date: Tue May 22 18:06:10 2001 from: Mike Hulme subject: Fwd: quick question about near term climate change to: t.mitchell Tim, Can you help on this one? Refer to me only if you get stuck. Mike X-Mailer: Novell GroupWise Internet Agent 5.5.5.1 Date: Tue, 22 May 2001 16:01:28 +0100 From: Sari Kovats To: "m.hulme" Cc: Tony McMichael Subject: quick question about near term climate change X-Virus-Scanned: by AMaViS snapshot-20010407 Dear Mike Hope you are well. We are struggling a little bit about climate change before 2010 and need some advice. We have this project for WHO - the global burden of disease fro climate change - which requires estimates for the following years 2000 2001 2005 2010 2020 2030 We have agreed to use estimates based on predictive modelling - mainly the fast track work plus some other models. Therefore, we decided in advance to set all climate change and climate impacts at zero at 2000 - which is the reference point for the global burden of disease exercise. Our problem is the that the baseline climate is the 1961-90. Which could mean that all deviation from this normal is defined at "climate change". Although this would be observed and not modelled. I think it is conceptually easier for set climate change and impacts at zero for 2000. We could say that some climate change has occurred but we are not confident enough to say that health impacts have occurred (and do not want to estimate them). We are using the fast track climate scenarios - HadCM2 ensemble mean and s750 and s550 scenarios - for 2050s and 2020s. In order to provide estimates for years before 2020 we need to interpolate back to 1990 - which is when climate change starts (according to models). This means that we would be able to estimate impacts for 2000 - as curve would not go through zero - but we probably wont present them to WHO. Having just answered our own questoins - do you think this is OK? In one of the tables you produced for the fast track work (see attached) - does figure 0.3 refer to observed change in annual global surface mean temperature (average for decade 1990 to 1999) compared to 1961-1990 30 year average? We could apply this to our models and see if it fits with our interpolated estimates. thanks very much. Sari ******************* Sari Kovats Research Fellow Centre on Globalisation, Environmental Change and Health Dept of Epidemiology and Population Health London School of Hygiene and Tropical Medicine Keppel St London WC1E 7HT tel: +44 20 7612 7844 fax: +44 20 7580 6897 sari.kovats@lshtm.ac.uk