Stefan et al. (2021) High-Resolution SMAP-Derived Root-Zone Soil Moisture Using an Exponential Filter Model Calibrated per Land Cover Type

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Remote Sensing
• Year: 2021
• Authors: Vivien Stefan, Gianfranco Indrio, Maria‐José Escorihuela, Pere Quintana Seguí, Josep María Villar Mir
• DOI: 10.3390/rs13061112

Research Groups

Not explicitly detailed in the provided text. The study utilizes data and models associated with SMAP (Soil Moisture Active Passive) and the ISBA (Interaction Soil Biosphere Atmosphere) modeling community.

Short Summary

This study tests a simple recursive exponential filter, calibrated by land cover type using ISBA-DIF data, to derive 1 km resolution Root-Zone Soil Moisture (RZSM) from downscaled SMAP Surface Soil Moisture (SSM), finding strong validation results over rainfed crop sites in Catalonia, NE Spain.

Objective

• To test and validate a simple recursive exponential filter method, calibrated per land cover type using a long-term ISBA-DIF dataset, for retrieving high-resolution (1 km) Root-Zone Soil Moisture (RZSM) estimates from downscaled SMAP Surface Soil Moisture (SSM).

Study Configuration

• Spatial Scale: 1 km resolution; Regional scale (Catalonia, NE of Spain).
• Temporal Scale: Systematic basis (remote sensing); Long-term (implied by the use of the ISBA-DIF dataset for calibration).

Methodology and Data

• Models used: Recursive exponential filter (using time constant $\tau$); ISBA-DIF (Interaction Soil Biosphere Atmosphere—Diffusion scheme); DISPATCH (DISaggregation based on a Physical and Theoretical scale CHange).
• Data sources: SMAP (Soil Moisture Active Passive) Surface Soil Moisture (SSM) retrievals (downscaled to 1 km resolution via DISPATCH); Long-term ISBA-DIF dataset (used for filter calibration); Scaled in situ observations (at 5 cm, 10 cm, 25 cm, and deeper) over rainfed and irrigated crop areas (used for validation).

Main Results

• The RZSM estimates showed good agreement with observations over rainfed crops, particularly at shallower depths (10 cm and 25 cm).
• Rainfed Crops (10 cm depth): Nash–Sutcliffe (NS) scores ranged between 0.33 and 0.58. Correlation coefficients (R) ranged between 0.76 and 0.91.
• Rainfed Crops (25 cm depth): NS scores ranged between 0.37 and 0.56. Correlation coefficients (R) ranged between 0.71 and 0.90.
• Irrigated Sites (25 cm depth): Validation results were poorer, attributed partly to high heterogeneity. NS scores ranged between -2.57 and 0.16, and correlations (R) ranged between -0.56 and 0.48.
• Sensitivity analysis confirmed that increasing the time constant ($\tau$) led to increased NS scores and correlations with deeper soil layers in the rainfed site, suggesting better coupling.
• Strong correlation with very shallow depths (5 cm or 10 cm) in certain sites suggests the filter may lack the skill to fully represent deeper soil column processes, such as the effect of evapotranspiration in the profile.

Contributions

• Demonstrated the adequacy of calibrating the recursive exponential filter based on land cover type ($\tau$ per land cover) for deriving RZSM estimates over large, heterogeneous areas.
• Provided a validated methodology for generating high-resolution (1 km) RZSM estimates from downscaled SMAP SSM data.
• Quantified the performance differences of the RZSM retrieval method between rainfed and highly heterogeneous irrigated agricultural environments.

Funding

Not mentioned in the provided text.

Citation

@article{Stefan2021HighResolution,
  author = {Stefan, Vivien and Indrio, Gianfranco and Escorihuela, Maria‐José and Quintana‐Seguí, Pere and Mir, Josep María Villar},
  title = {High-Resolution SMAP-Derived Root-Zone Soil Moisture Using an Exponential Filter Model Calibrated per Land Cover Type},
  journal = {Remote Sensing},
  year = {2021},
  doi = {10.3390/rs13061112},
  url = {https://doi.org/10.3390/rs13061112}
}

Generated by BiblioAssistant using gemini-flash-latest (Google API)