finding the tallest trees from space

project goal

The tallest known tree in the world is 380 feet tall (115 m). It's a California coastal redwood tree and is located in Redwood National Park. The tree was found and measured in in 2006, and named Hyperion after the greek Titan of heavenly light. In an attempt to avoid damage from excess tourists or vandalism, the location of the tree is kept a secret among a select few arborists.

In 2019, a new sensor called GEDI (Global Ecosystem Dynamics Instrument) was mounted aboard the International Space Station. Its purpose is to assess forest structure and health from space. I looked to see whether I could use the metrics on tree canopy height from this sensor to find the location of this trees.

methods

GEDI is a Laser Imaging, Detection and Ranging system (LIDAR). As with most ranging systems, it works by pulsing a laser and measuring the time it takes for that light to be reflected. For objects that are further away, the difference in time between the pulse and its return will be greater.

In addition to a single distance metric, LIDAR can also capture density profile characteristics for complex 3D objects. In particular, if a LIDAR beam illuminates a tree canopy, some percentage of the reflected return will come from the tree top, while another portion will come from the ground beneath the canopy. By analyzing the strength of this signal response, we can identify the ground height, the tree canopy density and structure, and the tree canopy height.

Classically, LIDAR systems for forest monitoring have been mounted in an airplane that flies across a region of interest. These scans can be incredibly detailed, but come at a high cost. The GEDI instrument extends the aerial survey concept, but mounts the LIDAR on the International Space Station rather than an aircraft. This allows for continuous measurement of a ground track beneath the orbiting satellite.

To amplify the coverage, GEDI splits the single beam into eight separate components, and takes a measurement from each of these beams. Each of these beams is separted by 600 m. GEDI makes 242 measurements per second, and each point of evaluation has a radius of 25 m. These points are separated by 60 m along the path of the orbit.

The first batch of data was released in January of 2020. Given a couple peculariarities in the way the data is structured, it can be challenging to work with. I have fully documented the process in a Google Colab notebook.

results

The presently available GEDI data contains only a portion of Redwood National Park. From other sources providing information about Hyperion's location, it seems that there is no data coverage from GEDI in that region.

In areas where there is coverage, it is impressive to see the detailed resolution. It seems like we can identify the Orion Grove in one of the hotspots.

Though I have found no documentation on the topic, I'm curious whether trees taller than 100 meters are not picked up by the sensor. In the 1900 measurements meeting the quality thresholds, no tree is taller than 100 meters, and one tree gets suspiciously close at 99.85 m.

Map of GEDI data in Redwood National Park where the tallest tree is located. Click for full screen view

conclusion

I continue to be blown away by the remote sensing data that the public can access. Measuring the height of trees worldwide with centimeter precision from space is remarkable, and it amazes me that an average person can access and analyze the data.

I am disappointed that the region of interest needed to find Hyperion is not yet covered by the sensor. That said, the mission continues, and new measurements continue to be added. I'll probably return to this analysis later.

I've also become a bit obsessed with this world of tall trees and peculiar characters who find them. I'm excited to visit Redwood National Forest to see these trees for myself.

what will i do next

• Re-run this analysis as new data from GEDI is released.
• Continue research into whether reported canopy height is limited at 100m.
• Visit Redwood National Park!