Global Evaluation of a Soil Moisture Retrieval Algorithm Incorporating Improved Dielectric and Surface Roughness Models Using Amsr2 Observations

Key Points

• The aim is to evaluate a new soil moisture estimation algorithm that incorporates improved dielectric and surface roughness models.
• Developed a new soil moisture retrieval algorithm based on JAXA's existing framework.
• Utilized ground observation data from 14 global sites for evaluation.
• Compared algorithm performance using metrics like cumulative distribution function and mean absolute error.
• Assessed four dielectric models and two surface roughness models.
• The new algorithm showed better accuracy than the traditional JAXA method in most regions except Mongolia.
• Some areas exhibited decreased accuracy, highlighting the need for region-specific roughness parameters.