UN Handbook on Remote Sensing for Agricultural Statistics

The UN Handbook on Remote Sensing for Agricultural Statistics serves as a practical guide for applying remote sensing techniques to agricultural data. It addresses the increasing availability of big Earth Observation (EO) data from cloud-based services, emphasizing the transition from single snapshots to Satellite Image Time Series (SITS) for monitoring ecosystem dynamics and improving agricultural statistics.

This resource is designed to equip readers with the necessary methodologies for generating high-quality land use maps and crop yield predictions. A primary objective is to engage national statistical and mapping authorities, fostering a collaborative environment where these institutions can leverage EO data to enhance national estimates and support sustainable development initiatives.

The handbook is structured into major sections, covering both theoretical foundations and practical applications. The "Foundations" section delves into the technical aspects of satellite imagery, including optical and Synthetic Aperture Radar (SAR) imagery, Earth observation big data sources, and the use of data cubes. It also explores land cover and crop classification schemas, quality control for training sets, machine learning algorithms for image time series, spatial map uncertainty estimation, and map validation for area estimation. The "Use Cases in Crop Type Mapping" section provides real-world examples of crop classification across various regions.