The presentation, An Improved Assessment of Record-Setting US Daily Temperatures Using Homogenized Data, outlines the use of a novel homogenized gridded Berkeley Earth daily temperature data record to improve the assessment of record-setting daily US temperatures; correcting for systemic biases throughout historical temperature records, this novel daily temperature data set finds that more maximum daily temperature records have been set in the last decade (2010-2019) than in the 1930-1939 “dust bowl” period.
While monthly homogenized data sets have been available and widely used for some time, this work will provide the first set of daily homogenized data, as well as produce the first detailed methodology behind this homogenization process. Final publication expected early 2022.
This work is presented alongside co-authors Dr. Robert Rohde, Dr. Michael Mann, Dr. Don Wuebbles, and Dr. Mark Boslough. The full abstract of the paper is available below.
Changes in the occurrence of record-setting daily maximum (TMax) and minimum (TMin) temperatures can provide an important indicator of the change in extremes. Prior analyses have found that the number of all-time daily TMax records set in the conterminous United States (CONUS) were notably higher in the 1930s dust-bowl era than in recent decades. However, historical daily temperature records are subject to large systemic biases caused by station moves, instrument changes, time of observation changes, and other inhomogeneities. While monthly homogenized records that correct these biases have long been widely used, homogenized daily records have previously been unavailable. Using a novel Berkeley Earth gridded homogenized daily temperature record, we find that more all-time daily CONUS TMax records have been set in the past decade (2010-2019) than in the 1930-1939 period. The years 1934 and 2012 are effectively tied for the most daily TMax records set. We also find a robust decline in the number of all-time daily CONUS TMin records set over time. The application of homogenization techniques to daily temperature data is important to accurately understand the evolution of temperature extremes over the past century.