Fischer et al. (2026) Quantifying evaporation of intercepted rainfall: a hybrid correction approach for eddy-covariance measurements

Identification

• Journal: Hydrology and earth system sciences
• Year: 2026
• Date: 2026-02-18
• Authors: Stefanie Fischer, Ronald Queck, Christian Bernhofer, Matthias Mauder
• DOI: 10.5194/hess-30-965-2026

Research Groups

• Technische Universität Dresden, Institute of Hydrology and Meteorology, Department of Meteorology, Tharandt, Germany

Short Summary

This study quantifies evaporation of intercepted rainfall at a coniferous forest site, revealing a systematic underestimation by eddy-covariance measurements (24% of precipitation) compared to model estimates (45%). A novel hybrid correction approach is proposed to reconcile eddy-covariance data with both energy and water budgets during interception events.

Objective

• To compare eddy-covariance (EC) measurements of evaporation with classical canopy water balance (WB) estimates during interception events.
• To assign EC data to a dynamic source area, considering stand heterogeneity, and integrate it with spot measurements from the WB method.
• To identify and implement the most effective method for adjusting and gap-filling latent heat flux (LE) measurements from EC under interception conditions.
• To assess the impact of the proposed adjustment on the overall water budget and the partitioning of precipitation within the study area.

Study Configuration

• Spatial Scale: ICOS site DE-Tha, a dense Norway spruce forest (72% Norway spruce, 15% Scots pine, 10% European larch, 1% birch) in Tharandt Forest, Germany. EC flux measurements at 42 m height. Throughfall collected by two 10 m troughs (total area 3.18 m²). Gross precipitation measured 130 m west of the tower in a 50 m x 90 m forest clearing. Model domain of 1140 m x 800 m with 10 m spatial resolution for vegetation structure (Plant Area Index, PAI). Dynamic EC flux footprint areas were considered.
• Temporal Scale: Data analyzed for the period 2008 to 2010. Throughfall measurements restricted to frost-free periods (typically April to September). Data resolution of 10 minutes for precipitation and throughfall, and 30 minutes for EC fluxes. Long-term climate records (1991-2020) were used for context.

Methodology and Data

• Models used:
  • 2D Rutter Model: A conceptual framework (Rutter et al., 1971) applied dynamically to calculate canopy water balance, incorporating spatially variable vegetation characteristics (PAI) at 10 m resolution. It uses the Penman-Monteith equation for evaporation.
  • Kljun et al. (2015) model: Used to calculate the extent and relative contribution of the source area (footprint) to EC flux measurements.
  • REddyProc: Software package used for gap-filling of latent heat flux data.
• Data sources:
  • Eddy-covariance (EC) measurements: Turbulent fluxes of sensible (H) and latent heat (LE) measured using an ultrasonic anemometer (SA-Gill-R3-50) and a closed-path infrared gas analyzer (LI-COR-LI7000).
  • Meteorological measurements: Global radiation (Rg), net radiation (Rn), air temperature (T_air), relative humidity (rH), and wind speed (u) at various heights.
  • Canopy water balance (WB) measurements: Gross precipitation (Pg) and throughfall (Tf) collected by trough systems.
  • Terrestrial laser scanning (TLS) and airborne laser scanning: Used to derive a 3D representation of the forest and Plant Area Index (PAI).
  • ICOS processing chain (Sabbatini et al., 2018) and EddyPro®: Used for post-processing and quality control of EC raw data.
  • FLUXNET and ICOS data platforms.

Main Results

• Eddy-covariance (EC) measurements showed a systematic underestimation of evaporation during and shortly after rainfall events, accounting for only 24% of precipitation for selected frost-free liquid interception events.
• The 2D Rutter model, incorporating spatially variable Plant Area Index (PAI) derived from laser scanning, provided reliable estimates of interception evaporation. For the EC footprint area, modelled interception evaporation accounted for 45% of precipitation for the evaluated events.
• Standard energy balance adjustment methods (Bowen ratio-based and energy balance residual approach) were found to be insufficient for correcting EC fluxes during interception events, particularly under high relative humidity (rH > 75%).
• A strong non-linear decrease in the latent energy ratio (LER = LEEC / LEEB) was observed for rH exceeding 75%, directly correlating with interception conditions.
• The proposed hybrid correction approach, which combines 2D Rutter model estimates for interception conditions with Bowen ratio-based adjustments for dry conditions, yielded an average annual total evaporation of 638 ± 16.4 mm. This value accounts for 59 ± 6% of annual precipitation and 68 ± 1% of available energy, significantly higher than the uncorrected EC measurements (375 ± 27.4 mm/year, 35 ± 3% of precipitation).
• The 2D model estimated that for the whole study domain, annual interception evaporation (EI) contributed 28 ± 3% and transpiration (ET) 29 ± 5% to mean annual precipitation, with EI and ET contributing 49 ± 1% and 51 ± 1% respectively to modelled total evaporation.

Contributions

• Developed and validated a novel hybrid correction approach for eddy-covariance latent heat flux measurements, specifically addressing the systematic underestimation during rainfall interception events.
• Integrated a spatially explicit 2D Rutter model, utilizing high-resolution terrestrial laser scanning data for detailed canopy structure (PAI), to provide independent and reliable estimates of interception evaporation for varying source areas (plot to EC footprint).
• Demonstrated that standard energy balance closure methods (Bowen ratio, residual) are inadequate for correcting EC fluxes under wet, high-humidity interception conditions.
• Provided a more consistent dataset of water and heat fluxes that simultaneously satisfies both energy and water budgets, thereby improving the parametrization of surface fluxes in weather and climate models and supporting water management under climate change.
• Highlighted the significant spatial heterogeneity of interception within a forest stand and the limitations of traditional plot-scale measurements for representing larger EC footprint areas.

Funding

• Deutsche Forschungsgemeinschaft (DFG) (grant no. BE-1721/23-1).

Citation

@article{Fischer2026Quantifying,
  author = {Fischer, Stefanie and Queck, Ronald and Bernhofer, Christian and Mauder, Matthias},
  title = {Quantifying evaporation of intercepted rainfall: a hybrid correction approach for eddy-covariance measurements},
  journal = {Hydrology and earth system sciences},
  year = {2026},
  doi = {10.5194/hess-30-965-2026},
  url = {https://doi.org/10.5194/hess-30-965-2026}
}