## Friday, May 6, 2011

### years are not brothers

#### 4 comments:

Oale said...

On top of the hat attempt to smooth the ends of a dataset, likely done somewhat wrong. there's no data prior the start of the year nor after the year here.

Oale said...

Yes, I made a mistake in calculating the boundaries, but I think that's not impossible to solve.

I probably could do some improvements to this statistical model if I had more maths skills. F.e., if I assumed that earth revolves around the sun, or rather, that the hemispheric insolation varies according to a simple elliptical gravitational equations (not talking bout relativistic effects), I would have a model for all the hemispherically averaged measures of energy, thus assuming that winters are on average colder than summers, and assuming years are somewhat brothers, I could likely remove most of the uncertainty in the ends of the dataset. But in doing this, I would also have to allow a lag on the equation because it looks like the dataset has its minimum in september and not in June 22nd, which would be the assumed minimum from a strict insolation equation.

Oale said...

The improvement mentioned in the previous comment of course assumes the dataset is from a such hemisphere that has a boundary that does NOT cross the equator perpendicularly, and is gathered from such distance of the other possible energy sources, that the sun is the dominant energy provider. Likely there is also some internal variability involved as the lag between the dataset and the insolation equation is about 83 days.

Oale said...

A 'boundary that crosses the equator perpendicularly' in the last comment would be a case where the tangent on the maximum circumference of the ellipsoid (presumed to form from matter in a two-body system in the long run as the entropy increases by time(ref to thermodynamics)) projected on a plane forms a straight angle with the projection of the maximum circumference on the same plane.

But this is going a bit too deep for me ('Plane' is defined here as a 2-dimensional entity, or, a surface of an archimedean solid...), and I don't want to go to the psychology of '2-dimensional thinking', whatever it is, so I'll stop.