Originally posted by oxmixmudd
View Post
Congratulations, you've finally given me cause to call you stupid. Happy?
And then there's the 'we need to correct our raw data' bit - which brings the raw data into question.
GAH!
Even if - and I'm not even close to conceding this based on what you wrote - correcting were somehow fundamentally different (hard to prove when you DESCRIBE THE SAME THING) it would still have similar issues - when you (general) think you're fixing something, you are running the risk of entering your own bias.
But you clearly don't understand the difference between weighted data and corrected data. For example, the TOBS adjustment I am talking about is not a weight, it is a correction. Likewise, if a thermometer was moved from a site that due to urban heat island trended 2 degrees warmer than the sight is was moved to, the data itself is also corrected for that shift over time. Homogenization is also a bias removal technique, were sites that tend to give biased readings are adjusted based on readings from areas around it (this has to be done carefully, a thermometer in a valley legitimately gives warmer readings that the 10 thermometers around it that are located in the hills above the valley).
But again, in summary, the majority (I can't guarantee all) of what we are talking about, and certainly the most significant contributors to the adjustments made, are not weights. They are know bias corrections.
Jim
YOU are the one that doesn't know what the bleep WEIGHTING actually IS!
Comment