Since people have been asking me how I put together my imagery, I’ve written this tutorial to find, download and process Landsat 8 satellite data, along with some explanation of the image processing techniques behind what you will be doing. There are a lot of great tutorials out there, but most require Photoshop at a certain stage. I’ve spent some time put together an app you can download for free (Mac only for now) that bypasses the need for any fancy software, automatically merging different satellite bands, pan sharpening the image from 30m to 15m resolution, and adjusting the input values such that the output image can be easily edited for final color corrections in any basic image editing program.
Landsat 8 gathers an immense amount of data every day that have been transformed into some of the most stunning images of our planet. The best part is that anyone can explore the Earth with this satellite.
1. Get Data:
Go to www.earthexplorer.usgs.gov. This has a free database of all sorts of satellite and remote sensing data and imagery; you will have to make a free account to get download access. Today we will focus on Landsat 8.
On the main page, there are 4 tabs on the left. Starting with “Search Criteria”, you can specify dates, locations, etc. They key here is to narrow your geographical search enough to not overload the system in returning loads of data. Use the map to zoom to somewhere on Earth you are interested in, then under “Polygon” on the left, select “Use Map Extent”. This will define your region of interest as the area you see on your map on the right.
Advance to the next tab, “Data Sets”. Select “Landsat”, then “Landsat Collection 1 Level 1”, then finally “Landsat 8 OLI/TIRS”. OLI & TIRS are the imaging instruments aboard the satellite.
Advance to the next tab, “Additional Criteria”. Here, you may want to consider defining “Cloud Cover” to be below 10%, unless you want your final results to be bloated with too many choices. Once done here, select “Results”!
Scroll through the tiles, press the preview buttons, etc. Once you find a scene you like, press the download icon. Choose “Level-1 GeoTIFF Data Product”.
2. Edit Your Imagery:
The following steps normally involve software such as Photoshop or QGIS. There are plenty of tutorials out there for these tools. If you want control of the entire process and have Photoshop, this is a great one.
Otherwise, I made a Mac app that you can download from here: https://github.com/derekpickell/LandsatProcessing/releases. Click “Assets” and download the .app zipped file. *OS Catalina no longer allows apps to open not from the App Store: the way around this is to navigate to the app in Finder, right click the app icon, select open, then accept the message. Simply clicking from the dock or Desktop won’t work.
Open the app (it takes a while), and select the folder location of your downloaded (and unzipped) data. It will output a true color, pan sharpened image. It has to process a lot of data, so it’s best to not run a million things on your computer at the same time. The spinning wheel of death may also appear—just be patient.
From here, the output should be in a state where you can continue editing in your software of choice (Apple Photos, Preview, etc.). It’ll take some adjustments to make things look “natural” as atmospheric properties will make the image a bit more blue than an equivalent photo taken from closer to Earth. You will likely want to further crop the scene to an area of interest to help reduce file size.
What does this app do, specifically?
While our day-to-day cameras capture just about the entire visible spectrum (from red to violet), Landsat records data into smaller bands (chunks of light), which is much more useful for research. To produce a true color image, we take Landsat’s red, green and blue bands and combine them together into a full spectrum visible image. B4 is red (0.64–0.67 µm), B3 is green (0.53–0.59 µm) and B2 is blue (0.45–0.51 µm) if you are curious.
Out pops a true color image that encompasses the colors we are used to seeing (the visible spectrum). Behind the scenes, this app also performs some limited histogram clipping and doubles the resolution by “pan sharpening”. This technique overlays Landsat’s Band 8, which has twice the resolution as the others, with our true color image, effectively doubling our resolution! You can do this in Photoshop here.
Ultimately, Landsat 8 is exciting because it offers us weekly views of just about the entire planet. Its data is free, the views are stunning, and anyone can engage with the our planet.
*Feel free to make suggestions for my app. It’s certainly a work in progress, but it’s my hope that it bypasses the Photoshop finessing that is otherwise required—allowing us to get straight to more artistic editing.
Satellite data adapted from ESA’s Copernicus program and NASA/USGS’s Landsat 8. CC BY-SA 3.0 IGO.