The U.S. government is set to invest $80 million in supercomputing for weather and climate models, yet predicting specific weather beyond 10-15 days remains limited due to the chaotic nature of Earth’s atmosphere. This challenge relates to meteorologist Edward Lorenz’s “butterfly effect,” which illustrates how tiny changes in initial conditions can lead to vastly different outcomes in weather models. As errors in models propagate over time, predictions begin to diverge significantly, causing models to “lose memory” of initial conditions.
Research indicates that as temperatures rise, these errors propagate more rapidly, linked to faster growth of atmospheric storms known as eddies. Understanding the finite limits of predictability is crucial for improving climate models and preparing for future changes. The study, involving researchers from Stanford and the Geophysical Fluid Dynamics Laboratory, emphasizes the socioeconomic value of weather forecasting, estimated at $160 billion annually.
Source link
