The Madden–Julian Oscillation Simulated in the NCEP Climate Forecast System Model: The Importance of Stratiform Heating

Abstract

Abstract This study investigates the capability for simulating the Madden–Julian oscillation (MJO) in a series of atmosphere–ocean coupled and uncoupled simulations using NCEP operational general circulation models. The effect of air–sea coupling on the MJO is examined by comparing long-term simulations from the coupled Climate Forecast System (CFS T62) and the atmospheric Global Forecast System (GFS T62) models. Another coupled simulation with a higher horizontal resolution model (CFS T126) is performed to investigate the impact of model horizontal resolution. Furthermore, to examine the impact on a deep convection scheme, an additional coupled T126 run (CFS T126RAS) is conducted with the relaxed Arakawa–Schubert (RAS) scheme. The most important factors for the proper simulation of the MJO are investigated from these runs. The empirical orthogonal function, lagged regression, and spectral analyses indicated that the interactive air–sea coupling greatly improved the coherence between convection, circulation, and other surface fields on the intraseasonal time scale. A higher horizontal resolution run (CFS T126) did not show significant improvements in the intensity and structure. However, GFS T62, CFS T62, and CFS T126 all yielded the 30–60-day variances that were not statistically distinguishable from the background red noise spectrum. Their eastward propagation was stalled over the Maritime Continent and far western Pacific. In contrast to the model simulations using the simplified Arakawa–Schubert (SAS) cumulus scheme, CFS T126RAS produced statistically significant spectral peaks in the MJO frequency band, and greatly improved the strength of the MJO convection and circulation. Most importantly, the ability of MJO convection signal to penetrate into the Maritime Continent and western Pacific was demonstrated. In this simulation, an early-stage shallow heating and moistening preconditioned the atmosphere for subsequent intense MJO convection and a top-heavy vertical heating profile was formed by stratiform heating in the upper and middle troposphere, working to increase temperature anomalies and hence eddy available potential energy that sustains the MJO. The stratiform heating arose from convective detrainment of moisture to the environment and stratiform anvil clouds. Therefore, the following factors were analyzed to be most important for the proper simulation of the MJO rather than the correct simulations of basic-state precipitation, sea surface temperature, intertropical convergence zone, vertical zonal wind shear, and lower-level zonal winds: 1) an elevated vertical heating structure (by stratiform heating), 2) a moisture–stratiform instability process (a positive feedback process between moisture and convective–stratiform clouds), and 3) the low-level moisture convergence to the east of MJO convection (through the appropriate moisture and convective–stratiform cloud processes–circulation interactions). The improved MJO simulation did improve the global circulation response to the tropical heating and may extend the predictability of weather and climate over Asia and North America.