Valk et al. (2025) Evaporation measurements using commercial microwave links as scintillometers

Identification

• Journal: Hydrology and earth system sciences
• Year: 2025
• Date: 2025-11-21
• Authors: Luuk D. van der Valk, Oscar Hartogensis, Miriam Coenders‐Gerrits, Rolf Hut, R. Uijlenhoet
• DOI: 10.5194/hess-29-6589-2025

Research Groups

• Department of Water Management, Delft University of Technology, Delft, the Netherlands
• Meteorology and Air Quality Group, Wageningen University & Research, Wageningen, the Netherlands

Short Summary

This study proposes and evaluates a novel, opportunistic method to estimate evaporation using commercial microwave links (CMLs) as scintillometers. It demonstrates the feasibility of using CMLs to estimate 30 min latent heat fluxes and daily evaporation, with the energy-balance method showing promising performance, though highly dependent on the quality of net radiation estimates.

Objective

• To obtain 30 min latent heat fluxes (LvE) and daily evaporation (E) estimates using a former commercial microwave link (CML) in combination with scintillometry, evaluating different correction methods, flux estimation methods (two-wavelength, energy balance), and Monin-Obukhov similarity theory scalings (complete and free-convection).

Study Configuration

• Spatial Scale: Measurements were conducted over an 856 meter path at the Ruisdael Observatory near Cabauw, the Netherlands. Instruments were installed at 10 meters (CML, scintillometers) and 3 meters (eddy-covariance system) above the surface, predominantly over grass fields.
• Temporal Scale: Data were collected from 1 April to 1 October 2024, corresponding to a full growing season. CML signals were sampled at 20 hertz (Hz), microwave and optical scintillometers at 1 kilohertz (kHz), and the eddy-covariance system at 10 Hz. Analysis was performed on 30 minute intervals for latent heat flux and aggregated to daily intervals for evaporation.

Methodology and Data

• Models used:
  • Scintillometry theory for deriving turbulent characteristics from signal fluctuations.
  • Two-wavelength method to retrieve structure parameters of temperature (CT T) and humidity (Cqq) from combined microwave and optical scintillometer signals.
  • Energy-balance method (EBM) using net radiation as a constraint to separate structure parameters into turbulent heat fluxes.
  • Monin-Obukhov Similarity Theory (MOST) for relating structure parameters to turbulent heat fluxes.
  • Free-convection scaling as a limit of MOST, eliminating the need for horizontal wind speed measurements.
  • Constant noise correction method and spectral noise correction method for processing CML signals to estimate the structure parameter of the refractive index (Cnn).
• Data sources:
  • A Nokia Flexihopper Commercial Microwave Link (CML) operating at 38 gigahertz (GHz).
  • A Radiometer Physics RPG-MWSC-160 microwave scintillometer (160.8 GHz) and a Kipp & Zonen LAS Mk-II optical scintillometer (352.7 terahertz).
  • An eddy-covariance (EC) system (Gill-R50 sonic anemometer and LICOR-7500 H2O/CO2 sensor).
  • In-situ meteorological measurements (temperature, humidity, horizontal wind speed, net radiation, ground heat flux) from the KNMI Data Platform.
  • Satellite Application Facility on Land Surface Analysis (LSA SAF) of EUMETSAT for satellite-derived radiation products (incoming shortwave and longwave radiation, albedo, emissivity, land surface temperature) to estimate net radiation.

Main Results

• All CML methods, particularly with the spectral noise correction, successfully estimated 30 min latent heat fluxes (LvE) and daily evaporation rates (E).
• The energy-balance method (EBM) using observed net radiation (EBM-OBS) generally outperformed the two-wavelength method for LvE estimates, exhibiting lower mean bias error (MBE) and interquartile range (IQR). However, EBM-OBS significantly overestimated sensible heat flux (H).
• The spectral noise correction method consistently improved the IQR and Pearson's correlation coefficient (r) for LvE estimates compared to the constant noise method. For instance, the IQR for the two-wavelength method compared to the reference MWS-2λ method decreased from 154 watts per square meter (W m⁻²) to 134 W m⁻².
• The EBM using LSA SAF net radiation (EBM-LSA) showed higher MBE and IQR than EBM-OBS, highlighting the strong dependence of this method on the accuracy of net radiation estimates.
• Application of free-convection scaling for the two-wavelength method resulted in an average reduction of latent heat flux estimates by 40 W m⁻², attributed to neglecting shear-driven turbulence. For EBM versions, free-convection scaling led to an increase in LvE estimates and a reduction in H.
• Aggregation of 30 min LvE estimates to daily evaporation rates (m s⁻¹) maintained comparable performance patterns across methods.

Contributions

• This study demonstrates the viability of using commercial microwave links (CMLs) as opportunistic scintillometers to estimate 30 min latent heat fluxes and daily evaporation, building upon prior Cnn correction methods.
• It provides a comprehensive evaluation and comparison of conventional two-wavelength and standalone energy-balance methods for CMLs against established reference instruments (scintillometers and eddy-covariance systems).
• The research quantifies the impact of different CML noise correction techniques and Monin-Obukhov similarity theory scalings on latent heat flux estimates.
• It highlights the potential of CML networks for large-scale evaporation monitoring in hydrological applications, while also identifying critical dependencies on the quality of net radiation data and CML sampling strategies for future development.

Funding

• Ruisdael Observatory (Dutch Research Council (NWO), grant number 184.034.015).

Citation

@article{Valk2025Evaporation,
  author = {Valk, Luuk D. van der and Hartogensis, Oscar and Coenders‐Gerrits, Miriam and Hut, Rolf and Uijlenhoet, R.},
  title = {Evaporation measurements using commercial microwave links as scintillometers},
  journal = {Hydrology and earth system sciences},
  year = {2025},
  doi = {10.5194/hess-29-6589-2025},
  url = {https://doi.org/10.5194/hess-29-6589-2025}
}