Search for: All records

Abstract We develop and demonstrate a new interpretable deep learning model specifically designed for image analysis in Earth system science applications. The neural network is designed to be inherently interpretable, rather than explained via post hoc methods. This is achieved by training the network to identify parts of training images that act as prototypes for correctly classifying unseen images. The new network architecture extends the interpretable prototype architecture of a previous study in computer science to incorporate absolute location. This is useful for Earth system science where images are typically the result of physics-based processes, and the information is often geolocated. Although the network is constrained to only learn via similarities to a small number of learned prototypes, it can be trained to exhibit only a minimal reduction in accuracy relative to noninterpretable architectures. We apply the new model to two Earth science use cases: a synthetic dataset that loosely represents atmospheric high and low pressure systems, and atmospheric reanalysis fields to identify the state of tropical convective activity associated with the Madden–Julian oscillation. In both cases, we demonstrate that considering absolute location greatly improves testing accuracies when compared with a location-agnostic method. Furthermore, the network architecture identifies specific historical dates that capture multivariate, prototypical behavior of tropical climate variability. Significance StatementMachine learning models are incredibly powerful predictors but are often opaque “black boxes.” The how-and-why the model makes its predictions is inscrutable—the model is not interpretable. We introduce a new machine learning model specifically designed for image analysis in Earth system science applications. The model is designed to be inherently interpretable and extends previous work in computer science to incorporate location information. This is important because images in Earth system science are typically the result of physics-based processes, and the information is often map based. We demonstrate its use for two Earth science use cases and show that the interpretable network exhibits only a small reduction in accuracy relative to black-box models. 

Abstract Few studies have utilized machine learning techniques to predict or understand the Madden‐Julian oscillation (MJO), a key source of subseasonal variability and predictability. Here, we present a simple framework for real‐time MJO prediction using shallow artificial neural networks (ANNs). We construct two ANN architectures, one deterministic and one probabilistic, that predict a real‐time MJO index using maps of tropical variables. These ANNs make skillful MJO predictions out to ∼18 days in October‐March and ∼11 days in April‐September, outperforming conventional linear models and efficiently capturing aspects of MJO predictability found in more complex, dynamical models. The flexibility and explainability of simple ANN frameworks are highlighted through varying model input and applying ANN explainability techniques that reveal sources and regions important for ANN prediction skill. The accessibility, performance, and efficiency of this simple machine learning framework is more broadly applicable to predict and understand other Earth system phenomena. 

Abstract The Propagation of Intraseasonal Tropical Oscillations (PISTON) experiment conducted a field campaign inAugust-October 2018. The R/V Thomas G. Thompson made two cruises in thewestern North Pacific region north of Palau and east of the Philippines. Using select field observations and global observational and reanalysis data sets, this study describes the large-scale state and evolution of the atmosphere and ocean during these cruises. Intraseasonal variability was weak during the field program, except for a period of suppressed convection in October. Tropical cyclone activity, on the other hand, was strong. Variability at the ship location was characterized by periods of low-level easterly atmospheric flow with embedded westward propagating synoptic-scale atmospheric disturbances, punctuated by periods of strong low-level westerly winds that were both connected to the Asian monsoon westerlies and associated with tropical cyclones. In the most dramatic case, westerlies persisted for days during and after tropical cyclone Jebi had passed to the north of the ship. In these periods, the sea surface temperature was reduced by a couple of degrees by both wind mixing and net surface heat fluxes that were strongly (~200 Wm −2 ) out of the ocean, due to both large latent heat flux and cloud shading associated with widespread deep convection. Underway conductivity-temperature transects showed dramatic cooling and deepening of the ocean mixed layer and erosion of the barrier layer after the passage of Typhoon Mangkhut due to entrainment of cooler water from below. Strong zonal currents observed over at least the upper 400 meters were likely related to the generation and propagation of near-inertial currents.