Where Your Dreams Become Reality
You are reporting the following post to the moderators for review and possible removal from the forum|
Poster: rose colored glasses?
Subject: RE: Greenland ice sheet melting faster than previously thought. PLEASE RECYCLE!
The danger in extending trends from stochastic data based on some "most probable" or "mean" or some other estimate of central tendency, particularly when dealing with data that clearly have a fractal structure and associated power law relationships, is that your estimates of central tendency may be a) wrong and b) meaningless (which I mean as two different things).
If you did some basic statistical analysis of the phenomena that (were inferred to have) caused those prior fluctuations and then did some kind of monte carlo simulation with your model, including these fluctuations in some measure that makes statistical sense, you would see a) significant fluctuation in the best guestimate trends b) huge growth in uncertainty in the model outcome past a few years.
Or so I think, anywy.
Hit the submit button below if you want us to review the post. If you feel this is urgent or want a reply, email us at email@example.com about the post and please include a link to the thread the post is on and what page number/post on that page it is