Estimated Statistical Variance of Least Squares Predicted (Extrapolated) Rising Temperatures From Climate Change
9 Pages Posted: 3 Feb 2019 Last revised: 16 May 2019
Date Written: January 20, 2019
Although NOAA global temperature change data are deterministic rather than statistical, this paper suggests a method of approximating what the variance might be were temperature actually described by a polynomial corrupted with statistical measurement noise (errors). It is a novel and simple technique of estimating noisy polynomials with unique orthogonal polynomial coefficient estimators. This allows simple estimates of the temperature at any point on the estimated polynomial and the statistical variance of the extrapolated temperature, extrapolation being of particular interest.
Keywords: Climate Change, Global Temperature, Polynomial Least Squares, Multiple-Model
Suggested Citation: Suggested Citation