Scientists copied the neural networks of the brain to map phytoplankton types in the Mediterranean Sea. A new study published in the Journal of Geophysical Analysis: Oceans showed a new method of classifying phytoplankton that depends on artificial intelligence clustering.
Phytoplankton blanket surface waters of the world’s oceans and pigments of their cells absorb specific wavelengths of sunshine, like the chlorophyll that provides plants their green color. Viewed from space, the color of the ocean’s floor modifies depending on the phytoplankton growing there. In the Mediterranean Sea, where the most recent study centered its efforts, an array of phytoplankton species bloom all year long.
Past analysis has mined satellite pictures of ocean color in the Mediterranean for frequent pigments found in phytoplankton. A mix of pigments can reveal dominant phytoplankton in the space, like specific species of diatoms that can be noticed due to their distinctive orange pigment, fucoxanthin. However, connecting the complex connections between satellite picture pixels, pigments, and phytoplankton varieties can make for a tricky analysis.
The newest study turns to artificial intelligence to parse via multidimensional data. The process copies the human brain’s ability to grasp new information and learn over time, giving the algorithm a chance to determine relationships in the knowledge that will not be readily apparent. The algorithms cluster similar nodes of data near each other, creating a two-dimensional diagram known as a “self-organized map.” The scientists trained two algorithms used in the research with 3 million pixels from satellite pictures and more than a thousand measurements taken by boat in the Mediterranean.
The outcomes show six kinds of phytoplankton and how they grow and diminish by season.