U.S. Climate Extremes Index (CEI): Introduction

NCEI replaced input data for the CEI mean monthly maximum and minimum temperature extremes on Thursday, October 6, 2016, coincident with the release of the September 2016 monthly monitoring report. For more information on transition from USHCN to nClimGrid data, please visit the CEI - Data Used page.


How has the climate changed over the past 50 or more years? In what ways and by how much? Many people, including climatologists, have been struggling with these questions for some time now, not only for scientific interest, but also to aid in policy decisions (IPCC 2001) and to inform the general public. In order to answer these questions, it is important to obtain comprehensive and intuitive information which allows interested parties to understand the scientific basis for confidence, or lack thereof, in the present understanding of the climate system. One tool, first developed as a framework for quantifying observed changes in climate within the contiguous Unites States, is the U.S. Climate Extremes Index (CEI).

The CEI was first introduced in early 1996 (Karl et al. 1996) with the goal of summarizing and presenting a complex set of multivariate and multidimensional climate changes in the United States so that the results could be easily understood and used in policy decisions made by nonspecialists in the field. The contiguous U.S. was selected as the focus for this study in part since climate change is of great interest to U.S. citizens and policy makers and since climate changes within the U.S. have not been given extensive coverage in intergovernmental or national reports which focus on climate change assessments (IPCC 2001; NRC 1992; NRC 2001).

In 2003, two notable modifications were made to the CEI. Indicators in the original CEI summarized trends in temperature, precipitation and drought data on an annual basis. The revised CEI now includes an experimental tropical system component and is calculated for multiple seasons. The newest indicator documents trends in tropical system activity based on the wind velocity of landfalling tropical storm and hurricanes. As of October 2004, CEI calculations begin in 1910 for all periods and are updated within a few weeks after the end of a particular season and include final quality controlled data as well as near-real-time data. In September 2005, the two components for each of four indicators (steps 1, 2, 3, and 5) are plotted separately to help in the identification of trends and variability of each component. All graphs are now plotted as bar graphs rather than dot plots. In December 2005, a year-to-date season was made available along with the other eight standard seasons. Additions and modifications made to the original CEI are explained in an article entitled "A Revised U.S. Climate Extremes Index", which was published in mid-2008 (Gleason et al. 2008).

In July 2011, a regional CEI (RCEI) was introduced, which computes the CEI across the 9 U.S. Standard Regions (Karl and Koss, 1984). Year-to-year varations in the regional index have higher amplitude swings and larger/smaller percentages of each region affected by extremes compared with the CEI. There is a good deal of spatial consistency among the RCEI indicators and similar extremes may span across or be absent from a region in any given season.

In October 2016, an update was made to the two components for which extremes based on mean monthly maximum and minimum temperatures are computed. In an effort to improve spatial resolution and serial completeness, the USHCN input data were replaced by NCEI's nClimGrid dataset. The nClimGrid dataset has a 5km resolution and is spatially and serially complete for all grids across the contiguous U.S. for the operational period of record (1910-present).