Shifting potential for high-resolution climate reconstructions under global warming

Abstract

Reconstructions of climate in pre-instrumental times are a cornerstone of earth-system science that relies critically on statistical relationships between meteorological observations and natural proxy archives. Recent studies have frequently reported that these relationships are not stable in time (non-stationarity), which might be due to environmental change (climate, atmospheric CO2), data resolution and quality, and statistical methods applied. Here, we assess the elusive impacts of these factors on the palaeoclimatological potential across the Northern Hemisphere. Scrutinizing spatiotemporal patterns in widely applied calibration metrics derived from 3781 tree-ring chronologies and 517 published dendroclimatic studies, we show that temperature and precipitation sensitivity have increased towards present. This increase was most pronounced in moisture-limited areas and accentuated when using daily rather than monthly instrumental data. An assessment of climate scenarios for 2021–2040 indicated further expansion of areas with strong water limitation (+5 ± 2%), whereas the areas with strong temperature limitation are projected to shrink by 8 ± 3% (tree-ring width proxy) and 3 ± 2% (maximum latewood density proxy). These findings indicate that further refinement of statistical methods will likely no longer compensate for trees’ decreasing temperature sensitivity. Further, a scenario of increased CO2 fertilization may mitigate water limitation on tree growth and weaken precipitation reconstructions. Finally, we discuss the potential drivers of non-stationarity and the consequences for high-resolution paleoclimatology.