Sodemann (2025) The Lagrangian moisture source and transport diagnostic WaterSip V3.2

Identification

• Journal: Geoscientific model development
• Year: 2025
• Date: 2025-11-24
• Authors: Harald Sodemann
• DOI: 10.5194/gmd-18-8887-2025

Research Groups

• Geophysical Institute, University of Bergen, Bergen, Norway
• Bjerknes Centre for Climate Research, Bergen, Norway

Short Summary

This paper introduces WaterSip V3.2, a diagnostic software tool that identifies evaporation sources and transport pathways of precipitation or water vapour based on Lagrangian model output. It provides a comprehensive reference for the method's foundations, technical setup, and interpretation, illustrated with a case study, to facilitate its wider application and inter-comparison with other moisture source diagnostics.

Objective

• To provide a consistent reference point for the use and interpretation of the Lagrangian moisture source and transport diagnostic WaterSip method and software, including its wide array of additional diagnostic output of moisture transport characteristics.

Study Configuration

• Spatial Scale: Regional to global domains (e.g., Northern Hemisphere, Scandinavia, Greenland, Antarctica, Amazon, Mediterranean, East Asian Summer Monsoon region). Output grids range from fine (e.g., 0.25° latitude/longitude) to coarse (e.g., 1.0° latitude/longitude).
• Temporal Scale: Event-scale to climatological studies, seasonal to inter-annual time scales, paleoclimate records. Trajectory lengths typically 5 to 20 days, with time intervals of 1 to 6 hours. The example case study covers a 10-day period (5 to 15 August 2022).

Methodology and Data

• Models used:
  • WaterSip V3.2 (diagnostic software tool)
  • Lagrangian trajectory models: LAGRANTO (Sprenger and Wernli, 2015)
  • Lagrangian particle dispersion models: FLEXPART (Stohl et al., 2005; Pisso et al., 2019)
  • Implemented diagnostic algorithms: Sodemann et al. (2008b), Stohl and James (2004), Gustafsson et al. (2010), Dirmeyer and Brubaker (1999).
• Data sources:
  • Reanalysis data (e.g., ERA5)
  • Numerical Weather Prediction (NWP) model output fields
  • Climate model data (e.g., CESM, NorESM)

Main Results

• WaterSip V3.2 is a C++ software with OpenMP parallelization, providing gridded output in netCDF format, and capable of processing large datasets of trajectories.
• It quantitatively identifies moisture source contributions to precipitation or water vapour, including geographic location and thermodynamic properties (e.g., temperature, relative humidity, d-excess) at the source, during transport, and at arrival.
• The software distinguishes between moisture uptakes within the boundary layer and in the free troposphere.
• It provides various output fields: moisture source footprints (total, boundary-layer, free-troposphere), particle mass density, moisture transport, air mass mixing, moisture rainout, Lagrangian precipitation estimates, accounted fractions, arrival conditions (temperature, pressure), and Lagrangian forward projections of source/transport properties (e.g., source latitude/longitude, distance, land fraction, skin temperature, d-excess, transport time, transport distance).
• A case study over Scandinavia (5–15 August 2022) illustrates these capabilities, showing dominant moisture sources from the North Atlantic and southern Scandinavia, with boundary-layer uptakes dominating.
• Sensitivity analysis reveals that the Lagrangian precipitation estimate is sensitive to the critical relative humidity threshold (RHc), while the total accounted fraction and source distance are sensitive to the uptake threshold (Δqc) and trajectory length. The time step (Δt) significantly impacts all these bulk properties.
• The software includes options for sectorisation of source regions and generation of histograms for detailed statistical analysis of uptake and transport properties.

Contributions

• Provides a comprehensive and consistent reference for the widely used Lagrangian moisture source and transport diagnostic WaterSip (V3.2), consolidating numerous technical and conceptual developments since its original publication.
• Makes the WaterSip software code publicly available, removing the need for re-implementation by other researchers and enabling community contributions.
• Offers practical guidance for setting up, running, and interpreting the diagnostic results, including a detailed example case study.
• Facilitates the comparison of the Sodemann et al. (2008b) algorithm with other moisture source diagnostics (Stohl and James, 2004; Gustafsson et al., 2010; Dirmeyer and Brubaker, 1999).
• Discusses key uncertainties and caveats of the method, supported by a sensitivity study, to promote valid and reliable interpretation of diagnostic results.

Funding

• Norges ForskningsrÃ¥d (grant no. 262710)
• European Research Council (ERC) within the EU's H2020 research programme (grant no. 773245)

Citation

@article{Sodemann2025Lagrangian,
  author = {Sodemann, Harald},
  title = {The Lagrangian moisture source and transport diagnostic WaterSip V3.2},
  journal = {Geoscientific model development},
  year = {2025},
  doi = {10.5194/gmd-18-8887-2025},
  url = {https://doi.org/10.5194/gmd-18-8887-2025}
}