Warm streaks in the U.S. temperature record: What are the chances?

Authors

  • Peter F. Craigmile,

    1. Department of Statistics, Ohio State University, Columbus, Ohio, USA
    2. School of Mathematics and Statistics, University of Glasgow, Scotland, UK
    Search for more papers by this author
  • Peter Guttorp,

    1. Department of Statistics, University of Washington, Seattle, Washington, USA
    2. Norwegian Computing Center, Oslo, Norway
    Search for more papers by this author
  • Robert Lund,

    1. Department of Mathematical Sciences, Clemson University, Clemson, South Carolina, USA
    Search for more papers by this author
  • Richard L. Smith,

    1. Department of Statistics and Operations Research, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, USA
    2. Statistical and Applied Mathematical Sciences Institute, Research Triangle Park, North Carolina, USA
    Search for more papers by this author
  • Peter W. Thorne,

    1. National Climatic Data Center, Asheville, North Carolina, USA
    2. Department of Marine, Earth and Atmospheric Sciences, North Carolina State University at Raleigh, Raleigh, North Carolina, USA
    3. Nansen Environment and Remote Sensing Center, Bergen, Norway
    Search for more papers by this author
  • Derek Arndt

    1. National Climatic Data Center, Asheville, North Carolina, USA
    Search for more papers by this author

Correspondence to: P. F. Craigmile,

pfc@stat.osu.edu

Abstract

A recent observation in NOAA's National Climatic Data Center's monthly assessment of the state of the climate was that contiguous U.S. average monthly temperatures were in the top third of monthly ranked historical temperatures for 13 straight months from June 2011 to June 2012. The chance of such a streak occurring randomly was quoted as (1/3)13, or about one in 1.6 million. The streak continued for three more months before the October 2012 value dropped below the upper tercile. The climate system displays a degree of persistence that increases this probability relative to the assumption of independence. This paper puts forth different statistical techniques that more accurately quantify the probability of this and other such streaks. We consider how much more likely streaks are when an underlying warming trend is accounted for in the record, the chance of streaks occurring anywhere in the record, and the distribution of the record's longest streak.