Go to main content

New technology improves the monitoring of harmful algae


Hege Skaar Lars Johan Naustvoll KI Alger montasje

Left: Engineer Hege Skaar monitors from land the “microscope” (on the right) that takes pictures of the algae in the seawater. Researcher Lars-Johan Naustvoll believes the innovative technology is a good supplement.

Photo: Naustvoll/Fagerbakke/IMR

“Submersible” microscope and artificial intelligence help marine researchers uncover harmful algal blooms.

At the Institute of Marine Research (IMR), new equipment provides new opportunities for even better algae monitoring.

“We are now installing advanced ‘microscopes’ that take pictures of microalgae found in seawater,” explains IMR researcher Lars-Johan Naustvoll. “Then we use artificial intelligence to identify the algae.”

Seawater is analysed – in the sea 

Researchers are busy testing the modern technology at the Flødevigen Research Station.

“This machine is set up to analyse directly from the sea. We do not need to collect water samples, bring them to the lab, and analyse them under a microscope.”

The seawater goes directly into the ‘microscope,’ and all the microalgae in the water sample are photographed inside the metal cylinder.

Artificial intelligence helps, so researchers can raise the alarm

“We use artificial intelligence to count the algae and identify them. This way, we know both which species are present and in what quantity. This allows us to detect the presence of harmful algae quickly and efficiently.”

“Thus, we can also quickly alert authorities and industries if we see a bloom of the more troublesome algae.”

Some algae are “problem algae”

Although most algae are completely harmless, some species can be harmful.

Early in the spring, algae bloom, which is the “kickstart” for the summer season in the sea, and the algae serve as food for many. The spring bloom in the sea is a natural and necessary phenomenon.

The “mix” of which algae bloom varies from place to place and year to year.

But if, after the spring bloom, there is a considerable proportion of a somewhat troublesome species, or if the quantity becomes too large, it means trouble.

A harmful bloom can discolour the water, algae can produce toxins, or use up the oxygen and thereby suffocate fish and other marine animals. The algae can also make shellfish toxic for us to eat.

In 2019, such a harmful bloom led to extensive salmon deaths in Northern Norway: What we know about the so-called "killer alga" in Northern Norway

 

Gul klump med svarte streker stikkende ut, på grå bunn
In 2019, a harmful bloom of Chrysochromulina leadbeateri led to extensive salmon deaths in Northern Norway. Photo: IMR

Does not replace water samples, but provides broader monitoring 

Both the quantity and “mix” of algae change rapidly, even within a week. Therefore, the innovative technology becomes a useful supplement to the more traditional water samples.

“This does not replace thorough analyses of water samples under a microscope, but we get information faster. A bloom happens so quickly that it is not possible to collect and analyse everything as fast using the ‘old’ method.”

“The ‘old method’ still has its advantages. Alongside the testing of artificial intelligence, human, experienced taxonomists are identifying microalgae in the images taken by the new machine. The foundation for the artificial intelligence is thus created by humans.”

“Today’s technology cannot identify all harmful algae; many are too small, and the machine cannot get good enough images of them. So, water samples and skilled professionals are still necessary,” explains the IMR researcher. “But together, these methods give us an increasingly better overview of algal blooms. This, in turn, makes us better able to provide good advice to management and industry more quickly.”