Saturday, October 5, 2013
Problems With Fits And Making Anomaly Projections
A fit works for the time interval for which we have data to work with but fits of comparable accuracy can have quite different projections. I used the previous fit with a functional form consisting of a cubic polynomial plus a simple sinusoidal function and made a projection of the land northern hemisphere anomaly based on it.
When we compare this projection with one with a functional form consisting of a linear term plus the simple sinusoidal function and its second harmonic we get a quite different projection.
The outcome is not at all certain and we are faced with a problem like that of describing what we see in a set of ink blots. The global warming people may be trying to err on the side of caution but are we in control of the situation if we do not actually know what is happening with respect to global warming?