The unprecedented Pacific Northwest heatwave of June 2021

Abstract

In late June 2021 a heatwave of unprecedented magnitude impacted the Pacific Northwest region of Canada and the United States. Many locations broke all-time maximum temperature records by more than 5 °C, and the Canadian national temperature record was broken by 4.6 °C, with a new record temperature of 49.6 °C. Here, we provide a comprehensive summary of this event and its impacts. Upstream diabatic heating played a key role in the magnitude of this anomaly. Weather forecasts provided advanced notice of the event, while sub-seasonal forecasts showed an increased likelihood of a heat extreme with lead times of 10-20 days. The impacts of this event were catastrophic, including hundreds of attributable deaths across the Pacific Northwest, mass-mortalities of marine life, reduced crop and fruit yields, river flooding from rapid snow and glacier melt, and a substantial increase in wildfires-the latter contributing to landslides in the months following. These impacts provide examples we can learn from and a vivid depiction of how climate change can be so devastating.

Fig. 1. June 2021 Pacific Northwest heatwave temperatures.

Heatwave daily maximum near-surface (2 m) temperatures (TX). a ERA5 reanalysis data maximum 3-day running mean (between 23 June and 02 July 2021) of TX anomalies relative to a daily 1981–2020 climatology (see Methods for more details). b Absolute TX values for 2021 (solid) and 1981–2020 climatology (dashed); black lines: spatial average of ERA5 data over the black box in a, with shading ±1, 2, and 3 standard deviations; red lines: observations from Lytton, British Columbia (BC), denoted by the red triangle in a—missing values from 1 to 5 July are likely due to the wildfire. Black letters and outlines in a show the main Canadian provinces and US states referred to in this study. Gray outlines and letters in a show the eight crop regions in BC, discussed further in the section on Agricultural Yields.

Fig. 2. Temperature record exceedances.

Exceedance of previous record high temperatures during a the June 2021 Pacific Northwest heatwave, b the July–August European heatwave of 2003, and c the July–August Russian heatwave of 2010. Filled contours show ERA5 since 1950, whilst individual markers show observational station data in Canada; see Methods for record lengths.

Fig. 3. Meteorological conditions leading up to heatwave onset.

Meteorological conditions over the North Pacific for a 00UTC 21 June, b 00UTC 23 June, c 12UTC 24 June, and d 12UTC 26 June. Data from ERA5 reanalysis showing: mean sea-level pressure (MSLP; contoured, hPa), 700 hPa relative humidity (RH; shaded, light blue >70%, dark blue >90%), and 250 hPa wind vectors (ms⁻¹, colored by wind speed). Coastlines and country borders are shown in green to distinguish them from the MSLP contours.

Fig. 4. Back-trajectories for Near-surface Heatwave Air.

a Nine representative four-day backward trajectories terminating at 500 m above ground level (AGL) within the strongest heat anomaly (boxed area in Fig. 1a). b Potential temperature of parcels along the trajectory, time from right to left. Potential temperature changes indicate diabatic heating/cooling. Small (large) markers indicate every 6 (24) h, and colors correspond to the trajectory terminating latitude. Trajectories were computed and plotted using GFS 0.25° forecast data and the NOAA HYSPLIT model.

Fig. 5. Subseasonal forecasts of extreme temperature and atmospheric blocking.

Fraction of ensemble members that predicted 2 m temperature (a–e) and number of blocked days (f–j) greater than the 95th percentile of the respective climatologies during 25 June to 1 July. For a 95th percentile event, 0.05 is the statistically expected value if there is no forecasting skill. Forecasts were initialized on 3 June (a, f), 7 June (b, g), 10 June (c, h), 14 June (d, i), and 17 June (e, j). Gray contours show the observed location of the heatwave based on ERA5 2 m temperature data.

Fig. 6. Heatwave impacts on marine life.

Thermal images showing extreme high surface temperatures during low tide on 28 June, 2021, on a a rocky intertidal shoreline and b within a mussel bed. Scale bars indicate the range in temperature from the coolest to warmest parts of the image, while the value at the upper left indicates the temperature in the cross-hairs at center. Note that the mussels in b have died and are gaping open. A subset of species impacted by the heatwave are shown in c–i, including c bay mussels, Mytilus trossulus, d Pacific oysters, Magallana (= Crassostrea) gigas, e heart cockles, Clinocardium nuttallii, f leather stars, Dermasterias imbricata, g kelp crabs, Pugettia producta, h dogwhelks, Nucella lamellosa, and i barnacles, Chthamalus dalli (upper portion of image) and Balanus glandula (lower portion of image). See Methods for locations and dates of photos in c–i.

Fig. 7. Heatwave impacts on wildfire conditions.

Model estimated Fire Weather Index (FWI; a, b) and smoke concentration and satellite hotspots (c, d) for pre-heatwave conditions on 20 June (a, c) and post-heatwave conditions on 3 July (b, d). Smoke concentration in c, d is represented by the concentration of particulate matter (PM) 2.5. Satellite hotspots in c, d, indicating likely wildfire activity, are shown by the red triangular markers.

Fig. 8. Heatwave impacts on crop health.

Weekly Normalized Difference Vegetation Index (NDVI) from the Crop Condition Assessment Program (see Methods), spatially averaged over agricultural divisions in British Columbia, organized alphabetically. 2021 values are shown in the black line, with the 2000–2020 climatology in blue and shading showing interannual variability for this 2000–2020 period (±1 and 2 standard deviations). The gray highlighted region shows the heatwave period, 20 June—3 July; data are reported as weekly averages and so the heatwave is split between weeks 20–26 June and 27 June—3 July. The locations of the 8 agricultural divisions are shown in Fig. 1a in gray contours.

Fig. 9. Heatwave impacts on streamflow.

Streamflow observations at nine stream gauge stations in 2021 (black line) relative to the 1979–2020 median (blue line) and 1 standard deviation range (shaded). Gauges are organized from top to bottom by basin glacier coverage: highly glaciated basins (a–c), lightly glaciated basins (d–f), and minimally or non-glaciated basins (g–i). See Supplementary Fig. S6 for locations of these gauges.

Fig. 10. Example of post-wildfire debris flows following the heatwave.

Post-wildfire debris flows in Nicoamen River watershed triggered by a rainstorm on 16 August 2021. Severely burnt vegetation is shown by blackened trees. Debris flows originated as slope wash and rilling on steep burnt slopes.