When searching for satellite imagery the date of acquisition is crucial to the return of accurate results whether this be year, month, day or time the sensing took place.
As a user, you may be looking for coverage over an event that took place on a specific date at a particular time. Unfortunately, not all suppliers include the time of acquisition in their products’ metadata. Our platform, EarthImages hosts imagery from over 176 satellites distributed amongst several different suppliers, which can lead to a large number of images missing crucial information we feel to be paramount. If our users are unable to determine the period of acquisition, this often triggers a request via our contact channels for further information. Previously, we were able to give a rough estimation based on the orbital path of the satellite; Now, based on the recent implementation of a new Geocento developed innovation, we can go one step further.
In order to improve the accuracy of some of our records we first looked at sourcing the Sun’s azimuth angle, the direction of the sun in the horizontal plane from any point on Earth, on any day and at any time of the year. This is a very straightforward function which can be developed using resources available online, such as those available here. This information was then integrated with our Daylight Terminator, a tool found by clicking on the sun icon on the map toolbar within EarthImages.
The daylight terminator displays the limit between night and day on the Earth’s surface at any given day and time of the year. This is handy to visually get an idea of the sun conditions on the day of image acquisition. The terminator is updated when the time changes, for instance, if you click on a product it gets automatically updated to fit the day and time of the acquisition of the product (where available).
From there, if you know the position of your satellite, the sun zenith angle and the day of the year the image was acquired, you can work out the actual time of day the scene was captured. We implemented the reverse function and ran many tests and found it to be accurate within 5-10mns. Being confident in the reliability of the result, we’ve run a campaign to augment our meta catalogue with this extra piece of information. Alongside this, we’ve also integrated the ability to determine the local time at the place of acquisition, want to know how? Look out for our next Blog Post!