How to Download Satellite Data: A Practical Guide (2026)

Published July 13, 2026 · 9 min read · By Earth Analytics

Downloading satellite data sounds simple until you try it. There are dozens of portals, each with its own account system, file formats, and quirks. This guide covers the main free sources of earth observation data, which one to use for which job, and the mistakes that cost beginners the most time.

The three main free sources of satellite data

1. Copernicus Data Space Ecosystem (Sentinel satellites)

The European Union's Copernicus programme operates the Sentinel satellites and offers all data free of charge through the Copernicus Data Space Ecosystem. This is your source for:

• Sentinel-2 — 10 m optical imagery every ~5 days, ideal for vegetation monitoring (NDVI), agriculture, and land cover.
• Sentinel-1 — radar imagery that sees through clouds, used for flood mapping, ground deformation, and ship detection.
• Sentinel-3 — ocean and land temperature at coarser resolution.
• Sentinel-5P — atmospheric gases such as NO₂, CO, and methane for air quality work.

You can browse and download interactively, or use the free APIs (OData and STAC) to automate downloads. Registration is free.

2. USGS EarthExplorer (Landsat)

The USGS EarthExplorer portal is the home of the Landsat archive — the longest continuous satellite record of the Earth's surface, going back to 1972. Landsat 8 and 9 provide 30 m imagery every 8 days combined. Use Landsat when you need historical depth: it's the only free option for studying changes before 2015.

3. NASA Earthdata (MODIS, VIIRS, and hundreds more)

NASA Earthdata aggregates hundreds of datasets, most notably MODIS and VIIRS, which trade spatial resolution (250 m–1 km) for daily global coverage. This is the go-to source for daily vegetation, fire, snow, and temperature products at regional to global scales.

Which satellite should you use?

Step-by-step: downloading your first Sentinel-2 scene

1. Create a free account at the Copernicus Data Space Ecosystem.
2. Open the Browser, draw your area of interest on the map.
3. Filter to Sentinel-2 L2A (this is the atmospherically corrected product — almost always what you want).
4. Set a cloud cover filter (e.g., under 20%) and your date range.
5. Pick a scene and download it. Expect a ~1 GB zip file per scene.

Five pitfalls that catch beginners

• Downloading Level-1 instead of Level-2 products. Level-1 data isn't atmospherically corrected; your vegetation indices will be wrong. For Sentinel-2, always choose L2A.
• Ignoring clouds. An optical scene that is 60% cloud may be useless for your location even if the average looks fine. Always check the cloud mask over your specific area.
• Underestimating file sizes. A year of Sentinel-2 over one region can easily exceed 100 GB. If you only need values at a few points, downloading full scenes is enormous overkill — see our guide on extracting data at points of interest.
• Wrong projections. Satellite tiles come in UTM zones; your coordinates are probably in WGS84 lat/lon. Mixing them up puts your data kilometres off target.
• Manual downloads for time series. If you need more than a handful of dates, use the APIs or a cloud platform — clicking through a portal 200 times is not a workflow.

What if you just want the numbers?

Most people don't actually want satellite images — they want a table: this location, this date, this value. If that's you, the whole download-process-extract pipeline can be skipped entirely. Our next guide shows how to convert satellite data to CSV, or you can let us do it for you.