Skip to content

SOMOSPIE Documentation

SOMOSPIE: SOil MOisture SPatial Inference Engine

Welcome to the documentation for the SOMOSPIE Library!

SOil MOisture SPatial Inference Engine (SOMOSPIE) is a data-driven tool designed to improve the representation of soil moisture. It addresses the limitations of remote sensing (i.e., satellites), such as coarse resolution and spatial gaps, through various machine learning algorithms based on meaningful terrain features.

SOMOSPIE supports reproducibility, explainability, and portability of results, contributing to FAIR principles. Its use of containerization and methods usable with publicly available datasets allows users to use and experiment with SOMOSPIE easily.

🚀 Quick Start

🧰 Features

  • 🛰️ Retrieve/Download Satellite Data - Fetch coarse-resolution soil moisture datasets from satellites (ESA-CCI).
  • ✂️ Crop to Region of Interest - Define an area of interest for efficient and localized analysis.
  • 🔍 Predict Missing Values and Downscale: Fill spatial gaps and downscale to a finer resolution with various machine learning algorithms.
  • 📊 Analysis and Visualization: Generate statistical outputs and visual maps for interpretation and evaluation.