Skip to main content
ARS Home » Northeast Area » Beltsville, Maryland (BARC) » Beltsville Agricultural Research Center » Hydrology and Remote Sensing Laboratory » Research » Research Project #446527

Research Project: Developing a Deep Learning-Based Framework for Root-Zone Soil Moisture Time Series Reconstruction

Location: Hydrology and Remote Sensing Laboratory

Project Number: 8042-13610-030-089-A
Project Type: Cooperative Agreement

Start Date: Sep 1, 2024
End Date: Aug 31, 2025

Objective:
The Cooperator will develop, test and apply a machine learning to track root-zone soil moisture at high-spatial resolution across the Central Valley of California. Application will include integration within an existing near-real-time system designed to output weekly soil moisture at five-day latency.

Approach:
The Cooperator has already developed a machine learning approach to describe the complex relationship between a range of environmental inputs data sources and root-zone soil moisture. The ARS PI has an existing operational soil moisture system for the Central Valley of California. Substantial interaction is required between the Cooperator and the ARS PI in order to integrate the two. In particular, the Cooperator and the ARS PI will work closely together to: i) adapt this existing model to work within irrigated agriculture sites, ii) additionally integrate remotely sensed evapotranspiration and surface soil moisture retrievals into the model, iii) improve model application at sites where it cannot be directly calibration, and iv) support the model's integration within an existing soil moisture monitoring system within the Central Valley of California.