cc: Tony McMichael , "sarah.randolph" date: Thu, 02 Dec 1999 17:00:58 +0000 from: Sari Kovats subject: Re: health report to: "m.hulme" , "david.rogers" To add in my two-penny worth. I do not think it is appropriate to discuss the relative merit of impact models in Mike's chapter (or even put it in the reference list). The place to do this would be in the methods chapter which I sent to Roger and Sarah last week and I am still waiting for your and Sarah's comments ( as per the meeting minutes)....[I will also send it to Steve L today] Mike's bit is ok with me otherwise.. Also, Roger, Sarah, and Steve, is it possible for this assessment to be cited in the IPCC chapters (9-health and 13-Europe) in the forthcoming report? As you correctly point out, attention is focussed on the Martens/Patz work - and part of the reason for this is that nobody else is publishing in this area! It would be great if you could write it up for a journal paper - would you also consider publishing the work on vectors in Mike's Southern Africa report in a journal to make it more accessible? bye for now best wishes Sari ******************* Sari Kovats Research Fellow Dept of Epidemiology and Population Health London School of Hygiene and Tropical Medicine Keppel St London WC1E 7HT tel: +44(0)171 612 7844 fax: +44(0)171 580 6897 >>> "David J Rogers" 12/02/99 04:10pm >>> Mike, Sari, I have no problem with this except that I do not really think that your first example should be called climate matching. It is a TEMPERATURE matching exercise only. I am happy with your multi-variate definition, perhaps with the slight caveat that the MORE variables you include in such a multi-variate description of climate the FEWER places you will match to them. It is importnat to choose ONLY those climatic variables that you think are important in determining (in our case) vector borne disease risk. Mike, you will also be aware that your criticism of the first approach neatly sinks more or less ALL the current predictions about the future spread of malaria from the Martens/Patz schools. Right on! You may, therefore, like to refer to their work in your reference list Regards David > David and Sari, > > Here is an amended comment on the use of analogue regions for the health > report. This comes from p.6 of the 1st draft. I hope this strikes a > reasonable balance and gives you enough degrees of freedom. Let me know if > you have any contrary thoughts. > > Thanks, > > Mike > > _________________________________________ > > > "Although GCMs provide the most credible basis for constructing climate > change scenarios, it is important to make some comments about the use of > spatial analogue climates. Spatial analogue climates have generally been > used in one of two ways in climate scenario and impact studies: i) as a > means of generating a climate scenario for a region, and ii) as a means of > considering the sort of impacts climate change may induce in a region. We > will comment briefly on each of these in turn. > > When using spatial analogues to construct climate scenarios, the > contemporary climate of a geographically distant location is used as an > analogue for the future climate of the target location. The analogue > region is usually identified simply on the basis of the assumed future > change in annual-mean temperature at the target location. Thus the climate > of southwest England by 2050 may be said to resemble the present climate of > the Bordeaux region in France following a 2ºC warming, simply because > Bordeaux is presently about 2ºC warmer than southwest England. This > approach to climate scenario construction is seriously flawed since climate > cannot simply be transposed in space simply on the basis of an annual-mean > temperature. For example, the seasonal and multivariate character of > climate at a location is intimately bound up with the local topography and > land surface characteristics and cannot be captured by annual-mean > temperature. > > The second use of spatial analogues to identify possible impacts has more > legitimacy. Here, multivariate future climate scenarios for the target > location are derived using GCM results for a given time period. The > question is then asked, 'Where, at the present time, are there climates > similar to those the target location will experience in the future?'. > Lessons may then be learned from these derived analogue regions about the > range of climate-dependent ecosystems, diseases, etc. that the target > location may experience in the future (Note: this analogue approach says > nothing about rates of change nor the ability of species, etc., to > migrate). As long as this search for analogue climates uses multivariate > techniques for climate matching - for example, by matching the annual range > of temperature, precipitation seasonality, humidity regimes - rather than > just the annual-mean temperature then this use of analogue regions may have > some value. This latter approach to analogue climates is commented on > further on p.33 and p.69 [?] in regard to particular health applications." > > > > ***************************************************************************** > Dr Mike Hulme > Reader in Climatology tel: +44 1603 593162 > Climatic Research Unit fax: +44 1603 507784 > School of Environmental Science email: m.hulme@uea.ac.uk > University of East Anglia web site: http://www.cru.uea.ac.uk/~mikeh/ > Norwich NR4 7TJ > ***************************************************************************** > Annual mean temperature in Central England for 1999 > is currently about +1.3 deg C above the 1961-90 average > *************************************************** > The global-mean surface air temperature anomaly for 1998 > was +0.57 deg C above the 1961-90 average, the warmest year yet recorded > ***************************************************************************** > > David J. Rogers Professor of Ecology TALA Research Group, Department of Zoology, South Parks Rd., Oxford OX1 3PS, England Tel/FAX (+44) 1865 271240 Attachment Converted: "c:\eudora\attach\METHS_5.doc"