NASA scientists have successfully used lasers installed on the International Space Station to scan fields of corn in the USA, China and France to map. Going forward, the research team intends to create a map of corn production worldwide that could be used to determine the harvest prospects of this agricultural crop important for maintaining food security, reports the website FutureFarming.
It could also help farmers and scientists to assess food security issues and select an agricultural technique that would help improve production in key regions - the corn producers.
The lasers on the International Space Station are part of NASA's Global Ecosystem Dynamics (GEDI) research. Every second they send 242 fast pulses of light to Earth, which bounce off our planet's surface and can be used to create three-dimensional profiles of the Earth's surface. The main task of GEDI is to measure the height of the trees and the structure of the forest to estimate the amount of carbon accumulated in forests and mangroves. However, a new study supported by NASA Harvest shows that this data can also be used to create maps on which different types of crops are grown.
Mapping the acreage of specific crops is important to assess their overall production in the world. However, according to NASA, correctly mapping the types of crops from space has been difficult because many plants could look the same in optical images.
David Lobell, an agroecologist at Stanford University, is the director of NASA's Harvest project. He and his team began using GEDI data to map corn. When fully grown, average corn stalks are about a meter taller than other crops, and this difference is noticeable in the GEDI profiles. Based on this fact, lidar profile data from GEDI was combined with optical images from the European Space Agency's Sentinel-2 satellites. They were able to remotely map the corn fields in three regions where reliable land data existed to support their observations: Iowa to the USA, Jilin in China and the Grand Est region in France.
The Stanford algorithm has Corn correctly distinguished from other cultures with an accuracy of more than 83%. The model using only Sentinel-2 data had an average overall accuracy of 64%.