Back to Blog
![]() ![]() Machine learning (and particularly deep learning) in Earth system science is now more capable of reaching the high dimensionality, complexity and nonlinearity of real-life Earth systems and is expanding the learning ability from Big Data. Data assimilation provides the framework for incorporation of high-resolution observations into Earth system models but lacks the decision-making interface and learning ability needed for the digital twin. ![]() In this Review, we summarize the methodological and cyberinfrastructure advances in Big Data that have advanced the progress towards a digital Earth twin. Although computational advances are rapidly progressing, digital twins of Earth have not yet been produced. The concept of a digital twin of Earth envisages the convergence of Big Earth Data with physics-based models in an interactive computational framework that enables monitoring and prediction of environmental and social perturbations for use in sustainable governance. ![]()
0 Comments
Read More
Leave a Reply. |