Active sensor imagery, such as Synthetic Aperture Radar (SAR), provides recent images that can be extremely useful for remote trip planning. SAR imagery penetrates through clouds and does not require daylight, making it especially valuable at higher latitudes (Alaska!).
Remote sensing is categorized as passive or active depending on the energy source:
- Passive: Electromagnetic radiation from naturally occurring energy sources, like the sun, reflects off the surface of the Earth. A common sensor that records (passive-source) waves in the visible spectrum is a handheld camera. Other sensors and satellites record infrared, ultraviolet, and other wavelengths.
- Active: A satellite creates its own energy source, such as a radar pulse or laser (LIDAR), and then records the reflected radiation, which reveals information about the texture of the Earth’s surface.
Spectral imaging (passive sensors) in the visible light wavelengths is easy to interpret, the images are like what the human eye would see from space. Most of these satellites collect signals outside of the visible spectrum, and these wavelengths can be mapped to red-green-blue to provide information about snow, vegetation, land use, etc.
Popular sources for passive-source imagery include NASA’s MODIS satellites, Landsat 8 and Sentinel-2. Refer to Satellite Imagery: Near-realtime Sources for directions on how to use these free resources. Visible-light imagery requires daylight, and what you see is what you get… including clouds. Image collection for Sentinel-2, the highest resolution source, dips to lower latitudes mid-winter, barely reaching southern Alaska.
Active sensors, like the Synthetic Aperture Radar of Sentinel-1, can be used at all times of day, any season. In addition, the radiation penetrates clouds, meaning that all images are cloud-free.
Sentinel-1 resolution varies from 5-20m (for the IW mode), with a revisit time of 6-12 days (less at higher latitudes). Signal processing provides an image of “smoothness.” A mirror-like reflection on a smooth surface reflects the incident wave away from the satellite, whereas a rough surface reflects some of the backscattered wave back to the satellite.
Sentinel-1 has two polar-orbiting satellites, each with four acquisition modes. I don’t fully understand the differences, and I don’t think I need to. I found the best combination for my needs by trial and error:
- Mode: Interferometric Wide Swath (“IW”: analyzing the interference of two crossing waves)
- Polarization: VV (vertical transmission polarization, vertical reception)
The easiest way to visualize Sentinel-1 is with Sentinel Hub Playground. Select “Sentinel-1” from the list of available satellites, refer to the more detailed instructions available on my Satellite Imagery: Near-realtime Sources page.
Select Sentinel-1 from satellite dropdown
More advanced features are available via Sentinel Hub EO Browser, most notably, the ability to save the satellite image as a jpg or georeferenced image. Georeferenced images can be viewed in Google Earth, the Avenza App, or GIS software.
From the Search panel, select Sentinel-1. I don’t see any advantage to the Advanced search option.
From the Results panel, either select Visualize from the tabular results, or click on a blue ‘coverage’ region.
The tabular list includes information about the mode and polarization:
The Visualization tab contains an overwhelming number of options. The color default, “Enhanced visualization” is a little hard for my brain to interpret. I prefer the black and white “VV – decibel gamma0 – orthorectified” option. “Decibel gamma0” appears to provide a higher contrast image than linear, and orthorectified means each pixel is positioned correctly on the ground using a Digital Elevation Model.
Application: Remote Ice Skating
My first application for Sentinel-1 was an ice skate trip to the Arctic Circle. We ended up skating 100 miles in 1.5 days, and then hiking/skating 25 more over the next two days.
The Swedes have already documented the application of Sentinel-1 for nordic skating (thanks to Bob French for these resources), but I was curious to apply this tool to a truly remote, Alaska-style, expedition skate trip.
I was able to use both SAR and visible spectrum satellites during my planning, barely catching the final seasonal pass of Sentinel 2. You can watch the freeze-up progression of Kobuk and Selawik Lake in this gallery:
Black surfaces in the Sentinel-1 imagery means “smooth,” but that could be smooth ice, water, or snow. In this case I still had to rely on the weather history and Sentinel-2 / Landsat images to be convinced that the lakes were covered with ice and not snow.
For a remote trip like this, I want to have as much info as possible to accommodate multiple contingencies (“alternate finishes”). I use Gaia GPS for phone navigation (refer to my Gaia tutorial), and wanted to have a layer showing the “good ice” polygons.
After signing in to EO Browser (create a free account), you can save the imagery as a georeferenced image. The georeferenced image can be loaded into Google Earth, the Avenza App, or GIS software. I traced the “good ice” polygons, saved the result as a kml file, and loaded it into Gaia. I could not have been happier with how this worked. We used our phones to connect the areas of good ice, making for much faster and enjoyable travel than just taking a direct route through rough ice.
Traced “smooth ice” polygons, waypoints, and route options imported into Gaia GPS
Sentinel-1 was very useful for this trip and the only option for monitoring conditions since the trip (when Sentinel-2 was no longer collecting imagery at that latitude). However, I still needed to collect information from the weather history and visible spectrum imagery to help interpret Sentinel-1’s roughness scale. In the image below, the smoothest ice, far east, appears more gray than the rougher ice furthest west. I know from our flight that some of the black terrain is smooth snow, not ice, but I can’t suss that information out of the Sentinel-1 image alone. I need to do more ground truthing, gaining a better sense for what the Sentinel-1 roughness means on the ground.
A Comment on SAR for SAR (Search and Rescue)…
A few days after our trip, I heard from a friend in Kotzebue that a local had lost a snowmachine in overflow. I suspect Sentinel-1 might be very useful helping remote communities anticipate dangerous travel conditions. For example, we were told in Selawik to avoid the eastern shore of Selawik Lake because it often contains open water. The most recent Sentinel-1 imagery shows black on that shore, likely overflow. In this case, the locals know about the hazard, but they might not when traveling further from home. If only the State of Alaska could find someone to fund… someone to plan trips to remote areas and then compare ground textures with Sentinel-1… no one comes to mind.
Overflow? “Smooth” (black) texture on the eastern shore of Selawik Lake