Space-based tech allows yield prediction and carbon monitoring

A new satellite-based technique to measure photosynthesis promises to make it easier for producers to predict yields and participate in carbon markets, while allowing developers to improve precision farming tools.

“Our technique purely uses satellite remote-sensing data, and is therefore completely observation-based as opposed to relying on complex, uncertain modelling methods,” said Kaiyu Guan from University of Illinois Urbana Champaign.

In an email interview, Guan, a bioenergy researcher at the university’s College of Agricultural, Consumer and Environmental Services, explained these methods must rely upon geographically imprecise climate data that is often out of date.

Working with his colleague Chongya Jiang, Guan measured something called gross primary production (GPP). This is what plants, primary producers, make. That is, they take molecules such as carbon dioxide and nitrogen from their environment and use sunlight to power the creation of leaves, stems and seeds, all of which contain carbon. GPP is used as a measure of carbon uptake by plants.

GPP is traditionally measured on the ground, either in the lab by putting plants into sealed chambers and measuring carbon dioxide levels, or outside by using sophisticated tower-mounted gas analysis instruments.

“Both measurements are expensive, complicated, and limited to a single location, and thus are impossible for civil applications,” Guan said.

To overcome these limitations, the researchers turned to satellite imagery. Such pictures don’t measure carbon directly but they capture sunlight conditions and high-resolution pictures in real time. With this, the researchers could figure out how much photosynthesis, and hence carbon capture, was happening.

“In other words, satellite measures quantities that can be used as proxy for GPP and thus can indirectly measure GPP,” Guan said.

Turning satellite signals into a valid GPP proxy was a challenge. The researchers turned to the latest research that described a better way to measure photosynthesis. They looked at both the satellite data and that of several ground-based observation networks, using novel analysis techniques and machine learning.

They named their technique SLOPE GPP, for SatelLite Only Photosynthesis Estimation Gross Primary Production. It offers field-scale spatial resolution, updatable daily.

Guan emphasized that SLOPE GPP is a big help in quantifying how much carbon is being sequestered by plants, but it’s not the endgame. For one thing, it measures how much carbon the plants take up, but not how much is released back into the atmosphere through plant respiration and from the soil. Models are still needed.

“The SLOPE GPP product can serve as good constraints to optimize ecosystem models for individual fields so that they can quantify carbon sequestration with higher accuracy that can satisfy the requirements for carbon markets,” he said.

More immediately, Guan pointed out that GPP is directly linked to how fast a plant is growing, or accumulating biomass.

“Therefore, if producers compare their historical yields with accumulated GPP, they will be able to predict their crop yields in real time.”

Crunching the vast amounts of data involved requires the use of the Blue Waters supercomputer, at the National Center for Supercomputing Applications (NCSA). Once the analysis is done, it’s posted for users.

“We routinely update the data, and we have a plan to allow farmers or normal users to get access to the data,” Guan said.

SLOPE GPP has drawn attention for its potential for policy makers and commercial applications such as precision farming. The university’s spinoff company, Aspiring Universe Corporation, is handling this aspect of the project.

About the author

explore

Stories from our other publications