An Honest Scientist At The University Of Wisconsin

“Given our level of ignorance about what’s going on, we don’t want to compound that with a level of arrogance by saying we know what’s going to happen in a month,”

Jonathan Martin

http://www.csmonitor.com/

 

About Tony Heller

Just having fun
This entry was posted in Uncategorized. Bookmark the permalink.

12 Responses to An Honest Scientist At The University Of Wisconsin

  1. Sense Seeker says:

    I wonder what he would say about the climate over a number of years. After all, predicting long term averages is a completely other business than predicting the weather in, say, a month.

    • Dave N says:

      He wasn’t even too keen on predicting what will happen in a year.

    • Paul H says:

      Why not ask him then.

    • Mike Davis says:

      It is easy to predict long term patterns if long term patterns are evident in historical records you just need to average out the observed variations and project those into the future. Care should be taken to use the longest period available of historic records and searching for ALL available patterns. It would probably be easier to do this on a regional rather than global basis as different regions follow different patterns.
      That is ONE of the evident problems with Climate models being used today.

    • Precting the future climate is infinitely more difficult than predicting the weather in a month.

    • Ian says:

      The average of erroneous data is still erroneous data. If someone uses a computer model to predict global daily temperatures for the next thirty years (the arbitrarily defined time period of “climate”) and then claims that the average of this data is a more accurate predictor of what mean global temperatures over the period will be than the daily predictions of temperature are for each day of the period, then they are utterly crazy. If the idiotic notion that “the errors tend to cancel each other out over time” were true, then it would follow that a weekly mean temperature prediction would tend to be more accurate than a daily prediction, monthly more accurate than weekly, yearly more accurate than monthly, and so on. This is a statistically testable proposition, both in theory and in practice. We don’t have to wait 30 years or longer to see if the climate modelers’ claims are true, they can, if they are right, show us right now that their models’ predictive powers improve progressively with time. So why don’t they?

      • Mike Davis says:

        When talking about Long Term, I am talking about Thousands of years and even hundreds of thousands of years. Thirty years is regional weather and only a portion of a pattern. You can not build a house using the pattern for a door or window and that is about what the current modelers are trying to do. They do not even get that correct because they are patterning their door from hanging beads. and their window from the curtains.

      • Latitude says:

        Good analogy Mike…..

    • Ian says:

      @Mike Davi

      I was actually addressing our self-described “Sense Seeker”. I’m not quite sure what your comment is relevant to here, so, at the moment, I have nothing to say about it.

Leave a Reply

Your email address will not be published. Required fields are marked *