GFDL - Geophysical Fluid Dynamics Laboratory

Skip to main content

A Model Study of Heat Waves over North America: Meteorological Aspects and Projections for the 21st Century

Key Findings

  • Typical synoptic features accompanying model-simulated warm episodes in various sectors of North America are in good agreement with observations.
  • Model projections indicate lengthening of the heat wave duration, as well as increases in the frequency of heat wave episodes in various sites, during the 21st century as compared to the 20th century.
  • Most of the projected trends in the heat wave statistics in different regions may be attributed to the long-term shift in the climatological norm of the local surface temperature.
  • The upward trends in heat wave duration and frequency are projected to be discernible in the early decades of the 21st century

Ngar-Cheung Lau and Mary Jo Nath. Journal: American Meteorological Society. doi:10.1175/JCLI-D-11-00575.1


The goal of this research is to evaluate the fidelity of GFDL climate models in reproducing the characteristics of summertime heat waves in North America, and to examine the model-projected changes of these characteristics in the 21st century.

North American incurs considerable economic costs due to heat waves. Understanding of the processes contributing to heat waves, and projecting changes in their characteristics in the 21st century, are an integral part of the NOAA mission to provide scientific information on the present and future states of the climate system.

This study analyzes output from comprehensive suites of model experiments with present and future climate settings, and considers heat waves in different parts of North America. It provides spatial and temporal details of the present and future behavior of heat waves.


Output from two climate models with 50- and 200-km resolution (HiRam and CM2.1, respectively) is subjected to rotated empirical orthogonal function analysis, so as to select regions with enhanced and coherent surface temperature variability. A heat wave identification scheme is then applied to these regions using model data for both present and future climate scenarios. Trends of heat wave intensity, duration and frequency are then deduced by comparing the heat wave statistics for the 20th and 21st centuries.

Known uncertainties

Present findings based on GFDL models need to be compared with those derived from other models. Robustness of the results presented herein needs to the evaluated by application of alternative identification schemes of heat waves.

Typical anomaly patterns of surface wind (arrows), sea level pressure (contours) and precipitation (green and brown shading) associated with heat waves over the Central Great Plains (left panels) and the Northern Great plains (right panels). The regions affected by the heat waves are indicated by a pink patch in each panel. Results are based on (top row) output from the C180 HiRAM model and (bottom row) observational reanalysis data.

Distributions of the averaged (left column) heat wave duration, (middle column) heat wave frequency, and (right column) number of heat wave days per year in the June-July- August season as identified at individual gridpoints. Results are based on (top row) observational reanalysis data, and output from the C180 HiRAM model for the (second row) 1971-2000 period and (third row) 2041-2070 period. Ratios of heat wave measures in the 2041-2070 period versus the 1971-2000 period are shown in the fourth row.