Ground Truth– 2019
GEDI & & Sentinel-2 Prediction– 2019
An instrument developed to measure tree height can likewise distinguish corn from other crops.
Every second, lasers mounted on the International Space Station send 242 quick pulses of light down to Earth. These safe beams from NASAs Global Ecosystem Dynamics Investigation (GEDI) instrument bounce off Earths natural and human-made surfaces and are reflected back to the instrument. By determining the time it takes for the signals to come back, researchers can derive the height of the surface area listed below.
Researchers use these light detection and varying, or lidar, measurements to develop three-dimensional profiles of Earths surface area. GEDIs main mission is to determine tree heights and forest structure in order to approximate the amount of carbon stored in forests and mangroves. New research supported by NASA Harvest reveals these information also can be used to map where different kinds of crops are being grown.
When David Lobell, a farming ecologist at Stanford University, saw scientists using GEDI data to estimate tree heights, he wondered how he might use the data to study agriculture. Stefania Di Tommaso and Sherrie Wang, scientists on his team, came up with the idea of using the information to differentiate various types of crops growing on farms.
Wang reached out to the GEDI science group at the University of Maryland to see if they were utilizing the instrument for farming research. They reacted that they were not sure GEDI information could be utilized for such an application. “But they did not say it was difficult,” stated Lobell, who assists lead crop yield research studies for NASA Harvest.
Mapping where certain crops are grown is very important for estimating the overall production of the worlds major crops. But it has been hard to reliably map crop types from area due to the fact that many plants can look the same in optical images.
Lobell and his group started with corn (maize). When totally grown, typical corn stalks are about a meter taller than other crops, a distinction that is noticeable in GEDI profiles. Using this insight, the Stanford team combined the lidar profile data from GEDI with optical images from the European Space Agencys Sentinel-2 satellites. They were able to remotely map corn in 3 areas where there was trusted ground-based information to validate their observations: the state of Iowa in the U.S., the province of Jilin in China, and the region of Grand Est in France.
The images at the top of the page show the circulation of corn and other crops near Truchtersheim, France, as determined from the ground (leading image) and from the GEDI-Sentinel model (lower image). The images listed below reveal the exact same strategy used to the 3 study websites.
2019
The Stanford algorithm correctly prominent corn from other crops with a precision above 83 percent. The model using Sentinel-2 information alone had a total typical accuracy of 64 percent. “Two years back, I never ever would have believed that GEDI could be utilized in this way,” Lobell said.
In the future, the research study group aims to map corn production around the world, which might be utilized to understand the harvest prospects of corn each year. It might also assist farmers and help firms assess food security concerns and get a sense of possible changes in management that could improve production in major corn-producing areas.
NASA Earth Observatory images by Lauren Dauphin, utilizing data from DiTommaso et al. (2021) and Landsat information from the U.S. Geological Survey. Story by Emily Cassidy, NASA Earthdata.
Wang reached out to the GEDI science group at the University of Maryland to see if they were using the instrument for agricultural research. They responded that they were not sure GEDI data might be used for such an application. Utilizing this insight, the Stanford group combined the lidar profile data from GEDI with optical imagery from the European Space Agencys Sentinel-2 satellites. The design using Sentinel-2 data alone had a total typical precision of 64 percent.
New research study supported by NASA Harvest exposes these information also can be used to map where various types of crops are being grown.
