The European Space Agency (ESA) is working on a “digital twin” of Earth in the hopes of better understanding our planet’s past, present, and future. The project, first announced in September of last year, will deploy AI, as well as quantum computing, to build Earth’s digital doppelgänger in virtual space. And the scientists hope this Digital Twin Earth will help them forecast extreme, climate change-induced weather events.
Popular Mechanics reported on the digital Earth, which ESA scientists discussed during the agency’s 2020 Φ-week event. The scientists say their digital model will help humanity to “monitor the health of the planet,” as well as simulate the effects of human behavior on the environment.
Kevin Gill / Christian Colen
The scientists are going to evolve the digital twin over the next decade, constantly feeding real-world data into the model; data that will come from the EU’s Copernicus program, which captures atmospheric data, such as air quality changes. They’ll then use neural networks (computer algorithms) to identify patterns in Earth’s weather systems, and hopefully begin making accurate predictions.
“Machine learning and artificial intelligence could improve the realism and efficiency of the Digital Twin Earth—especially for extreme weather events and numerical forecast models,” European Center for Medium-Range Weather Forecasts (ECMWF) Director General, Florence Rabier, said at the event. Rabier and her colleagues also noted that the satellites collecting the data for the models are deploying AI programs.
Our 10-year Strategy highlights our aim to focus on developing next-generation models to produce high-resolution #DigitalTwins of Earth.https://t.co/ilDnmg52nX— ECMWF (@ECMWF) February 2, 2021
Yesterday, a paper by Peter Bauer was published in @NatureClimate to explore this in detail.https://t.co/V6YLcFHKDr pic.twitter.com/LhdW0zkfdc
Speaking of which, the machine learning necessary to mimic Earth’s weather systems will be so computationally intensive, the scientists will have to use quantum computers. Which, at their core, distinguish themselves from non-quantum computers thanks to their using quantum bits, or qubits of information. (Unlike binary bits of information, which are made up of either zeroes or ones, qubits are “superpositions” of both zeroes and ones.)
Along with predicting weather events, the scientists say they hope the model helps humanity address the “urgent challenges and targets” in the Green Deal; a set of policies aimed at making Europe climate neutral in 2050. Which would be great, because extreme weather events are already kind of the norm. And they are terrifying.