| description abstract | Based on counts of record highs and lows, and employing reversibility in time, an approach to examining natural variability is proposed. The focus is on intrinsic variability; that is, variance separated from the trend in the mean. A variability index α is suggested and studied for an ensemble of monthly temperature time series around the globe. Deviation of ?α? (mean α) from zero, for an ensemble of time series, signifies a variance trend in a distribution-independent manner. For 15 635 monthly temperature time series from different geographical locations (Global Historical Climatology Network), each time series about a century-long, ?α? = ?1.0, indicating decreasing variability. This value is an order of magnitude greater than the 3σ value of stationary simulations. Using the conventional best-fit Gaussian temperature distribution, the trend is associated with a change of about ?0.2°C (106 yr)?1 in the standard deviation of interannual monthly mean temperature distributions (about 10%). | |
