Algae blooms, such as cyanobacteria blooms in freshwater lakes, are a growing problem in recent years, and blooms are projected to increase in both frequency and severity with rising global temperatures. With our partners at the Atlantic Coastal Action Program (ACAP) in Saint John, New Brunswick, we aim to improve cyanobacteria bloom detection and awareness.
Our final product aims to be a passive monitoring system of Atlantic Canadian freshwater systems using satellite imagery. A deep learning algorithm will issue an alert if a likely bloom is detected. This program would be a cost-effective addition to current monitoring techniques, which rely on direct sampling and can be expensive and time consuming.
Why satellite data:
Satellites are an incredibly useful tool, with a variety of programs capturing imagery of the Earth’s surface for decades. Monitoring using satellite data is independent of local infrastructure or social/political boundaries that can make direct monitoring difficult. Satellite imagery covers several 100s of km2 and enables large-scale research.
In this project we are using satellites that employ passive monitoring techniques. Passive satellites measure energy emitted from the planet’s surface as reflected light or thermal infrared signals. A set of special cameras onboard the satellite capture the signal strength, with each camera designed to only capture light at a particular wavelength. Comparing the strength of different wavelengths to each other allows the monitoring of ecosystem health, urban expansion, and tracking large-impact natural events like hurricanes, forest fires, and algae blooms.
In recent years there has been leaps of progress in the commercial development of nanosatellite networks. Nanosatellites are much smaller than previous satellite generations, generally weighing under 10 kg. Their small size allows for multiple nanosatellites to be put into orbit during the same rocket launch. Hundreds of nanosatellites now orbit the Earth, allowing for high-resolution images of any chosen area to be captured on an almost daily frequency.