Wang et al. (2026) Revisiting isotopic time series and linkages between precipitation and water vapor: High-resolution insights in a semi-arid setting

Identification

• Journal: Journal of Hydrology Regional Studies
• Year: 2026
• Date: 2026-02-11
• Authors: Zhilan Wang, Mingjun Zhang, Shengjie Wang, Lingling Liu, Cunwei Che, Yao Lu
• DOI: 10.1016/j.ejrh.2026.103225

Research Groups

• College of Geography and Environment Sciences, Northwest Normal University, Lanzhou 730070, China
• Key Laboratory of Resource Environment and Sustainable Development of Oasis, Lanzhou, Gansu Province 730070, China

Short Summary

This study investigates the dynamic variations and linkages between atmospheric vapor and precipitation stable isotopes (δ18O) at high resolution in the semi-arid Western Loess Plateau, China, using three years of continuous observation data. It reveals significant temporal variations, a strong precursor effect of vapor isotopes on precipitation, and distinct controlling factors (temperature, moisture transport, and convection) across different monsoon periods.

Objective

• To clarify the temporal variations of precipitation and vapor isotopes across different timescales using high-frequency isotope data, and to explore the relationship between precipitation and vapor isotopes using both equilibrium and non-equilibrium models.
• To identify the key controlling factors of regional isotopic variations from multiple perspectives, including local environmental factors, moisture sources, transport pathways, and convective activities.

Study Configuration

• Spatial Scale: Western Loess Plateau, China, centered at Dingxi station (104.59°E, 35.56°N, 1928.39 m elevation).
• Temporal Scale: Continuous observations from December 20, 2021, to October 30, 2024 (approximately 3 years). Vapor isotope data recorded every second, event-based precipitation samples collected at 30-minute intervals, meteorological data at 5-minute intervals, and daily/monthly scale analyses.

Methodology and Data

• Models used:
  • Lagrangian backward trajectory HYSPLIT (Hybrid Single-Particle Lagrangian Integrated Trajectory) model
  • Equilibrium fractionation model
  • Non-equilibrium fractionation model (incorporating Local Evaporation Line - LEL)
  • Hydrological calculator software (for non-equilibrium model)
• Data sources:
  • Continuous in-situ stable isotope observations of near-surface atmospheric vapor (Picarro L2130-i liquid water isotope analyzer) and precipitation (T-LWIA-45-EP liquid water isotope analyzer, NSA 181/S automatic precipitation sampler) at Dingxi station.
  • Integrated automated weather station data (temperature, precipitation, relative humidity) at 5-minute intervals.
  • ERA5 reanalysis dataset (specific humidity, zonal and meridional wind components at 850 hPa).
  • NOAA database for daily outgoing longwave radiation (OLR) with 1° × 1° spatial resolution.
  • NOAA Global Data Assimilation System (GDAS1) meteorological fields for HYSPLIT model.

Main Results

• The average δ18Ov was −18.03 ‰ and δ18Op was −7.93 ‰, with δ18Op consistently more enriched than δ18Ov.
• On a monthly scale, δ18Ov enriched from January to May, peaked in June, then fluctuated downward until December. On an hourly scale, δ18Op fluctuated significantly, while δ18Ov remained stable.
• The strongest correlation between δ18Ov and δ18Op occurred on the day of precipitation (r = 0.55, p < 0.05).
• δ18Ov showed a significant precursor effect on δ18Op, with the strongest correlation at a lag of −2 days (r = 0.311, p < 0.001). The δ18Op signal continued to modulate δ18Ov for approximately +3 days after the event.
• The non-equilibrium fractionation model (δ18Onon-equ = −17.99 ‰) provided a better simulation of observed vapor δ18O (δ18Oobs = −16.43 ‰) compared to the equilibrium model (δ18Oequ = −18.17 ‰), indicating significant kinetic fractionation effects.
• Temperature significantly impacted isotopes during the winter and transitional monsoon periods, while the summer monsoon period was mainly controlled by moisture transport processes, leading to gradual δ18O depletion due to rainout effects.
• Strong convection in upstream regions (indicated by low OLR values) led to isotopic depletion in the study area. Local convection intensity showed a significant positive correlation with δ18Ov from lag −5 days to −1 day, with the strongest correlation at lag −1 day (r = 0.299, p = 0.0002).

Contributions

• Provides the first systematic investigation of dynamic variations and linkages between vapor and precipitation isotopes at high-resolution (hourly to monthly) over a three-year period in the semi-arid Western Loess Plateau, addressing a critical data gap.
• Quantifies the temporal lag relationships between precipitation and atmospheric vapor isotopes, revealing a significant precursor effect of vapor on precipitation and a feedback effect of precipitation on vapor, highlighting an isotopic "memory effect" in the local moisture-recycling system.
• Systematically identifies and quantifies the multi-scale controlling factors (local meteorological conditions, large-scale moisture transport, and convective activities) influencing isotopic variations in an arid-semi-arid transition zone.
• Compares equilibrium and non-equilibrium fractionation models against high-resolution observations, confirming the prevalence of kinetic fractionation processes in the regional water cycle.
• Offers a theoretical basis and scientific support for understanding water cycle response mechanisms under climate change and for regional water resource management in arid-semi-arid regions.

Funding

• National Natural Science Foundation of China (42071047)
• Basic Research Innovation Group Project of Gansu Province (22JR5RA129)

Citation

@article{Wang2026Revisiting,
  author = {Wang, Zhilan and Zhang, Mingjun and Wang, Shengjie and Liu, Lingling and Che, Cunwei and Lu, Yao},
  title = {Revisiting isotopic time series and linkages between precipitation and water vapor: High-resolution insights in a semi-arid setting},
  journal = {Journal of Hydrology Regional Studies},
  year = {2026},
  doi = {10.1016/j.ejrh.2026.103225},
  url = {https://doi.org/10.1016/j.ejrh.2026.103225}
}