Temperature variability and early clustering of record breaking events

As the number of studies using record breaking statistics in climatology is growing rapidly, the criteria for choosing time period become essential. To that end, here we examine the evolution of monthly mean temperatures and its dependence on beginning and final year. Specifically, we use the variability index Alpha (Anderson and Kostinski, J, Appl. Met. and Clim., 2010) such that <Alpha> = 0 indicates no trend in variability. Generally, Alpha has decreased between 1900 and 2010 (indicating decreasing variability) for stations from the contiguous United States (United States Historical Climatology Network, version 2). We find, somewhat surprisingly, that the observed decrease is due to an early clustering of records. While detailed results depend on whether the data is gridded, detrended, etc., the general finding appears remarkably robust and holds globally as well.