 |
 |
Ray, I'd personally like to see you provide some discussion on the difference between a prediciton and a projection. To me, and I do lots of forecasting, the two terms are pretty much synonymous. Further, from my readings on statistical models, the only real test of a model is how well its *predicitions* match up with the real data. Steve --- Ray Lanier <[log in to unmask]> wrote: > Hello Steven, > > It seems to me that your comments are not well-grounded, perhaps I am > wrong. > Let me offer my experience in an effort to clarify for us both. I'm > obviously biased! :-) > > All models that I have been involved with - on agriculture water & land > use > issues - have begun with many trial runs to test validity of the data > and > assumptions used in model construction. These analyses were then > analyzed > by professionals in the several fields involved: agr production and > yield > assumptions, current and over time, technology, water available, costs & > prices, land available and productivity, environmental factors, etc. > For > example, I remember particularly one model 35 years ago. When I > presented > it for review by the several specialists one old agricultural specialist > sorta drawled: "Well, I see you have projected a move of the dairy > industry > from Wisconsin to Iowa!" For various institutional reasons that was not > a > valid outcome in the time-frame involved. So go back to the drawing > board > to see where the data and/or assumptions were not acceptable. > > Once the model was judged "reasonable" in data and basic assumptions the > analysis could begin. First we would evaluate the situation based on > the > assumption that there would be no change in poliicy/program. This is > extremely important because it establishes a baseline that suggests the > magnitude of, for example, the potential water supply problems, if any. > Then it offers a baseline against which the several proposals to > alleviate > any water supply problems that might be proposed. These programs offer > us > an opportunity to evaluate, relatively easily, many policy/program > options > to enable us to better select those that are most helpful and least > damaging. > > None of these analyses were "predictions" - implying factual accuracy. > They > were *projections*, implying best estimates of the relative consequences > that could be expected under the assumptions and the data available at > the > time of study. > > Those models were extremely important because they enabled us to > incorporate > interrelationships among many activities in a way that the old hand > calculators could not provide. And, in today's modelling world, they > were > very primitive. The model outlined above was a matrix of about 200+/- > columns and about 100+/- rows with a large percentage of empty cells. > Yet > it still took over 12 hours to run on an IBM computer complex taking up > most > of a building at U. Illinois Champaign-Urbana - and that was just to see > if > we had a potential feasible solution. And we ran the problem on the > weekends to save money - it would take a weekend to get to a feasible > solution. Today, I could run that problem on my little old desktop in > minutes. > > Now the analytical capabilities are far in advance of the little work we > did. But for me, the significance of the great expansion in computer > power > is, first, the savings in time & money; second, the opportunity to > incorporate more of the complexity of the environmental problems in the > analysis. > > The important ethical issues for me are whether the data and assumptions > are > valid in the eyes of the professionals reviewing the work and whether > the > problem is appropriate to the particular computer complex used. In > today's > world, I think that the organization and review of data and assumptions > in > such large problems must be horrendous. We should expect errors; the > test - > whether there is adequate review by competent professionals of the base > program and the analyses. > > And these are only projections to help society visualize the potential > problems and to get some understanding of the *relative* worthwhileness > of > proposed policies/programs. > > I think it would be unethical not to use the newest computer technology > as > it becomes tested and available. > > Steven, and all, I would appreciate any comments. > > Ray > ------------- > > ----- Original Message ----- > From: "Steven Bissell" <[log in to unmask]> > To: <[log in to unmask]> > Sent: Wednesday, July 25, 2001 10:16 AM > Subject: more on global warming models > > > > Here is another article (I haven't seen the cited article in > _Science_) > that > > seems to me to be based on the use of models. This is not new > evidence, it > > is just more modeling of old data and, worse, it is based on > assumption > > about policy. I'm not sure this serves anyone very well. > > > > Steven > > > http://www.enn.com/news/enn-stories/2001/07/07242001/warming_44399.asp?site= > > email ===== "In a nutshell, he [Steve] is 100% unadulterated evil. I do not believe in a 'Satan', but this man is as close to 'the real McCoy' as they come." --Jamey Lee West __________________________________________________ Do You Yahoo!? Make international calls for as low as $.04/minute with Yahoo! Messenger http://phonecard.yahoo.com/