Unprecedented 21st century heat across the Pacific Northwest of North America

Global surface air temperatures and the frequency and intensity of extreme heat events are rising due to anthropogenic climate change. The IPCC reports that global mean temperatures in the early 21st century exceeded pre-industrial levels by nearly 1°C and are projected to surpass 1.5°C above pre-industrial levels by 2040. The last two decades have witnessed a substantial increase in extreme summer temperatures worldwide. The summer 2021 heatwave in the PNW of the US and Canada was exceptionally prolonged and severe, reaching temperatures as high as 49.6°C. This event led to record-breaking summer temperatures, increased wildfire activity, elevated heat-related deaths, and significantly impacted the region's ecosystems and communities. The PNW's historically temperate climate lacks adequate infrastructure and societal preparedness for such prolonged extreme heat events, highlighting a critical vulnerability. This study uses tree-ring data to reconstruct PNW summer temperatures over the last millennium to contextualize the 2021 event and assess the likelihood of future occurrences using climate model projections.

Existing literature, using limited observational data and modeling, indicates the 2021 PNW summer heatwave was highly anomalous, lacking a comparable modern analog. Rapid attribution studies suggest anthropogenic climate change played a crucial role, but quantifying the event's rarity using only historical observations is challenging. Studies utilizing tree-ring data to reconstruct past temperatures have proven valuable in characterizing past climate variability, particularly in the Northern Hemisphere. Tree growth variability, especially wood density, strongly reflects summer temperature variations, allowing for the creation of millennial-length temperature estimates to be compared with modern observations and future climate projections. This research leverages this approach to improve the understanding of long-term temperature variability in the PNW, including comparisons to past warm periods like the Medieval Climate Anomaly (MCA).

Tree-ring samples (total ring width and latewood blue intensity) were collected from 29 conifer collections across 17 sites in the PNW. Samples were processed, dated using COFECHA, and standardized using the SignalFree (SF) framework with an age-dependent spline (ADS) to remove age-related trends. A nested principal components regression (PCR) model was used to reconstruct millennial-length (950-2021 CE) summer temperature estimates, leveraging the strong relationship between tree-ring data and ERA5 reanalysis data (1950-2021). Model validation involved cross-calibration and verification statistics (RE, CE, CRSQ, VRSQ, VRE, VCE) and maximum entropy bootstrapping (MEBoot) to estimate uncertainty. The reconstruction was compared with an independent Northern Hemispheric temperature reconstruction and the Alberta Icefields reconstruction. CMIP6 SSP2-4.5 and SSP3-7.0 projections were used to evaluate future changes in PNW summer temperatures and the likelihood of 2021-like events. Statistical analysis included probability density functions, LOESS smoothing to identify multi-year warm/cool periods, bivariate analysis of warm periods, and assessment of multi-decadal trends. Risk calculations were based on multi-model ensemble data from CMIP6 projections.

The tree-ring reconstruction shows that the 2021 PNW summer temperatures were unprecedented since 950 CE. Probability density estimates based on both instrumental and reconstructed data place the 2021 summer average far outside the range of historical variability. The reconstruction reveals several multi-decadal warm periods during the last millennium, particularly during the MCA. However, the magnitude and duration of the recent warming (1979-2021) are significantly greater than past warm periods, with the 2021 value being at least 1.5°C warmer than any MCA departure. The rate of warming over the last few decades is also unprecedented. The comparison of the reconstruction with CMIP6 model projections (SSP2-4.5 and SSP3-7.0) indicates a strong agreement in trends over their shared period. The 2021 anomaly exceeds the 95th percentile of CMIP6 summer temperatures for SSP2-4.5 (2005-2035). CMIP6 projections suggest a 50-60% likelihood of 2021-like extreme summers in the PNW by 2050 under SSP2-4.5 and SSP3-7.0 scenarios. Under SSP3-7.0, the risk increases dramatically to 94% by 2100.

The findings corroborate previous studies linking extreme weather events to anthropogenic climate change. The tree-ring reconstruction provides a longer-term perspective, confirming the exceptional nature of the 2021 event while highlighting the occurrence of past warm periods, particularly during the MCA. Although past warm periods are not perfect analogs to modern conditions, understanding the magnitude and frequency of past extremes is crucial. The agreement between the reconstruction, instrumental data, and CMIP6 simulations strengthens the evidence for human influence on the observed warming. The current warming rate in the PNW aligns with the SSP2-4.5 scenario, suggesting a high likelihood of 2021-like events becoming increasingly common without significant emission reductions. The study emphasizes the importance of empirical tree-ring data in better characterizing internal climate variability to improve future projections of extreme summer temperatures.

The 2021 PNW heatwave, analyzed within the context of the last millennium, illustrates the unprecedented warming trend and increasing frequency of extreme summer temperatures. This underscores the significant impact of anthropogenic climate change. Future projections, absent substantial mitigation efforts, point towards a stark increase in the probability of such events. This highlights the urgent need for climate change adaptation strategies, particularly in regions historically less exposed to extreme heat. Future research should focus on refining climate model projections to better capture internal variability and regional-specific climate dynamics, particularly regarding extreme events.

The reconstruction's accuracy decreases with time, especially prior to 1400 CE, due to a reduced number of tree-ring chronologies. While the model is robust, some uncertainty exists in the earliest part of the reconstruction. Although the study uses multiple models for comparison, there is always some uncertainty associated with climate model projections, including uncertainties in future greenhouse gas emissions.