Improved Precision of Δ17O Measurements by Laser Spectroscopy

Stable water-isotope ratios reflect atmospheric distillation processes and are an important tool for understanding earth's hydrologic system. At polar precipitation sites, stable water-isotope ratios vary as a consequence of the integrated history of the entire hydrologic cycle, from evaporation at marine sites to condensation and precipitation at the poles. The novel measurement of Δ¹⁷O - which is defined as the deviation in δ¹⁷O from the global meteoric water line, ln(δ¹⁷O +1) - 0.528*ln(δ¹⁸O +1) - offers considerable potential for separating the various influences on water-isotope variations recorded in polar ice cores. However, generally low signal-to-noise ratios have complicated the interpretation of existing Δ¹⁷O records. Despite recent advances in laser spectroscopy that simplify the measurement of δ¹⁷O and the determination of Δ¹⁷O, noise associated with sample introduction can cause measurement errors that are comparable in magnitude to the natural variability of Δ¹⁷O. We use a Picarro L2140i in combination with a customized continuous-flow analysis (CFA) system to characterize measurement limitations for Δ¹⁷O in ice cores. We find that reducing variability in the humidity of the vaporized sample stream at the instrument inlet significantly reduces measurement noise and measurement errors for Δ¹⁷O. For continuous-flow measurement of ice cores, careful removal of particulates and bubbles is required to ensure that the sample vapor stream humidity is minimally affected. Additionally, routine cleaning or replacement of system components can prevent liquid flow rate inconsistencies from salt precipitates. This work shows that when a constant humidity stream can be generated at the instrument inlet, continuous-flow laser spectroscopy can achieve high-precision measurements (<6 per meg) for Δ¹⁷O.