Heistermann et al. (2026) Soil moisture monitoring with cosmogenic neutrons: an asset for the development and assessment of soil moisture products in the state of Brandenburg (Germany)

Identification

• Journal: Natural hazards and earth system sciences
• Year: 2026
• Date: 2026-01-23
• Authors: Maik Heistermann, Daniel Altdorff, Till Francke, Martin Schrön, Peter Martin Grosse, Arvid Markert, Albrecht Bauriegel, Peter Biró, Sabine Attinger, Frank Beyrich, Peter Dietrich, Rebekka Eichstädt, Jakob Terschlüsen, Ariane Walz, Steffen Zacharias, Sascha E. Oswald
• DOI: 10.5194/nhess-26-465-2026

Research Groups

• Institute of Environmental Science and Geography, University of Potsdam, Potsdam, Germany
• UFZ – Helmholtz Centre for Environmental Research GmbH, Leipzig, Germany
• Landesamt für Bergbau, Geologie und Rohstoffe Brandenburg, Cottbus, Germany
• Deutscher Wetterdienst, Meteorologisches Observatorium Lindenberg – Richard-Aßmann-Observatorium, Tauche, Germany
• Ministerium für Wirtschaft, Arbeit, Energie und Klimaschutz (MWAEK), Potsdam, Germany
• Landesamt für Umwelt (LfU) Brandenburg

Short Summary

This study introduces a new cosmic-ray neutron sensing (CRNS) soil moisture monitoring network in Brandenburg, Germany, and evaluates its initial 2024-2025 data against large-scale soil moisture products and a local hydrological model. It concludes that model-based products, particularly a calibrated local model, offer the highest potential to overcome the inherent spatio-temporal limitations of a sparse instrumental network for water risk management.

Objective

• To introduce a novel soil moisture monitoring network based on cosmic-ray neutron sensing (CRNS) in Brandenburg, Germany, including its design, evaluation, and initial operational results from 2024 to 2025.
• To assess how widely used large-scale soil moisture products (ERA5-Land, Copernicus Land Monitoring Service Soil Water Index, Copernicus Climate Change Service surface soil moisture) capture observed soil moisture dynamics compared to a local soil hydrological model (Soil-Water-Atmosphere-Plant, SWAP) set up with region-specific data.
• To investigate whether CRNS-based soil moisture estimates can improve these products, for example, through bias correction.
• To discuss the implications of this evaluation for the application of soil moisture products in managing water-related risks in Brandenburg, identifying products with the best prospects for different application contexts.

Study Configuration

• Spatial Scale: The federal state of Brandenburg, Germany (area of 29 479 km²). Eight CRNS monitoring sites were instrumented by June 2024, with four more by the end of 2025. CRNS sensor footprint radius is approximately 130–240 metres.
• Temporal Scale: CRNS data collection and evaluation from 2024 to September 2025. Model simulations and long-term analysis for soil water storage and groundwater recharge from 1993 to 2024.

Methodology and Data

• Models used:
  • Cosmic-Ray Neutron Sensing (CRNS) conversion function (general calibration based on Desilets et al., 2010, and Heistermann et al., 2024).
  • Soil-Water-Atmosphere-Plant (SWAP) model (van Dam et al., 2008) for local soil hydrological simulations.
  • ERA5-Land reanalysis (Muñoz Sabater, 2025) for large-scale soil moisture.
  • Soil Water Index (SWI) from Copernicus Land Monitoring Service (CLMS, 2025) for large-scale soil moisture.
  • Copernicus Climate Change Service (C3S) surface soil moisture product (Copernicus Climate Change Service, 2018) for large-scale soil moisture.
  • PARMA model (Sato, 2015) for spatial cosmic radiation correction.
  • ROSETTA pedotransfer function (Schaap et al., 2001) for estimating soil hydraulic parameters.
• Data sources:
  • CRNS observations from 8 stationary sites in Brandenburg.
  • Local meteorological sensors at CRNS stations (barometric pressure, temperature, humidity).
  • Conventional point-scale soil moisture sensors (SMT100, ThetaProbe ML2x) for additional reference.
  • Soil sampling campaigns for gravimetric soil water equivalents of soil organic carbon and lattice water, and soil dry bulk density (from upper 30 cm).
  • Dry above-ground biomass (AGB) density estimates.
  • Neutron monitor on the Jungfraujoch (JUNG) for temporal cosmic radiation correction.
  • Daily climate observations from the German Meteorological Service (DWD) (minimum/maximum air temperature, average relative air humidity, sunshine hours, average wind speed).
  • DWD’s radar-based quantitative precipitation product RADOLAN.
  • State’s soil map BUEK300 (LBGR, 2024) for soil texture classes.
  • Groundwater table depth data (LfU, 2013).
  • OpenStreetMap contributors (2024) for forest coverage.
  • Rail-based CRNS systems for potential future comparisons.

Main Results

• A CRNS-based soil moisture monitoring network was successfully established in Brandenburg, with 8 sites operational by June 2024 and 4 more by the end of 2025, providing openly accessible data.
• CRNS-based soil moisture estimates (θCRNS) using a general calibration function showed a root mean squared error (RMSE) of 0.037 m³ m⁻³ and a mean error (ME) of −0.025 m³ m⁻³ compared to ground-based reference measurements, indicating satisfactory agreement but a systematic underestimation.
• The network captured consistent and plausible seasonal soil moisture dynamics, illustrating significant spatial and temporal variability across Brandenburg. CRNS measurement depth dynamically varied between approximately 25 cm and 60 cm.
• Native large-scale soil moisture products (SWI, ERA5-Land, C3S) exhibited substantial systematic bias and generally lower performance compared to the local SWAP hydrological model.
• The local SWAP model significantly outperformed all other products in terms of percent bias (PBIAS), RMSE, and Nash–Sutcliffe Efficiency (NSE), both with and without bias correction/calibration. It demonstrated a strong ability to capture spatial variability, achieving an NSE of 0.68 for pooled data across all locations.
• A simple multiplicative bias correction substantially improved the performance of large-scale products, with ERA5-Land benefiting most and becoming superior to SWI and C3S after correction.
• Model-based products (ERA5-Land and, in particular, the calibrated SWAP model) were identified as having the highest potential to overcome the inherent limitations of a sparse instrumental network (e.g., short time series, limited spatial coverage, dynamic penetration depth) for water-related risk management.
• The calibrated SWAP model was successfully applied for long-term analysis (1993–2024) of soil water storage at different integration depths (0–50 cm, 0–100 cm, 0–150 cm) and for estimating groundwater recharge, highlighting strong spatio-temporal variability and the importance of accounting for shallow groundwater tables.

Contributions

• Establishment and detailed introduction of the first long-term, CRNS-based soil moisture monitoring network in Brandenburg, Germany, designed for regional representativeness and open data access.
• Independent evaluation of CRNS-based soil moisture estimates using a general calibration, providing a rare assessment of its accuracy without local calibration.
• Comprehensive benchmarking of widely used large-scale soil moisture products (ERA5-Land, SWI, C3S) against CRNS observations and a region-specific local hydrological model (SWAP) in a drought-prone region.
• Demonstration of the utility of CRNS data for evaluating and improving soil moisture products, particularly highlighting the potential of model-based products (especially a calibrated local model) to overcome the spatio-temporal limitations of instrumental monitoring for water risk management.
• Practical lessons learned from the establishment and first year of operation of a transdisciplinary monitoring network involving research institutions and state agencies.
• Provision of openly accessible observational and simulated soil moisture data to foster future applications and collaborations.

Funding

• Deutsche Forschungsgemeinschaft (Research unit FOR 2694 Cosmic Sense, project number 357874777)
• Ministerium für Landwirtschaft, Umwelt und Klimaschutz des Landes Brandenburg (project “Einfluss des Klimawandels auf die Grundwasserneubildung in Brandenburg: Anpassungsbedarfe und Hebelpunkte”)
• Helmholtz-Centre for Environmental Research GmbH (UFZ)
• University of Potsdam

Citation

@article{Heistermann2026Soil,
  author = {Heistermann, Maik and Altdorff, Daniel and Francke, Till and Schrön, Martin and Grosse, Peter Martin and Markert, Arvid and Bauriegel, Albrecht and Biró, Peter and Attinger, Sabine and Beyrich, Frank and Dietrich, Peter and Eichstädt, Rebekka and Terschlüsen, Jakob and Walz, Ariane and Zacharias, Steffen and Oswald, Sascha E.},
  title = {Soil moisture monitoring with cosmogenic neutrons: an asset for the development and assessment of soil moisture products in the state of Brandenburg (Germany)},
  journal = {Natural hazards and earth system sciences},
  year = {2026},
  doi = {10.5194/nhess-26-465-2026},
  url = {https://doi.org/10.5194/nhess-26-465-2026}
}