prediction using data to August 1996 (Allen, Stott, Mitchell, Schnur, Delworth, 2000) Observed decadal mean temperature September 1999 to August 2009 inclusive

© Crown copyright Met Office Human-induced warming has been detected on every continent (except Antarctica) AR4

© Crown copyright Met Office Detection of human influence on temperatures at sub-continental scales Hot summers that during the period were experienced 1 year in 10 are now being experienced at least 3 years in 10. Jones et al, 2008 Source Gareth Jones, Hadley Centre

© Crown copyright Met Office Detection and attribution : from global to regional Wealth of evidence now available shows there is an increasingly remote possibility that climate change is dominated by natural rather than anthropogenic factors (Stott et al, WIRE, 2010) Human influence on Antarctic warming has been detected Increasing evidence that human influence on temperature is becoming significant below continental scales Major challenges still remain in obtaining robust attribution results at scales needed for evaluation of impacts Climate models often lack processes needed to realistically simulate regional details Observed changes in non-climate quantities could be the result of additional influences besides climate thus complicating attribution studies Extremes pose a particular challenge since they are rarely observed and models don’t necessarily represent droughts, floods and hurricanes At regional scales, many challenges remain Lower signal to noise ratios Difficulties of separately attributing effects of different forcings relevant at the region Limitations of models in capturing some aspects of regional climate variability