Kwon et al. (2026) Synergistic impact of simultaneously assimilating radar- and radiometer-based soil moisture retrievals on the performance of numerical weather prediction systems

Identification

• Journal: Hydrology and earth system sciences
• Year: 2026
• Date: 2026-03-03
• Authors: Yonghwan Kwon, Sanghee Jun, Hyunglok Kim, Kyung-Hee Seol, In-Hyuk Kwon, Eunkyu Kim, Sujeong Cho
• DOI: 10.5194/hess-30-1261-2026

Research Groups

• Korea Institute of Atmospheric Prediction Systems (KIAPS), Seoul, South Korea
• Department of Environment and Energy Engineering, Gwangju Institute of Science and Technology (GIST), Gwangju, South Korea

Short Summary

This study evaluates the synergistic impact of simultaneously assimilating radar-based (ASCAT) and radiometer-based (SMAP) soil moisture retrievals into the Korean Integrated Model (KIM) using a weakly coupled data assimilation system. The findings demonstrate that multi-sensor soil moisture assimilation leads to more balanced and improved analyses and forecasts of specific humidity, air temperature, and precipitation compared to single-sensor assimilation.

Objective

• To evaluate the relative (individual) and combined performance of C-band radar-based (ASCAT) and L-band radiometer-based (SMAP) surface soil moisture products in improving global soil moisture analysis and atmospheric analysis/forecast via assimilation within the KIM-LIS coupled system.

Study Configuration

• Spatial Scale: Global domain, with a horizontal resolution of 25 km for the Land Surface Model (LSM) and KIM deterministic component, and 50 km for the KIM ensemble component.
• Temporal Scale: 6-hour cycling data assimilation experiments from March to July 2022, with the first month used for spin-up. 5-day forecasts were performed every 00:00 and 12:00 UTC cycle.

Methodology and Data

• Models used:
  • Korean Integrated Model (KIM) version 3.9 (global non-hydrostatic dynamical core with a cubed-sphere grid system).
  • National Aeronautics and Space Administration (NASA) Land Information System (LIS) version 7.4.
  • Noah Land Surface Model (LSM) version 3.3 (within LIS and KIM, 2 m total soil depth, 4 layers: 0.1, 0.3, 0.6, 1.0 m).
  • Ensemble Kalman Filter (EnKF) for land data assimilation (1-dimensional, ensemble size of 20).
  • Hybrid four-dimensional ensemble variational (hybrid 4DEnVar) data assimilation method with 4DIAU for atmospheric data assimilation.
• Data sources:
  • Assimilated Soil Moisture:
    • ASCAT (Advanced SCATterometer) near-real time soil moisture product (MetOp-B/C, 12.5 km swath grid, C-band 5.3 GHz, 0-2 cm soil depth, soil wetness index). Bias corrected using monthly Cumulative Distribution Function (CDF) matching.
    • SMAP (Soil Moisture Active Passive) Level 2 (L2) Radiometer Half-Orbit 36 km EASE-Grid soil moisture data (SPL2SMP version 9, L-band 1.4 GHz, 0-5 cm soil depth, volumetric soil moisture). Bias corrected using anomaly correction method.
  • Atmospheric Observations (for atmospheric DA): Advanced Microwave Sounding Unit-A (AMSU-A), Atmospheric Motion Vectors (AMVs), Microwave Humidity Sounder (MHS), Global Positioning System Radio Occultation (GPS-RO), Infrared Atmosphere Sounding Interferometer (IASI), Advanced Technology Microwave Sounder (ATMS), Cross-track Infrared Sounder (CrIS), surface, aircraft, and sonde observations.
  • Reference Data for Evaluation:
    • Soil Moisture: Triple Collocation Analysis (TCA) using ASCAT soil moisture Climate Data Record (CDR) version 7, SMOS-INRA-CESBIO (SMOS-IC) version 2, and AMSR2 X-band Land Parameter Retrieval Model (LPRM) product.
    • Specific Humidity and Air Temperature: ECMWF Integrated Forecasting System (IFS) analysis.
    • Precipitation: Gauge-based global daily precipitation analyses from the National Oceanic and Atmospheric Administration (NOAA) Climate Prediction Center (CPC).
  • Land Surface Inputs for Noah LSM: Shuttle Radar Topography Mission (SRTM) elevation, MODIS-IGBP land cover, National Centers for Environmental Prediction (NCEP) green vegetation fraction and surface albedo, blended State Soil Geographic (STATSGO)/Food and Agriculture Organization (FAO) soil texture.
  • Initial Conditions/Forcing: Global Land Data Assimilation System (GLDAS) meteorological forcing fields, KIM atmospheric forcing, ECMWF atmospheric reanalysis (ERA5).

Main Results

• Soil Moisture Analysis:
  • Single-sensor ASCAT data assimilation (SGAT) reduced global mean fractional Mean-Square Error (fMSE) by 4.0% compared to the control (CTL).
  • Single-sensor SMAP data assimilation (SGSP) reduced global mean fMSE by 10.5% compared to CTL.
  • Both single-sensor DA cases showed significant improvements in soil moisture estimates in croplands.
  • SMAP DA generally exhibited higher skill than ASCAT DA for soil moisture analysis.
• Specific Humidity and Air Temperature Analysis/Forecast:
  • ASCAT DA had more beneficial impacts on air temperature analysis, while SMAP DA had more beneficial impacts on specific humidity analysis.
  • Multi-sensor data assimilation (MTATSP) improved both specific humidity and air temperature analyses relative to CTL, compensating for degradations seen in single-sensor cases.
  • MTATSP achieved a more balanced improvement for both specific humidity and air temperature analyses compared to single-sensor DA.
  • The most pronounced synergistic improvements by simultaneously assimilating both soil moisture products were observed in the 2 m air temperature analysis and forecast, especially when both soil moisture products had a positive impact.
  • Soil moisture DA was more effective in improving 2 m air temperature than specific humidity, particularly for the 00:00 UTC cycle, and generally performed better in the Northern Hemisphere.
• Precipitation Forecasts:
  • Multi-sensor soil moisture DA (MTATSP) improved precipitation forecast skill, as measured by the Equitable Threat Score (ETS), across various precipitation intensity thresholds.
  • MTATSP showed higher ETS skill than CTL (up to 1.8%) and single-sensor DA (up to 2.4% relative to SGAT and 0.6% relative to SGSP).
  • The impact on precipitation frequency bias (FB) was marginal globally, with regional variations.

Contributions

• This study is among the first to investigate the combined use of ASCAT and SMAP soil moisture products in a land-atmosphere coupled data assimilation system (KIM-LIS), demonstrating their feasibility and synergistic impact on Numerical Weather Prediction (NWP) performance.
• It highlights the complementary advantages of radar-based (ASCAT) and radiometer-based (SMAP) soil moisture data, showing that their simultaneous assimilation leads to more balanced improvements in atmospheric variables (specific humidity, air temperature, precipitation) compared to single-sensor assimilation.
• The research provides insights into the regional and variable-dependent performance of different soil moisture products and their combined use, emphasizing the need for accurate observation error specification in multi-sensor data assimilation.
• The study suggests a flexible framework for incorporating various combinations of soil moisture data sources, which is beneficial for near-real-time operational forecast systems.

Funding

• Korea Meteorological Administration R&D project "Development of a Next-Generation Data Assimilation System by the Korea Institute of Atmospheric Prediction Systems (KIAPS)" (grant no. KMA2020-02211).
• National Research Foundation of Korea (NRF) funded by the Korean government (MSIT) (grant nos. RS-2025-24535700 and RS-2025-02363044).

Citation

@article{Kwon2026Synergistic,
  author = {Kwon, Yonghwan and Jun, Sanghee and Kim, Hyunglok and Seol, Kyung-Hee and Kwon, In-Hyuk and Kim, Eunkyu and Cho, Sujeong},
  title = {Synergistic impact of simultaneously assimilating radar- and radiometer-based soil moisture retrievals on the performance of numerical weather prediction systems},
  journal = {Hydrology and earth system sciences},
  year = {2026},
  doi = {10.5194/hess-30-1261-2026},
  url = {https://doi.org/10.5194/hess-30-1261-2026}
}