Overview

This project focuses on estimating soil moisture using satellite-based Sentinel-1 Synthetic Aperture Radar (SAR) data. Soil moisture is a key environmental variable for agriculture, water management, and climate studies.

We begin by showing how radar backscatter signals (VV and VH polarizations) are linked to ground soil moisture. Small, simplified examples are used to build understanding step by step.

Next, we apply retrieval methods such as Least Squares and Tikhonov regularization to estimate soil moisture across pixels. We then compare results and use parameter selection techniques like the L-curve and Generalized Cross Validation (GCV) to make the retrieval more stable and realistic.

Through tables, figures, and visualizations, the project demonstrates how radar data can be transformed into meaningful soil moisture information for environmental applications.

Suitable For

• BS Geoinformatics / Remote Sensing – satellite data processing and geospatial analysis

• BS Agriculture – soil monitoring for irrigation and crop health

• BS Hydrology / Water Resources – water balance and flood/drought studies

• BS Climate and Environmental Studies – soil moisture as a climate variable

Technologies Used

• Data Source: Sentinel-1 SAR dual-polarization data (VV, VH)

• Programming: Python

• Libraries: NumPy (data processing), Matplotlib (visualization), SciPy (inverse problem solving)

• Techniques:

  • Linking radar backscatter to soil moisture

  • Estimation using Least Squares and Tikhonov regularization

  • Optimal parameter selection with L-curve and GCV

  • Visualizing soil pixels and retrieved soil moisture values

Features

• Converts radar signals into soil moisture estimates

• Uses simple, step-by-step examples for clarity

• Compares different retrieval approaches

• Explains why regularization is important for reliable estimates

• Provides tables, diagrams, and figures for easy interpretation

• Highlights the practical value for agriculture and environmental monitoring

Deliverables

• Python code for soil moisture estimation

• Documentation with theory, worked examples, and results

• Figures showing feature matrices, soil pixels, and retrieval outputs

• Comparison tables of different estimation methods

• Optional interactive notebook for testing with different radar inputs

Benefits for Students

• Understand how satellite radar can be used to monitor soil conditions

• Gain hands-on experience in geospatial data analysis with Python

• Learn soil moisture retrieval techniques relevant to agriculture and hydrology

• Build a portfolio-ready capstone project connecting remote sensing, environmental science, and data analysis