Friday, November 09, 2012

Temperature News

GISS reports the second-warmest October since 1879. They now have the last 10 years (120) months) as 0.17°C than the previous 10 years.

With HadCRUT4's September data, the "David Rose Hole" I mentioned earlier has shrunk from 10 months to 8. (Autocorrelation not included; I'm working on that, but haven't yet figured out if lags > 1 matter to the statistical uncertainty. Can anyone point me to a paper that discusses how to calculate the effective number of degrees of freedom of a time series (equation 9 in this Tom Wigley paper) for lags up to some number M, given all the autocorrelation coefficients rk?)

No comments: