The AstroStat Slog

From arxiv/astro-ph, arXiv:0706.2590v1 Extreme Value Theory and the Solar Cycle by Ramos, A. This paper might drag a large attention from CHASC members.

The paper points out that the extreme value theory is scarcely applied in astrophysics; however, extreme events are the most interesting ones in astronomy. Their goal is forecasting solar cycles via statistical tools for extreme values. They adopt the generalized Pareto distribution and estimates parameters of the distribution with a maximum likelihood approach. As a consequence of estimating the cumulative distribution of extreme values, return time (waiting time until an event goes beyond a threshold) and return level (typical extreme event to be discovered) were reported to be matching the empirical distribution of solar cycle. Also, a warning on extrapolating on return times has been given due to the short period of observation (250 years).

Note that the tabulated data is publicly available as the International Sunspot Number. Sunspot number has been a standard example for time series studies but still accept challenges from many scholars.

An additional interesting astronomical paper with a bayesian methodology appeared in arxiv/astro-ph on the same date: arXiv:0703417 MCMC analysis of WMAP3 and SDSS data points to broken symmetry inflation potentials and provides a lower bound on the tensor to scalar ratio by Destri et.al.