Cavity ring-down spectroscopy (CRDS) is rapidly becoming an invaluable tool to measure hydrogen (δ²H) and oxygen (δ18O)
isotopic compositions in water, yet the long-term accuracy and precision of this technique remain relatively underreported.
Here, we critically evaluate one-year performance of CRDS δ²H and δ18O measurements at ETH Zurich, focusing on high
throughput (~200 samples per week) while maintaining required precision and accuracy for diverse scientific investigations.
We detail a comprehensive methodological and calibration strategy to optimize CRDS reliability for continuous, high-throughput
analysis using Picarro’s “Express” mode, an area not extensively explored previously. Using this strategy, we demonstrate
that CRDS achieves long-term precision better than ±0.5‰ for δ18O and ±1.0‰ for δ²H (±1σ) on three United States
Geological Survey (USGS) reference materials treated as unknowns.18 Specifically, reported results for each reference material
over this one-year period are: (i) USGS W-67444: δ2H = −399.32 ± 0.96‰, δ18O = −51.07 ± 0.45‰ (n = 30), (ii)
USGS W-67400: δ2H = 2.55 ± 0.49‰, δ18O = −1.85 ± 0.13‰ (n = 140), and (iii) USGS-50: δ2H = 33.68 ± 0.91‰,
δ18O = 5.03 ± 0.04‰ (n = 21). We also address challenges such as aligning our analytical uncertainties with the narrower
uncertainties of International Atomic Energy Agency reference materials, and mitigating inherent CRDS issues like memory
and matrix effects when analyzing environmental samples. Our review provides a practical framework for CRDS applications
in hydrology, paleoclimatology, and biogeochemistry, underscoring the importance of continuous evaluation and methodological
refinement to ensure accuracy and precision in δ²H and δ18O analyses.18