USGS EarthExplorer Tutorial: Download Free Landsat and Sentinel Imagery
- Primary Signal: USGS EarthExplorer Tutorial: Download Free Landsat and Sentinel Imagery
- Overview: EarthExplorer is the gateway to the largest free Earth observation archive, including Landsat back to 1972. This tutorial covers accounts, datasets, scenes and bulk downloads....
- Verification: Analyzed and compiled by FreeSky GeoData editorial monitoring desk.
If one website deserves a bookmark in every GIS analyst's browser, it is USGS EarthExplorer. It provides free access to Landsat, Sentinel, MODIS, ASTER, aerial photography and more than a hundred other datasets, forming one of the largest freely available Earth observation archives in the world. No subscriptions, no per-scene fees, no proprietary formats: search, filter, download and analyse.
Start by creating a free USGS account and logging in. Without an account you can search but not download, so registration is the first step. The interface looks dated, but it is powerful once you learn its layout: search criteria on the left, a map in the centre, and dataset selection panels that let you drill from broad collections down to specific products.
The search workflow has four steps. First, define your area: draw a polygon on the map, enter coordinates, or search for a place name. Second, set a date range, keeping it reasonably narrow for your first attempts. Third, choose datasets: Landsat Collection 2 Level-2 is the best choice for long-term land change, vegetation and thermal work, with 30-metre multispectral imagery and a 16-day revisit stretching back to 1972. Sentinel-2 sits under the Sentinel category and offers 10-metre resolution with a five-day revisit for more recent, detailed work.
Fourth, filter the results. EarthExplorer shows each scene's cloud cover, acquisition date and footprint, and you can preview thumbnails before downloading. For Landsat, choose Level-2 science products when you want surface reflectance and surface temperature ready for analysis; Level-1 products are uncorrected and need more processing. A scene with less than 10 percent cloud over your area of interest is usually a good pick.
Downloading a single scene is straightforward: add it to your cart or download directly, and EarthExplorer delivers a compressed archive. Landsat scenes arrive as separate GeoTIFF band files, while Sentinel-2 scenes arrive in the SAFE folder structure with JPEG2000 files. For many scenes, install the USGS Bulk Download Application, which queues downloads, resumes interrupted transfers and organises files into named folders automatically.
Once downloaded, load the bands into QGIS. For Landsat 8 and 9, a natural-colour composite uses Bands 4, 3 and 2, while a standard false-colour vegetation composite uses Bands 5, 4 and 3. Check the metadata file that comes with each scene: it records acquisition time, sun angle, cloud cover and processing level, all of which matter when comparing scenes across years.
A practical tip for time-series work: use the same dataset, product level and season across years to keep comparisons fair. Mixing Landsat 5 scenes from the 1990s with Landsat 9 scenes from today without accounting for sensor differences will produce apparent changes that are really sensor artefacts. Consistency is what turns a pile of scenes into a credible change analysis.
EarthExplorer rewards patience on the first visit and pays it back forever after. Learn the dataset tree, save your search criteria, and build a habit of downloading scenes as you need them. Within a few sessions you will be pulling decades of Earth history onto your own hard drive, free, for any project that needs it.
EarthExplorer also rewards users who learn its less obvious corners: the additional criteria panel lets you filter by cloud cover, day or night acquisition and processing level before you ever see a result list, and the metadata export gives you a spreadsheet of candidate scenes for planning a multi-year study. Spend one session exploring those filters and every later search gets faster and more precise, which matters when a project needs dozens of scenes.
📖 Continue Reading the Full Story
Continue reading this article on FreeSky GeoData.
⚡ Access Full Content Now