This project aims at developing fundamental understanding and synthesis of the processes that drive Pacific decadal variability (PDV) in Earth System Models (ESMs) under different external climate forcing scenarios using simulations from the Coupled Model Intercomparison Project (CMIP) and E3SM v1. We will diagnose, and compare across ESMs, the role of coupled ocean-atmosphere processes and feedbacks, determine how they contribute to predictable PDV dynamics and how they are affected by unpredictable stochastic forcing, and evaluate mean state dependencies including those associated with anthropogenic external forcing.

Approach: Testing an observationally-based hypothesis that outlines the dominant processes of Pacific Decadal Variability (PDV), we will examine the mechanisms and teleconnections that energize the modes of PDV in ESMs under historical and projected future conditions. The data and methods for this project include (a) data archives from the latest available CMIP, (b) data archives from existing large ensembles from CESM, GFDL ESM2M, IPSL CM5A, and CanESM (c) a hierarchy of reduced complexity stochastic models, (d) a set of simulations from a new hybrid coupled climate model based on the CESM and E3SM, and finally (e) a 15-member large ensemble of the DOE E3SM v1 model, newly generated in this project, where each member undergoes the same historical/projected 1850-2100 external forcing. Our approach is to first construct an integrated stochastic model that incorporates the mechanisms and teleconnections of the observationally-based hypothesis, using it to quantify and synthesize the dominant pathways energizing PDV in each ESM. Next, individual processes of PDV in the tropics (e.g. stochastic forcing, ENSO precursors and feedback dynamics) and extra-tropics (e.g. tropical/extra-tropical teleconnections, stochastic forcing, and memory dynamics) will be further examined using the hybrid simulations and advanced budget analyses. The influence of anthropogenic external forcing (e.g. changes in mean state) will also be investigated using similar approaches applied to the existing large-ensemble simulations from CMIP5-era models and the new CMIP6-era E3SM large ensemble. Substantial internal climate variability has been shown to impact regional estimation of forced signals within ESMs, requiring large ensembles for robust identification of the links between climate change and PDV. Our approach moves beyond simple statistical comparisons and analyses of the patterns of the PDV modes to deliver mechanistic understanding and synthesis of the dynamics and processes that energize their variance.

Science Team: The PIs bring complementary expertise in the theory and stochastic modeling of PDV (Di Lorenzo, Newman), ENSO dynamics and sensitivity to climate change (Stevenson, Newman), hybrid coupled ocean-atmosphere modeling (Newman) and coupled climate modeling (Stevenson and Di Lorenzo). The PIs will also collaborate with Dr. Luke Van Roekel, an E3SM developer at Los Alamos National Laboratory, to conduct the E3SM large-ensemble and E3SM hybrid coupled simulations.

Broader Implications: PDV controls both large-amplitude transitions in marine ecosystems and the statistics of climate and weather extremes in the Pacific basin. More recently, PDV has been implicated as a driver of the warming hiatus, yet the mechanisms controlling its phase and predictability remain unclear because the wide-ranging expressions of PDV within ESMs have not been diagnosed from a mechanistic point of view. This proposal will provide the basis for a dynamical interpretation of the differences in the current suite of coupled climate models and associated representation errors. The improved understanding and modeling of the dynamics of PDV will provide a mechanistic-based framework for predicting societally relevant indicators linked to droughts, heatwaves, and ecosystem services. Furthermore, the tools and diagnostics developed in this project can be generalized for other regions (e.g. tropical Atlantic and Indian oceans) and will be made available through a project website. In addition, the proposed work will result in a new E3SM 20th/21st-century ensemble, which will be provided as a resource to the community, creating new opportunities for widespread use of E3SM results.