AI and machine learning can be well used to monitor real-time environmental changes, identify and protect species, and predict natural disasters.
• Real-time environmental monitoring
It is not always clear how a single stressor, such as salt runoff from roads, affects a natural ecosystem, let alone multiple stressors. In the past, environmental assessments are often manually collected over time, making it very difficult to monitor cause and effect.
To solve this problem, scientists are using machine learning to analyze data from environmental sensors to build and refine computer models of the ecosystem, to better understand the physical and biological processes. Over time, as more data is collected, machine learning will enable a better understanding of how stressors such as salt runoff from roads, invasive species, land-use changes, and climate change impact the ecosystem. This enables decision-makers to conduct real-time environmental monitoring and more targeted remediation, and provide insights to policymakers about the economic impact of their actions such as road treatment practices.
• Species identification and protection
Species are becoming extinct at a rapid pace, which will have devastating impacts on the ecosystem on which humans rely. What is more, what we don’t know is even more worrying – today we only know roughly 1% of all species on Earth. Fortunately, AI technology and machine learning can close this information gap cost-effectively and time-efficiently.
Traditionally, one of the activities in wildlife conservation is monitoring species with camera traps, which take hundreds of shots around the clock; once images are collected, they then need to be analyzed by filling in tables on the type of animals recorded. This rather tedious process can now be automated: we can teach machines to detect and classify flora and fauna from camera trap videos and images. Early work is proving that algorithms can sift through massive amounts of data streaming back from the monitoring systems, and humans and machines can begin to identify the species captured by these remotely deployed cameras and microphones. Automated animal identification performs at the same 96.6% accuracy level of human volunteers, saving approximately 8.2 years of human labeling effort on a 3.2- million-image dataset, according to a research institution.
• Natural disaster predictions
Predictions on major changes in our ecosystem are extremely valuable. To understand and predict where, how, and why environmental changes happen could help us mitigate one of the main threats to our planet. Machine learning can also help to create early warning systems that help us to respond to large-scale natural disasters – tsunamis, hurricanes, tornadoes, flooding, earthquakes, to name just a few.
The National Science Foundation is using machine learning to create a 3-D living model of the entire planet, called EarthCube. The digital representation will combine data sets provided by scientists across a whole slew of disciplines — measurements of the atmosphere and hydrosphere or the geochemistry of the oceans, for example — to mimic conditions on, above, and below the surface. Because of the vast amounts of data the cube will encompass, it will be able to model different conditions and predict how the planet’s systems will respond. And with that information, scientists will be able to suggest ways to avoid catastrophic events or simply plan for those that can’t be avoided, such as flooding or rough weather, before they happen.