Recent summer warming in northwestern Canada exceeds the Holocene thermal maximum

Trevor J Porter,Duane G Froese,Lauren J Davies,Spruce W Schoenemann,Sasiri Bandara,Eric J Steig

doi:10.1038/s41467-019-09622-y

Abstract

Eastern Beringia is one of the few Western Arctic regions where full Holocene climate reconstructions are possible. However, most full Holocene reconstructions in Eastern Beringia are based either on pollen or midges, which show conflicting early Holocene summer temperature histories. This discrepancy precludes understanding the factors that drove past (and potentially future) climate change and calls for independent proxies to advance the debate. We present a ~13.6 ka summer temperature reconstruction in central Yukon, part of Eastern Beringia, using precipitation isotopes in syngenetic permafrost. The reconstruction shows that early Holocene summers were consistently warmer than the Holocene mean, as supported by midges, and a thermal maximum at ~7.6–6.6 ka BP. This maximum was followed by a ~6 ka cooling, and later abruptly reversed by industrial-era warming leading to a modern climate that is unprecedented in the Holocene context and exceeds the Holocene thermal maximum by +1.7 ± 0.7 °C.

Highlights

Eastern Beringia is one of the few Western Arctic regions where full Holocene climate reconstructions are possible
Noise is inherent to all proxies, but recovery of the climate signal can be optimised through replication and a diverse, multi-proxy approach
Pollen and midges agree there was a climate optimum at ~7–5 ka BP, followed by a long-term cooling. These proxies show large differences before 8 ka BP with pollen indicating that early Holocene summers were persistently colder than the Holocene mean (ΔT range = –3.3 to –0.5 °C; mean ΔT = –1.6 °C), and midges showing that early Holocene summers were similar to the Holocene mean (ΔT range = –0.1 to + 0.4 °C; mean ΔT = + 0.1 °C)

Summary

Introduction

Eastern Beringia is one of the few Western Arctic regions where full Holocene climate reconstructions are possible. The two records show small systematic differences in their mean δD and δ18O values during the period of overlap from 8.5 to 6.4 ka cal BP (composite depths 380–290 cm), the upper core being relatively depleted compared to the lower core in δD and δ18O (Fig. 2a, b).

Results

Conclusion