Trees play one of the most important roles in the Earth’s ecosystem, providing a number of functions to assist the symbiosis of natural processes. These include (but are not limited to) providing homes for various birds and animals, creating a canopy for plants and shrubs growing on the forest floor, and acting as a negative emitter by storing CO2 from the atmosphere and releasing oxygen. Forests are therefore essential for a healthy planet. However, they have been placed under continual pressure as a result of rising deforestation and climate change which triggers forest fires, droughts and pest outbreaks. Traditionally, forest inventory management projects will operate manual field sample surveys of small forest patches across thousands of acres. SilviaTerra, on the other hand, uses Artificial Intelligence (AI) to provide aerial imagery in cloud-hosted databases, allowing the entire forest’s health to be monitored and recorded without physically travelling to remote areas.

Since starting operations in 2010, SilviaTerra’s goal has been to create an innovative service for conservationists, governments, and landowners by reducing manual fieldwork and improving data quality and precision with AI. Running out of Microsoft Azure, a cloud-based service for big data analytics which processes large datasets, SilviaTerra combines high-resolution satellite imagery with pre-existing field data, applying machine-learned classification models to create highly detailed forest maps with an unprecedented fine-grained resolution of 15 meters. SilvaTerra’s first initiative, Basemap, was released in 2019 and covered the entire continental United States, equating to 600 million acres of forest and 92 billion individual trees. In addition to forest density, Basemap provides information on tree species and diameter, annual carbon storage per acre, and wildlife habitat susceptibility metrics. Each of these functions can help forest managers strategize where to plant certain tree species and identify areas that can mature into old growth and create plans for efficiently controlled burns or forest fires.

SilviaTerra has produced world-class data that has helped determine practical forest mitigation efforts by targeting specific forested areas in United States national parks. For example, in New Mexico, Carson National Forest pinyon pine trees have always been highly susceptible to bark beetles. Extended droughts and the suppression of natural wildfires have caused dangerously high tree density, leading to a recurring bark beetle outbreak in the area. However, the AI-powered Basemap has successfully enabled foresters to pinpoint areas of overgrowth to reduce density and map the spread of bark beetles. This also assisted in the development of other practices which protect the remaining healthy trees and let the forest thrive. Beyond Basemap, SilviaTerra has also provided two precision forestry tools for stakeholders, Plot Hound and CruiseBoost. They use the AI-powered Basemap for managing, reporting, mapping, and analyzing timber cruises.

The disadvantage of SilviaTerra, like many AI-based geotechnologies, is that its coverage is limited to the continental United States, covering only 3.1 million km2 of the world’s 16.2 billion km2 of forested land. Areas susceptible to illegal deforestation, such as the Amazon rainforest, could benefit from a comprehensive AI map like SilverTerra’s Basemap. Yet, there is no mention of plans for extending the Basemap’s coverage globally. In the meantime, SilviaTerra provides machine learning models and incredibly detailed maps that are helping US-based foresters, conservationists, governments, and landowners discover more about how forests operate. Future developed solutions could be tested and applied to SilverTerra’s Basemap as more technology like this comes into play for the rest of the world. As an AI-based future continues to pick up steam, forest conservation will continue to benefit from satellite imagery that produces clear images of forests around the globe.

References

Microsoft Azure. (2020). SilviaTerra.  Retrieved from https://www.microsoft.com/en-us/ai/ai-for-earth-partners?activetab=pivot1:primaryr8

Microsoft. (2020). AI helps bring forests into focus: Microsoft In Culture. Retrieved May 29, 2020, from https://focusforests.microsoft.com/#carson

Microsoft. (2020). Cloud Computing Services: Microsoft Azure. Retrieved from https://azure.microsoft.com/en-ca/

Photo by Courtenay Crane