AWS Public Sector Blog

Managing the world’s natural resources with earth observation

A screenshot of the terraPulse dashboard featuring satellite imagery of land cover.

The global land market comprises a multi-trillion dollar economy of buyers, sellers, and brokers of real estate, natural resources, and ecosystem services—a set of values becoming known collectively as “natural capital.” Investors, lenders, insurers, and the various other operators in this market require information on the value of land for forestry, carbon sequestration, and other uses. Carbon markets, trading systems for buying and selling units of greenhouse gas credits to meet global climate goals, are creating a new opportunity for land owners to derive additional value from their land through conservation.

With increasing pressure from climate change, loss of biodiversity, and demand for natural resources from already stressed ecosystems, it has become essential to understand and address environmental changes by making sustainable land use decisions with the latest and most accurate data.

As part of the Amazon Sustainability Data Initiative (ASDI), Amazon Web Services (AWS) invited Joe Sexton, chief scientist and co-founder of terraPulse, to share how AWS technologies and open data are supporting terraPulse’s efforts to provide accurate and up-to-date information on the world’s changing ecosystems.

Making sustainability decisions and seeing the big picture requires data

Individual land owners make decisions on how to use land, as well as governments, conservation organizations, and private corporations. These decision-makers haven’t always had access to up-to-date or complete data on the resources they control, such as forests or wildlife habitat. With incomplete data, blind spots emerge and we can’t clearly see the effects of landscape changes such as from over-harvesting of timber. But accurately mapping and updating changing landscapes is a costly endeavor. Effective decision-making, especially in governmental sustainability policies, requires simple and affordable access to spatiotemporal data on changing ecosystems. Being able to visualize specifically the change of an environment over time, such as changes in surface water over the last ten years, allows for more flexibility and a quicker response to environmental changes—and possible consequences such as droughts and food insecurity.

Understanding the pulse of our planet with satellite data and AI

Established in 2014, terraPulse is a technology company devoted to creating timely, accurate information about the changing world to address the need for more accurate land use data. Using algorithms vetted through rigorous peer review and published in major scientific journals, terraPulse mines long-term, global satellite imagery with artificial intelligence (AI) to map changes in ecosystems over time. TerraPulse’s process delivers globally consistent information on current land conditions, as well as historical baselines and robust detection of changes.

TerraPulse originated in the Global Land Cover Facility (GLCF), a NASA-sponsored laboratory at the University of Maryland. At GLCF, terraPulse’s co-founders created the world’s first global maps of tree and water cover from the Landsat satellites, as well as Landsat-based time-series of impervious surface maps and MODIS-based maps of phenology and wildlife habitat. TerraPulse spun out of GLCF and is now a private company, with roots in academia and branches spreading across the geospatial information economy.

The terraPulse platform, built on AWS, is a solution for scientists, businesses, and governments to monitor and analyze Earth’s surface, and it leverages open satellite imagery hosted on the Registry of Open Data on AWS. The terraPulse platform is composed of three main components. The first is the data “cube”—a digital map of ecological variables like trees, surface water, impervious cover, biomass, and wildlife habitat. The data is hosted on Amazon Simple Storage Service (Amazon S3). The second component, the terraView dashboard, uses Amazon Elastic Compute Cloud (Amazon EC2) to display the current value of each map layer and retrieve the historical values for data points. The third component, the terraServe API, uses Amazon API Gateway with AWS Lambda to provide automated, big-data analytics and insights to data scientists. TerraPulse’s process offers consistent and current data on past and present land conditions worldwide, which is key information in accelerating our global response to environmental change.

The terraPulse platform is used worldwide by organizations such as the World Bank, the US Forest Service and Bureau of Land Management, the International Union for the Conservation of Nature (IUCN), World Wildlife Fund (WWF), and NASA.

Mapping for carbon markets

TerraPulse’s maps of current and historical forest cover are helping countries participate in emerging carbon markets. TerraPulse provided maps of historical forest cover and change to the Philippines and the Dominican Republic to serve as baselines of forest activity, information necessary for registering the countries’ carbon in regulated markets. And now, a new project is empowering Belize, Costa Rica, and the Dominican Republic to map each nation’s forest activity as well. TerraPulse is applying AI to nearly forty years of satellite imagery to map and monitor the countries’ forest cover, loss, and gain over time. The project fuses data from local sources with global satellite data available on AWS, and the results are enabling carbon modeling and the development of greenhouse gas inventories for the United Nations’ Land Use, Land-Use Change and Forestry (LULUCF) initiative.

The new project is supported by the Forest Carbon Partnership Facility, which empowers countries’ efforts to reduce emissions from deforestation and foster sustainable management of forests. “The challenge is that repeatable (short-cycle), large-scale (jurisdictional) monitoring using in situ measurements is not easily scalable. Countries participating in emissions reductions need a robust, transparent, and replicable approach to provide reliable estimates of land use and land cover change for estimating forest reference emission levels and reporting periodically about emissions and emission reductions,” says Dr. Julian Gonzalo, senior carbon finance specialist at World Bank Group. “Companies like terraPulse help bridge that information gap by using remotely sensed data and machine learning to provide reliable insights.”

In order to adapt, the land-use economy must accelerate its current pace of response to environmental change. Land markets—including brokerage, investment, insurance, operation, and governance functions—require current, accurate, and consistent information to value assets and to match land use to current and expected future conditions. The necessary information must be consistent over large areas and over the long term, it must be regularly updated, and it must accurately resolve the state and changes of the ecosystem at the scale of management action. TerraPulse brings information liquidity to natural resource management, facilitating participation in green markets and ultimately accelerating the resilience of the land-use economy to global change.

The datasets used by this project can be found in the Amazon Sustainability Data Initiative (ASDI) data catalog. ASDI seeks to accelerate sustainability research and innovation by minimizing the cost and time required to acquire and analyze large sustainability datasets. Joe Sexton will be presenting at the upcoming Amazon re:MARS event, from June 21-24, 2022. If you are attending re:MARS and want to attend the session, look for Taking the planet’s pulse: Using AI to empower adaptation to global change (Session ID MLR219) in the session catalog and favorite the session to add it to your agenda.

Read more stories about open data and sustainability on AWS:

Subscribe to the AWS Public Sector Blog newsletter to get the latest in AWS tools, solutions, and innovations from the public sector delivered to your inbox, or contact us.

Please take a few minutes to share insights regarding your experience with the AWS Public Sector Blog in this survey, and we’ll use feedback from the survey to create more content aligned with the preferences of our readers.