By Jack Simkin
In 1972, the first planetary image of Earth was taken.
Known as the ‘blue marble’ shot, the impact of this image has been monumental.
For the first time, people realised Earth isn’t some endless resource - its potential had finite boundaries.
One of the key growth areas of satellite imagery is for conservation purposes.
Known as ‘remote sensing’, satellite imagery and Unmanned Aerial Vehicle imagery data sources are changing the way we view the world once again, particularly for conservation.
How do they work?
Most observation satellites function like our eyes.
The light we perceive is a combination of absorbed and reflected energy. Energy is transmitted from the sun, and various objects reflect and absorb energy in different wavelengths.
‘Visible’ light is around 400-700 nanometres (10-6m) on the electromagnetic spectrum.
That green leaf you see? That appears green because the leaf absorbs electromagnetic energy in the blue and red wavelengths, and reflects energy in the green wavelengths.
But there are far more wavelengths than the visible range. And that means that there is heaps of electromagnetic energy being absorbed and reflected beyond what we can see.
But satellite sensors can.
Microwave satellites can give topographic information to map out elevation. Thermal infrared sensors can provide information about vegetation health and soil moisture.
The cool thing about satellite observation is that energy absorption is as important as reflection.
Minor differences between how an object reflects and absorbs electromagnetic energy can be used to describe the characteristics of an object, known as its ‘spectral signature’.
These signatures have become the primary purpose of ‘hyperspectral’ satellite imagery, where thousands of data layers are collected per pixel.
With increasing technologies, there are more and more applications for remotely sensed data.
What can’t we monitor?!
Take the longest running earth observational satellite – the Landsat series. With a resolution of 30 metres, one pixel on an image represents 30m on the ground.
Great for large scale research, not for small-scale applications. Fine for monitoring tree cover, useless for counting trees.
But resolution is improving. And fast.
In 2017, a researchers used Worldview-3 satellite data to survey Albatrosses on two remote islands in the southern hemisphere.
With a resolution of 30cm, these results were comparable with previous human-led surveys.
These findings could pave the way for high frequency, low disturbance, cheap and effective animal and habitat surveys (providing the animals are big enough!)
Remember those spectral signatures I mentioned?
Well those may be a way around pixel resolution issues.
Understanding the unique spectral characteristics of habitats may identify locations for targeted field work. Instead of trekking through the jungle for months, research efforts can be focused on known spectral characteristics that indicate certain habitat types.
A new NovaSAR satellite to be launched later in 2017 will overcome some of these deficiencies, offering resolutions of 6-30m, and capable of penetrating cloud cover (a huge issue for most satellites).
The future for remote sensing conservation looks bright – but the image is still a little grainy to switch across completely just yet.
This post originally appeared in the University of Melbourne's Scientific Scribbles blog
Banner image: Sourced from NASA via Wikimedia