Official websites use .gov
A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS
A lock ( ) or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.

Home » Understanding Tropical Pacific Biases in Climate Simulations and Initialized Predictions
Climate Variability & Predictability (CVP) logo

Understanding Tropical Pacific Biases in Climate Simulations and Initialized Predictions

We propose a collaborative study between GFDL and NCEP, to advance understanding, simulation, and forecasting of tropical Pacific climate and its variability. Motivated by the central role that the tropical Pacific plays in climate variability worldwide — in particular via the El Niño / Southern Oscillation (ENSO) — there is an urgent need to advance mechanistic understanding of the tropical Pacific climatology and its impacts on climate variability, and to improve the coupled general circulation models (CGCMs) upon which society relies for seasonal-to-interannual (SI) forecasts and decadal-to-centennial predictions and projections. A unique strength of this proposal is close coordination between two of NOAA’s premier institutions for simulation, assimilation, and prediction of tropical Pacific climate and ENSO. In particular, a critical aspect of the proposed work is the development of common metrics, and a coordinated design and analysis of focused simulation and forecast experiments leveraging next generation models. This coordination will facilitate assessment of the robustness of the model results and underlying mechanisms, to accelerate improvements in NOAA’s SI simulation and prediction capabilities. Our goals are to (1) diagnose the spatiotemporal structure of tropical Pacific climatological biases in GFDL’s and NCEP’s coupled simulations, reanalysis systems, and forecasts; (2) identify similarities and differences among GFDL’s and NCEP’s model biases, and understand how differences can be linked to model parameterizations, assimilation methods, and observational inputs; (3) understand the processes which seed and amplify tropical Pacific biases; (4) assess how these biases affect the simulation and prediction of climate fluctuations; and (5) develop methods to mitigate these biases and their impacts on forecast skill. We will use our findings to evaluate existing hypotheses for the emergence of tropical Pacific biases, and to assess the applicability of our results to the broader set of community models and forecasts, including those available from the CMIP5 and NMME projects.

Relevance: The proposed work is highly relevant to the NOAA CPO. We seek to improve scientific understanding and prediction of the climate system, by evaluating and advancing methodologies used for simulations and forecasts. We directly address several objectives of NOAA’s Next Generation Strategic Plan (NGSP), including (1) Improved scientific understanding of the changing climate system and its impacts, by elucidating the causes and effects of simulation biases, and advancing climate modeling, predictions, and projections; (2) An integrated environmental modeling system, by advancing fundamental climate research and transitioning it toward NOAA’s production of seasonal forecasts, by coordinating within NOAA to enhance the accuracy of global models and predictions, and by evaluating and optimizing NOAA’s investments in observation and monitoring through the use of models; and (4) A climate-literate public that understands its vulnerabilities to climate, by elucidating the strengths and limitations of climate information affected by simulation biases. Relevance to the Competition: This proposal directly addresses the ESS CVP solicitation. The analysis and multimodel experimentation will focus on understanding tropical Pacific biases in two of NOAA’s leading coupled GCMs, which are widely used for climate simulation, reanalysis, and predictions. The work will leverage nearly all of the methods suggested in the proposal call, including advanced physical metrics, reduced-model experiments, and short-term forecasts, to diagnose the sources and amplifiers of model biases.

Scroll to Top