Doctoral student Henri Riihimäki has studied how high-resolution remote sensing data can be gathered and utilised in research in the Arctic area.
Remote sensing can be used to gather data on e.g. the landforms and vegetation of the Earth.
Imagery from Kilpisjärvi region, located in Finnish Lapland, shows that drones are effective, precise, cheap, and agile tools for gathering data.
Riihimäki photographed three 6.25 hectare study areas with the drone. The device flew at an altitude of some 30 metres along a pre-programmed route, and took hundreds of pictures from each study area.
These pictures contain information related to the vegetation and topography of the areas.
– There is no other way to achieve such detail, with a resolution of up to a centimetre, from such a large area. You can even identify individual plants from the drone imagery, Riihimäki says.
Traditionally, remote sensing data has been gathered with the help of satellites or aeroplanes. Vegetation, in its turn, has been examined from smaller study plots.
The resolution of traditional satellite images is comparatively coarse. A single pixel can cover the same area size that Riihimäki’s drone photographed.
– One satellite pixel can contain a huge amount of variations in topography and vegetation, says Riihimäki.
In his newest study, he combines the material from drone imagery with various satellite images, derived from different sensors. This lets him study the vegetation at different scales, and examine what these satellite pixels contain. For example, drone imagery can used to estimate what is the fractional vegetation cover of each satellite pixel. Satellite images can then be used to study vegetation properties over larger areas.
Building a virtual world
The photography is only the start of the research.
A three-dimensional model can be built from the hundreds of drone pictures. This is possible since the same objects have been photographed from different angles, which allows 3D reconstruction of the landscape.
– If you put on your VR glasses, you can step into that landscape, as it were. The world created this way is almost photo-realistic. If you look at it from a suitable distance, you can even identify individual plants, says University Instructor Arttu Paarlahti , who heads the digital development at the Department of Geosciences and Geography.
The photos can also be made into a two-dimensional mosaic that resembles traditional aerial photographs. Compared to them, however, the drone photos are many times more accurate.
A good tool for follow-up research
Riihimäki specifically wanted to test the system in the Finnish Lapland.
– Arctic nature is vulnerable. The effects of climate change are seen in the north sooner than elsewhere. Big changes have already been observed, but there will be more. However, these areas are often hard to reach, and gathering data there is expensive and time-consuming, he explains.
In future, the data may help to follow the advance of climate change, but also follow the effects of reindeer husbandry or tourism on the environment of Kilpisjärvi.
– The same areas could be photographed by the same method in, for example, ten years’ time to see how it has changed, says Riihimäki.
The use of drones is not only suitable in geography, but in many other disciplines, as well. The images and 3D material could be utilised in e.g. urban research, geology, or even biology.
According to Paarlahti, drones are part of the teaching at the Department of Geosciences and Geography. First-year undergraduates are already taught how to use drones.
– This is one cost-effective research method among others. It is good to start practising it from the beginning, he says.
You might imagine that the drones used in research are especially expensive. However, Riihimäki and Paarlahti say they mainly use ordinary drones intended for the general market in their work.
– I’ve noticed that expensive devices aren’t necessarily much better than the ordinary ones when it comes to features or reliability. There is a variety of software available for cheap drones intended for recreational use, says Paarlahti.
The article Estimating fractional cover of tundra vegetation at multiple scales using unmanned aerial systems and optical satellite data, based on Riihimäki’s thesis, was published in the Remote Sensing of Environment in February.
The research has been funded by National Geographic Society and the Academy of Finland.
Estimating fractional cover of tundra vegetation at multiple scales using unmanned aerial systems and optical satellite data, Henri Riihimäki, Miska Luoto, Janne Heiskanen. https://doi.org/10.1016/j.rse.2019.01.030
Janne Heiskanen (PhD), Department of Geosciences and Geography, and Institute for Atmospheric and Earth System Research