Search for: All records

Abstract Global climate models parameterize a range of atmospheric‐oceanic processes, including gravity waves (GWs), clouds, moist convection, and turbulence, that cannot be sufficiently resolved. These subgrid‐scale closures for unresolved processes are a substantial source of model uncertainty. Here, we present a new approach to developing machine learning (ML) parameterizations of small‐scale climate processes by fine‐tuning a pre‐trained AI foundation model (FM). FMs are largely unexplored in climate research. A pre‐trained encoder‐decoder from a 2.3 billion parameter FM (NASA and IBM Research's Prithvi WxC)—which contains a latent probabilistic representation of atmospheric evolution—is fine‐tuned (or reused) to create a deep learning parameterization for atmospheric gravity waves (GWs); a process unseen during pre‐training. The parameterization captures GW effects for a coarse‐resolution climate model by learning the fluxes from an atmospheric reanalysis with 10 times finer resolution. A comparison of monthly averages and instantaneous evolution with a machine learning model baseline (an Attention U‐Net) reveals superior predictive performance of the FM parameterization throughout the atmosphere, even in regions excluded during pre‐training. This performance boost is quantified using the Hellinger distance, which is 0.11 for the baseline and 0.06 for the fine‐tuned model. Our findings emphasize the versatility and reusability of FMs, which could be used to accomplish a range of atmosphere‐ and climate‐related applications, leading the way for the creation of observations‐driven and physically accurate parameterizations for more earth system processes. 

Semi-crystalline plastics undergo necking followed by stable drawing under tensile forces. In contrast, a rubber extends many times its original length uniformly under tension. Previously we have shown experimentally that the behavior of rubber-plastic composites in tension is intermediate between that of the rubber. Here we conduct finite element simulations of plastic-rubber-plastic trilayers laminates under tension. Using relatively simple constitutive equations for the rubber and the plastic, we examine how the composite mechanics changes as the ratio of rubber to plastic thickness is varied. We show that at small rubber thickness, the composites show necking, whereas beyond a certain rubber thickness, necking is completely eliminated. 

We examine the stretching behavior of rubber–plastic composites composed of a layer of styrene–ethylene/propylene–styrene (SEPS) rubber, bonded to a layer of linear low density polyethylene (LLDPE) plastic. Dog-bone shaped samples of rubber, plastic, and rubber–plastic bilayers with rubber : plastic thickness ratio in the range of 1.2–9 were subjected to uniaxial tension tests. The degree of inhomogeneity of deformation was quantified by digital image correlation analysis of video recordings of these tests. In tension, the SEPS layer showed homogeneous deformation, whereas the LLDPE layer showed necking followed by stable drawing owing to its elastoplastic deformation behavior and post-yield strain hardening. Bilayer laminates showed behavior intermediate between the plastic and the rubber, with the degree of necking and drawing reducing as the rubber : plastic ratio increased. A simple model was developed in which the force in the bilayer was taken as the sum of forces in the plastic and the rubber layers measured independently. By applying a mechanical energy balance to this model, the changes in bilayer necking behavior with rubber thickness could be predicted qualitatively.