SAFE (Sentinel Satellite Data Format)
Sep 24,2026

GISBox is a one-stop 3D GIS data editing, conversion and publishing platform that supports editing in multiple GIS formats such as OSGB/GEOTIFF/RVT, converting to 3DTiles/Terrain and publishing.

Introduction

SAFE (Sentinel Satellite Data Format) is a European standard archive format specifically designed by the European Space Agency (ESA) for the Sentinel series of satellites. It is essentially a directory folder encapsulating complete remote sensing data. It uses the XML-format manifest.safe as its core metadata entry point and internally integrates binary image data, preview images, KML geographic overlay layers, and various other file types. It offers excellent extensibility and can uniformly carry satellite products across the entire Sentinel series at different levels and in different modes. It is the universal standard format for ESA's Earth observation data archiving and distribution.

File Structure

The SAFE format is essentially a structured archive folder. Product structures vary slightly across different Sentinel satellites. The core components are described below:

Root directory core files: The root directory contains the manifest.safe file, which stores the product's general metadata in XML format, covering basic information such as sensor parameters, data acquisition time, and processing workflow. It also contains the main metadata file for the corresponding satellite product—for example, Sentinel-2's MTD_MSIL2A.xml, which is the core entry point for tools such as GDAL to read data and records key attributes including band information, projection parameters, and cloud cover percentage.
Image data subdirectories: For Sentinel-1, this directory is the measurement folder, which stores radar image data in GeoTIFF format. For Sentinel-2, the IMG_DATA folder under the GRANULE subdirectory stores multispectral band files in JPEG2000 format, organized into separate folders by resolution (10 m, 20 m, and 60 m), along with corresponding tile-level metadata files such as MTD_TL.xml.
Auxiliary and annotation subdirectories: Sentinel-1 includes an annotation subfolder storing annotation metadata such as calibration parameters and attitude information captured during satellite imaging. Sentinel-2 comes with directories such as DATASTRIP and AUX_DATA, which store quality-related files including cloud masks and atmospheric correction auxiliary data.
Preview and support subdirectories: The preview folder stores PNG-format quicklooks, KML overlay files for display in Google Earth, and HTML-format interactive preview pages. The support subdirectory stores extended support files required for XML parsing. Some products also include auxiliary content such as HTML directories and rep_info version descriptions.

Pros

Strong data self-containment: Image data, metadata, quality annotations, and preview images are all encapsulated within a single directory, eliminating the need to search for companion files separately and ensuring data integrity and traceability.
High degree of standardization: ESA uniformly defines the directory structure and XML metadata specifications, so all Sentinel products are logically consistent, reducing the learning curve for users.
Good extensibility: Different sensors (S1 SAR, S2 MSI, S3 OLCI, etc.) can all use the same framework, and future satellites can be seamlessly integrated.
Rich metadata: manifest.safe and the various MTD files record sensor parameters, orbit information, processing levels, and more, allowing users to quickly understand data quality and processing history.
Supports multi-resolution hierarchical storage: For example, Sentinel-2 stores data separately at 10 m/20 m/60 m, allowing users to read on demand and flexibly choose data volume.

Cons

Redundant directory structure and relatively large size: Contains numerous XML metadata files and preview files, with actual imagery accounting for a relatively low proportion, resulting in lower storage and transmission efficiency.
Insufficient native software support: Mainstream GIS software such as ArcGIS and QGIS cannot directly open SAFE directories. Tools such as GDAL or SNAP are required to first convert them into general formats like GeoTIFF.
Scattered multi-resolution files: Sentinel-2's different-resolution bands are dispersed across different subfolders, requiring additional scripting for band merging and resampling during batch processing.
Long processing chain: Extracting usable imagery from SAFE typically requires multiple steps (unpacking → reading metadata → extracting bands → format conversion), which is less direct than a single GeoTIFF file.
Complex naming conventions: File names contain extensive information such as product level, processing baseline, and timestamps, which is not beginner-friendly and prone to errors.

Application Scenario

As the official archive standard for Sentinel series satellite data, the SAFE format's most core application scenario is the global free remote sensing data distribution and archiving under ESA's Copernicus program. Sentinel-1 radar imagery, Sentinel-2 multispectral imagery, and Sentinel-3 ocean and land observation data downloaded by users through platforms such as the Copernicus Open Access Hub and SciHub are all delivered in SAFE format. It is widely used in environmental and Earth science research fields such as land use change monitoring, crop growth assessment, flood disaster emergency response, glacier retreat tracking, and ocean color remote sensing analysis. It is also the standard input format for data parsing and product generation in SNAP, ESA's official toolchain, as well as third-party remote sensing processing platforms such as GDAL and Sentinel Hub.