cc: m.hulme@uea.ac.uk
date: Mon, 1 Nov 1999 14:46:53 BST
from: "Simon J Shackley" <mcysssjs@fs1.sm.umist.ac.uk>
subject: scenario choice 
to: paula.harrison@ecu.ox.ac.uk, pam.berry@ecu.ox.ac.uk, terry.dawson@ecu.ox.ac.uk, Janet.Sells@bbsrc.ac.uk, eric.audsley@bbsrc.ac.uk, anton.debaets@bbsrc.ac.uk, i.holman@cranfield.ac.uk, tom.downing@ecu.ox.ac.uk, p.loveland@cranfield.ac.uk, r.nicholls@mdx.ac.uk, rounsevell@geog.ucl.ac.be, robert.w.wood@man.ac.uk, t.wilson@mdx.ac.uk

dear Colleagues 

I spoke to Peter Fox at Environment Agency about the scenario issue 
and he had quite strong views - as indicated in his message below.  
I'm sure that Peter would welcome any responses.  

I have also spoken to Sustainability North West who are also in 
favour of including as many socio-economic scenarios as possible.  
Richenda at UKCIP is going to contact Merylyn at COP-5 to give us 
advice from UKCIP.  I think we should ask MAFF, DETR and UKWIR 
about their preferences before the december meeting.  (Do you want me 
to do this Peter?). 

In response to Mike Hulme's / Paula's proposal, as a non-expert  I 
wasn't sure why it was necessary to test both L and ML and MH and 
H.  Will this really tell us that much mroe than one of those 
scenarios? 

Also,  I don't think it is so clear that you would not get a Global 
Sustainability world with the MH or H scenario. It depends on whether 
the scenarios are about PROCESSES (attempts to achieve sustainability 
/ growth,etc) or the END POINTS (actually achieved sustainability 
according to a priori definitions, etc.).  Using scenarios to think 
about the processes of reaching different end points / objectives is 
perhaps more in tune with how political processes work in practice.  
We can imagine, for example, that a MH level of climate change might 
be one of the anomalies which would face a world which thinks itself 
moving towards global sustainability.  Of course, we can use the end 
points definition of scenarios which, by definition, would make MH 
and H inconsistent with Global sustainability (though uncertainty / 
surprise could still perhaps disrupt our expectations?). 

I think it follows less clearly - even for end point definition of 
scenarios - that the Regional Enterprise world is inconsistent 
with L or ML.  Entreprise in one region is likely to co-exist with 
lack of enterprise or attempts to achieve sustainability in other 
regions.  Only if you assume that all regions in the world are 
committed to enterprise, and manage to succeed at being enterprising, 
can you assume MH or H climate scenarios.  That doesn't seem a very 
plausible future scenario to me.  And how different would it be from 
the global markets scenarios? 

Yours 

Simon 


Message from Peter Fox, Environment Agency 

Simon,

Further to our telephone call I thought I would drop you a note with
my views on the need to sample the climate change and socio-economic
scenarios for interactions study.

I think you have chosen the wrong balance, we should do all 4
socio-economic scenarios with 2 climate change.

The overall differences between the four climate change scenarios can
be expressed in one dimension: hot, hotter, hottest. Sorry for
simplifying this but you understand what I mean. Hence, choosing 2
will allow us to explore and report the range of different responses.
However, the socio-economic scenarios express differences in two
dimensions, global vs local and stewardship vs economic growth.
Dropping even one of these will mean that one or other
between-scenario comparisons will not be possible. 

Also, it seems likely that the socio-economic scenarios will contain
the greatest overall variance. This strengthens the need for more
scenarios to be included. In my experience of modeling, you need to
include more information where variance is greatest.

Lastly, if we loose the chance to make climate change a reality to
people in the regions we will have missed a major trick in REGIS. The
study of socio-economic impacts is critical to success.

I'm sorry I cannot advise you on a choice, however, I hope this is
helpful.

Peter


