What Kind Of Data?

An excerpt from Duncan Steel’s article on the NZ Aerospace Challenge. Duncan is from our supporting partner, Xerra (formerly known as Centre for Space Science Technology).

There is a vast, vast amount of satellite data available for teams who have bright ideas they’d like to trial and prove in the Challenge. Here I will give just a few examples.

Imagery is freely available for download from the Sentinel-2 satellites operated by the European Space Agency, with resolution (i.e. pixel size) 10 metres. Here is an image covering Christchurch and Banks Peninsula:

S2_20190213_wide.png

If imagery covering that whole area were made available, participants might use a lot of bandwidth downloading and then handling the data. In order to make things easier, and the files smaller, the good people at Spacebase chose a limited region around Lake Ellesmere, including parts of Banks Peninsula and also the Lincoln University Dairy Farm, which might be an area of interest in that soil measurements of various descriptions will surely be available; and in terms of water pollution, the seasonally-changing colours of Lake Ellesmere might be telling us something. The resultant reference area* is as shown in the next graphic.

(*There is nothing to stop participants from selecting another part of NZ for study; perhaps a region near where they live, or else a location where there are known problems, such as Lake Wakatipu.)

That set of six images obviously indicates that Lake Ellesmere changes colour over the months and years. The next graphic is intended just to show that frequent image collection occurs:

Most of those thumbnail images were obtained between February 2018 and the same month this year; just a few at the end were added from 2016 and 2017 so as to indicate that lots more historical images are available.

Now, the above frames were essentially showing ‘true-colour’ images (red/green/blue) made up from data collected in the visible spectrum. In fact there are many wavelengths of electromagnetic radiation that our eyes cannot detect directly, and yet we can operate sensors on satellites that will pick them up. The Sentinel-2 satellites (there are two of them, A and B, covering the whole globe every five days) have a capability to collect images covering 13 different bands, across the visible and then out into the near infrared (IR) and thence the short-wave IR. At right is a full set, greyscales indicating the intensity in each band. It is obvious that the fluxes depend heavily on the wavelength, with Lake Ellesmere appearing bright at the blue end of the spectrum (i.e. it reflects a lot of blue sunlight) but dark in the IR.

 

 

 

 

 

 

 


 Digital elevation data

A rather different type of measurement that satellites can assist in making is the elevation above sea-level of points beneath. We call this ‘terrain’ data, and often data files are labelled as ‘DEM’ (digital elevation model). Following is a graphic showing terrain data obtained using three different satellites: one radar (carried on the Space Shuttle) and two using optical methods whereby pairs of images can be cross-correlated and the parallax measured can be interpreted to deliver DEM data (just as your two eyes enable you to determine depths and distances).

The resolution in these datasets is not very detailed, with 30-metre height elements (hexels). However, lidar (laser radar) data collected from aircraft on behalf of Land Information New Zealand (LINZ) has resulted in terrain data with one-metre hexels across at least part of this reference area:

 Many other forms of remote sensing data

The above examples pertain to a small subset of all the datasets available, from many satellites and other sources including aircraft and drones. Satellite imagery with better spatial resolution than the above will be available to Challenge participants courtesy of Airbus, and that company also is responsible for the commercial operation of several synthetic-aperture radar (SAR) satellites built by national space agencies in Europe, SAR data being able to deliver different insights as to what is happening at ground level than optical (visible/infrared) satellites.

The story I am trying to tell here – or at least its opening chapter or two – is that there are many sorts of data available to Challenge participants, and much that can be done.


You can read the full article here.