Understanding and Improving GCM Simulations of MJO Initiation over the Tropical Indian Ocean using DYNAMO Field Observations

Despite the numerous improvements made to General Circulation Models (GCMs) in recent years, current models continue to be deficient in simulating and predicting the Madden-Julian Oscillation (MJO), a leading tropical intraseasonal mode of variability that interacts with and influences a wide range of weather and climate phenomena. Particularly problematic is the simulation of the initial phase of the MJO over the tropical Indian Ocean ± a deficiency that limits both forecast skill and the ability to simulate a wide range of scale interactions linked to the MJO. The Dynamics of MJO (DYNAMO) field campaign, which provided in-situ observations over the central tropical Indian Ocean during the period October 1, 2011-March 31, 2012, will provide a unique opportunity to confront GCMs with in-situ observations, with the ultimate goal of improving GCM simulations and predictions of the MJO.

It is proposed to use in-situ observations collected during the DYNAMO field campaign and the NASA Goddard Earth Observing System Model, Version 5 (GEOS-5) model system, to better understand the key physical processes associated with the MJO initiation over the tropical Indian Ocean and identify the main model deficiencies in simulating the MJO initiation. First, we will use the latest GEOS-5 data assimilation system to produce the global high-resolution DYNAMO Reanalysis so as to place the DYNAMO in-situ observations in a global context. The DYNAMO Reanalysis will be analyzed to investigate the key physical processes associated with the MJO initiation over the tropical Indian Ocean, and assess the GEOS-5 Atmospheric GCM (AGCM) deficiencies through a detailed examination of the analysis increments that should largely reflect model bias. Second, we will perform a series of AGCM and coupled replay experiments, to investigate the relative roles of atmospheric processes, and the impact of air-sea interaction, for the MJO initiation over the tropical Indian Ocean. Third, a series of coupled MJO reforecasts will be performed to assess the contribution of the DYNAMO in-situ observations to an enhanced MJO forecast skill through improvements in atmospheric and oceanic initial conditions. The root cause(s) of GEOS-5 limited forecast skill will be assessed through in-depth case-study analyses of when and where in the early stages of the MJO life cycle the model begins to lose skill. Lastly, we will perform a series of GCM sensitivity experiments to investigate the model convective sensitivity to environmental moisture as well as to address model issues revealed in the above data diagnosis.

The above work with the GEOS-5 model will be coordinated with on-going modeling efforts at NOAA Geophysical Fluid Dynamics Laboratory (GFDL) that address the simulation of the MJO in the GFDL Atmospheric Model version 3 (AM3) and Coupled Model version 3 (CM3) models. It is anticipated that frequent interactions between Global Modeling and Assimilation Office (GMAO) and GFDL on lessons learned, as well as some joint model experimentation, will be key to achieving fundamental improvements that extend beyond any model-specific bias, and point to fundamental improvements in the representation of the MJO in both the GEOS-5 and GFDL GCMs.

The proposed work targets the focus area “Understanding and Improving Prediction of Tropical Convection using Results from the DYNAMO Field Campaign” solicited by FY 2013 NOAA Earth System Science (ESS) Program, and directly addresses the motivation and goal of the DYNAMO program. It will contribute to NOAA’s long-term goal of climate adaptation and mitigation through “Improved scientific understanding of the changing climae system and its impacts”.