Coupled GCMs (CGCMs) exhibit systematic errors in the climate of the tropical Pacific, including an equatorial cold tongue (ECT) that is too strong and too far west, and an ENSO with incorrect behavior and teleconnections. These biases limit climate forecast skill and confidence in future climate projections. The modeling community has identified deficiencies in ECT air-sea interactions, upwelling, and mixing as key contributors to emergent CGCM biases, citing a lack of detailed observations as a major factor. To address this, TPOS 2020 aims to enhance observations of upper-ocean and air-sea processes in the tropical Pacific, especially in the ECT region as part of the PUMP process study. TPOS 2020 has specifically requested hierarchical modeling to assess simulation capabilities, attribute biases to underlying processes, and refine observing strategies. A coupled approach is needed to assess how biases and air-sea feedbacks interact in models. Thus we propose using a hierarchy of GFDL CGCMs and ocean models with varying resolutions, complexity, and observational constraints, to identify links among emergent biases and processes in the ECT, and refine observing strategies to improve model simulations, reanalyses, and outlooks. Our project builds upon recent field programs, ocean model studies, diagnostic frameworks, and research that illuminate how ECT biases interact with climate variations and climate change in models. We intend to leverage the latest modeling, reanalysis, and forecast systems developed by GFDL to support the NMME and CMIP6. By focusing on ECT processes, air-sea coupling, and model resolution and parameterizations, the proposed studies will lend critical insights into strategies for observing the processes relevant to the ECT, and for improving NOAA’s models, products, and science.