While Radiant.Earth provides a box of powerful tools to use with imagery, there are some functions, such as classification, that require desktop software like QGIS or ERDAS Imagine. 

You can export imagery as it originally appears, color corrected, or with a number of indexes applied, as well as at different resolutions.

To export, go into a project, select Exports and New Export.

Then, click Define Area to Export and select New Area before clicking on the image to create points of a polygon. Click on the original point to close the shape, and then Save. You can repeat this process to include several areas in your export.

Then, choose your export settings (see below for details) and Start Export.

Export Settings

When exporting, you can choose from a number of settings to make the imagery work best for you.

Output Processing

This will affect the values of each pixel that is processed. 

If you want to use the images as they originally appeared, choose Raw which will export the bands as they originally were.

If you choose Color Corrected, an RGB GeoTIFF is created and exported using the color choices you made--either as a false-color image or corrected image. Note that these adjustments for display purposes only--you cannot use them for analysis in the same way you can with the raw imagery.

NDVI, NDWI, or NDMI use preset formulas to present images that give a sense of vegetation and water in an image. You can read more about them in Advanced Analysis.

Export to...

You can export either as a direct download to your computer, or to an S3 Bucket or Dropbox.

Approximate Resolution

You can choose the resolution that you want to export. This will resample the pixels from each image. While higher resolution is usually better, the original scenes may not have imagery available at high resolution--Landsat images are generally 15 meter resolution. Additionally, higher resolution leads to larger files. Especially if you are displaying a large area, you may want to consider using a lower resolution, which will allow for a more portable image.

Did this answer your question?