Hamid Moradkhani Alton N. Scott Professor of Civil and Environmental Engineering · Director, Center for Complex Hydrosystems Research, The University of Alabama

Home Contact

Downscaled Soil Moisture Active Passive (SMAP) Product @ 1-km Resolution using Machine Learning

A remotely sensed soil moisture data set at 1-km resolution, produced by machine-learning-based downscaling for hydrologic and water resources applications.

Abstract soft gradient of lavender and pale blue tones blending into deep purple, evoking a smooth continuous field of moisture data

Overview

The research group distributes a downscaled Soil Moisture Active Passive (SMAP) product at 1-km resolution developed using machine learning. The product provides remotely sensed soil moisture estimates at a spatial resolution finer than the native SMAP sensor footprint, supporting work on hydrologic modeling, data assimilation, drought assessment, and flood prediction.

Machine learning is applied within the group's broader research effort, which also covers predictive science, uncertainty analysis, data analytics, and high-performance computing.

Smooth muted green and slate blue gradient surface with faint diagonal banding, suggesting a gridded satellite data field
Fine-resolution soil moisture fields are consistent with the group's focus on data assimilation and remote sensing for hydrologic prediction.

Product Details

Attribute Description
Product Downscaled Soil Moisture Active Passive (SMAP) product
Resolution 1 km
Method Machine learning
Variable Remotely sensed soil moisture

The product description follows the methodology described in Abbaszadeh et al., 2019, Water Resources Research.

Applications

  • Data assimilation and remote sensing for hydrologic state estimation
  • Drought modeling and probabilistic drought monitoring
  • Extreme events and flood modeling
  • Hydrologic modeling and water resources system analysis

Related