Tree cover is declining in parts of the United States, forests have been burned by wildfires on the West Coast and drowned by rising sea levels in the East. From the ground it is difficult to assess the extent of the loss and the impact of disappearing trees on atmospheric carbon dioxide levels and climate change. NASA research scientist John Ranson is working to develop new technologies to study trees from above so future Earth-observing missions can more accurately assess the state of the forest.

“Trees do a huge service to the planet in terms of absorbing carbon dioxide, removing it from the atmosphere and putting it into wood,” said Ranson, “but trees are very sensitive to climate change. We’re trying to see changes in forest ecosystems. If you notice something early enough, it’s You can do something about it.”

By measuring the wavelengths of light that plants reflect from airplanes and satellites, they tell scientists how much photosynthesis takes place, and therefore how much carbon dioxide in the atmosphere is absorbed and stored by trees. The current standard for studying vegetation is called the NDVI or Normalized Differential Vegetation Index, which is the average of two broad parts of the infrared spectrum. NASA’s NDVI records date back 40 years, providing a low-resolution but accurate picture of forest conditions.

While NDVI is good at broadly estimating the amount and energy of vegetation, Ranson said, by splitting infrared and visible light into many other wavelengths, the technology called hyperspectral imaging can provide information about plant water content, chlorophyll, and even changes. in health

“Vegetation has these broad spectral features,” said Ian Adams, a technologist in the Earth Sciences Division at NASA’s Goddard Space Flight Center. “With hyperspectral imaging, you get a lot of different measurements at lower, closer frequencies. If we can get better spectral resolution, there’s a lot more information we can get.”

Greatly increasing the number of frequencies at which researchers can work, Ranson’s work aligns with the priorities of the National Academy of Earth Sciences’ most recent decade of research, which sets the industry’s long-term priorities. The study lists “surface relief and vegetation,” including forests, as a key research area requiring more advanced technology.

“Strategically for NASA, and for the remote sensing community more broadly, hyperspectral is an area we see as the future,” Adams said. Said.

To use hyperspectral imaging in future orbital missions, data analysis techniques are first tested closer to the ground.

Ranson’s team equipped a Skyfish drone (UAV) from partner company Virginia Tech with a visible infrared (VIS/IR) hyperspectral camera and lidar equipment. They flew the imaging equipment over the forest near Blacksburg, Virginia, in an area called Mountain Lake. By comparing the drone observations with actual CO2 levels recorded by sensors in a nearby tower at the National Environmental Observatory, Ranson’s team was able to refine their calculations of how much carbon the forest had removed from the atmosphere.

These comparisons allowed them to improve their technique for interpreting hyperspectral data. For example, plants that suffer from too much sunlight may secrete pigments to protect their chloroplasts, which can be detected by their sensors. If plants receive too much shade, leaves with a larger surface area can grow, causing the sensor to overestimate plant performance. Ranson wants to add shortwave infrared sensors to better distinguish reflections from leaves and other plant parts, eliminating another possible source of error.