Dinh et al. (2026) Evaluation of ASCAT soil moisture retrievals and their potential to detect intraday variability

Identification

• Journal: Earth Observation
• Year: 2026
• Date: 2026-09-17
• Authors: Thi Lan Anh Dinh, Filipe Aires, Victor Pellet
• DOI: 10.5194/eo-1-105-2026

Research Groups

• Estellus (Paris, France)
• LIRA (Observatoire de Paris, Université PSL, Sorbonne Université, CNRS, Paris, France)
• LMD (École Polytechnique, Palaiseau, France)

Short Summary

This study introduces a localized convolutional neural network (CNN-l) framework that enhances soil moisture estimates from Advanced SCATterometer (ASCAT) observations by exploiting spatial features and adapting to local conditions. The proposed approach achieves strong agreement with ERA5 reference SM, even at a sub-daily scale.

Objective

• Evaluate the potential of deep learning to improve sub-daily soil moisture retrievals from ASCAT observations over the contiguous United States (CONUS).

Study Configuration

• Spatial Scale: Continental-scale study area covering CONUS.
• Temporal Scale: Hourly temporal resolution for SM retrieval and analysis.

Methodology and Data

• Models used: Localized Convolutional Neural Network (CNN-l) framework.
• Data sources: ASCAT level-2 observations, ERA5 reanalysis data, in situ soil moisture measurements from the International Soil Moisture Network (ISMN).

Main Results

• The CNN-l framework achieves strong agreement with ERA5 reference SM, even at a sub-daily scale.
• The model outperforms the ASCAT H120 product in terms of correlation and error levels when evaluated against in situ measurements.
• The study demonstrates the ability of the CNN-l approach to capture intraday soil moisture dynamics during heavy precipitation events.

Contributions

• This study provides a novel approach for improving sub-daily SM retrievals from ASCAT observations by exploiting spatial features and adapting to local conditions.
• The results highlight the potential of deep learning techniques in enhancing SM retrieval accuracy and capturing short-term hydrological responses.

Funding

• European Space Agency (ESA) Climate Change Initiative for Soil Moisture (ESA CCI SM)
• EUMETSAT H SAF program

Citation

@article{Dinh2026Evaluation,
  author = {Dinh, Thi Lan Anh and Aires, Filipe and Pellet, Victor},
  title = {Evaluation of ASCAT soil moisture retrievals and their potential to detect intraday variability},
  journal = {Earth Observation},
  year = {2026},
  doi = {10.5194/eo-1-105-2026},
  url = {https://doi.org/10.5194/eo-1-105-2026}
}