Fig. 5

ΔT reconstructions and climate forcings. a Northern extratropical (30–90°N) multi-proxy ΔT reconstruction¹¹ (red; filled area = ±1σ uncertainty) and mean 1997–2016 GISTEMP³⁷ annual ΔT for the same area (red star); b E. Beringia multi-proxy ‘all-temperature’ composite⁹; c E. Beringia composite pollen- (blue) and midge-based (orange) summer ΔT reconstructions⁹; d DHP174 (Dempster Highway Peatland, near 174 km marker) δD_{pore ice}-based summer ΔT reconstruction (blue; filled area = ±1σ; this study) and smoothed 1900–2017 May–September GISTEMP ΔT anomalies for the study area (red; n.b., GISTEMP anomalies were smoothed with a 40-year cubic spline and normalised to the twentieth century mean of the δD_{pore ice} ΔT reconstruction; a 1991–2001 data gap in GISTEMP was filled with ERA-interim data⁶⁴); e 40-year smoothed Arctic (60–90°N) multi-proxy ΔT reconstruction⁷ (green; filled area = ±1σ) and mean 1997–2016 GISTEMP³⁷ annual ΔT for the same area (green star); f June–September insolation at 65°N⁶⁵; g Vostok, Law Dome and Mauna Loa composite pCO₂ record^66,67,68. Grey bars highlight warming intervals in the δD_{pore ice}-based ΔT reconstruction that are also evident in other ΔT reconstructions in E. Beringia and at broader spatial scales