Lake oxygen isotopes as recorders of North American Rocky Mountain hydroclimate: Holocene patterns and variability at multi-decadal to millennial time scales

Abstract

Lake sediment oxygen isotope records (calcium carbonate-δ 18 O) in the western North American Cordillera developed during the past decade provide substantial evidence of Pacific ocean–atmosphere forcing of hydroclimatic variability during the Holocene. Here we present an overview of 18 lake sediment δ 18 O records along with a new compilation of lake water δ 18 O and δ 2 H that are used to characterize lake sediment sensitivity to precipitation-δ 18 O in contrast to fractionation by evaporation. Of the 18 records, 14 have substantial sensitivity to evaporation. Two records reflect precipitation-δ 18 O since the middle Holocene, Jellybean and Bison Lakes, and are geographically positioned in the northern and southern regions of the study area. Their comparative analysis indicates a sequence of time-varying north–south precipitation-δ 18 O patterns that is evidence for a highly non-stationary influence by Pacific ocean–atmosphere processes on the hydroclimate of western North America. These observations are discussed within the context of previous research on North Pacific precipitation-δ 18 O based on empirical and modeling methods. The Jellybean and Bison Lake records indicate that a prominent precipitation-δ 18 O dipole (enriched-north and depleted-south) was sustained between ~ 3.5 and 1.5 ka, which contrasts with earlier Holocene patterns, and appears to indicate the onset of a dominant tropical control on North Pacific ocean–atmosphere dynamics. This remains the state of the system today. Higher frequency reversals of the north–south precipitation-δ 18 O dipole between ~ 2.5 and 1.5 ka, and during the Medieval Climate Anomaly and the Little Ice Age, also suggest more varieties of Pacific ocean–atmosphere modes than a single Pacific Decadal Oscillation (PDO) type analogue. Results indicate that further investigation of precipitation-δ 18 O patterns on short (observational) and long (Holocene) time scales is needed to improve our understanding of the processes that drive regional precipitation-δ 18 O responses to Pacific ocean–atmosphere variability, which in turn, will lead to a better understanding of internal Pacific ocean–atmosphere variability and its response to external climate forcing mechanisms.