Satellite-Based Data

Skylight is very dependent on satellite-based data to generate information for users. Space companies own many different types of satellites which organizations - like Skylight - can subscribe to for free or for a fee. Skylight does not own satellites itself.

Satellite-based data is any type of data that is collected by a satellite. Everyday examples include: 

• Cameras on satellites tracking temperature changes in the oceans

• “Satellite View” on Google Maps shows the Earth from satellite images 

• Tracking clouds from satellites to help predict the weather 

Here are examples more relevant to the marine context: 

1. Voluntary signals from ships, like AIS, VMS, and LRIT 

2. Involuntary signals from ships, like radio frequency (RF) and light emission (VIIRS)

3. Optical and radar imagery, which can capture objects in the sea like ships, structures, and pollution

#1 is unique in that there are usually enough satellites and/or a combination of receivers (satellite + terrestrial receiver + ship receiver) that make tracking a moving signal consistently possible. This is how Maritime Domain Awareness tools like Skylight can display vessel paths and determine behavior from the tracks. In comparison, #2 and #3 only provide snapshots of where vessels are a specific moment in time. For the rest of this page, we focus on #2 and #3. 

How do satellites collect this data?

Satellites collecting data are generally on pre-determined paths circling the globe. While they can be switched “on” and “off” at certain times and change what settings (“mode”) they use to collect the data, they cannot change their paths. That is why you cannot cue a specific satellite to collect information over a specific area at a specific time. However, as we add more satellites in space, the chances increase of there being a satellite over the area you are interested in at the time you are interested in. 

Compared to the past, we have many more satellites in space to gather information relevant to marine and maritime issues. Most of these satellites are owned by different companies and governments. Some make their data publicly available, such as SENTINEL-1 owned by the European Space Agency. Most satellite data is only commercially available, which means there is a fee to subscribe. These days, commercial companies offer packages where a single provider can coordinate with multiple satellite companies to provide the suite of data you are most interested in. 

Similarities among satellite-based data

There are a few things that satellites collecting information for use cases #2 and #3 have in common: 

• They have a ‘footprint’ or ‘collection area’ which is the area visible to the satellite during the snapshot.

Open

In Skylight, the footprints for satellite radar data appear as green boxes. 

• There is some delay introduced from the satellite collecting the information and the information becoming available to users.

  • This delay may be caused by the satellite needing time to reach a terrestrial station to “downlink” the information, weather, or other reasons.

• Algorithms are necessary to detect vessels. These algorithms might be simple (they are not always complex machine learning models), but the important point is that vessels are not detected automatically.

  • In the cases of RF and VIIRS, the radio and light emissions from ships must be separated from radio and light transmissions from other structures.

  • In the cases of optical and radar imagery, ships must be identified in the ocean, separate from waves, piers, and other natural and manmade objects. 

• When a vessel is detected, different types of information can be provided about the vessel depending on the type of satellite data.

  • Correlating these detections against the voluntary signals described in #1 can help identify “dark” vessels - vessels that are not transmitting their location on any known system available to the user.