MM5 Modeling of the Madden–Julian Oscillation in the Indian and West Pacific Oceans: Model Description and Control Run Results

Abstract

A new methodology to study the Madden?Julian oscillation (MJO) is introduced. While previous MJO studies typically have involved highly simplified mathematical models or general circulation models, this new approach seeks to reproduce the MJO by using a regional model with prescribed boundary conditions. This paper reports initial control run results for this methodology using the fifth-generation Pennsylvania State University (PSU)?National Center for Atmospheric Research (NCAR) Mesoscale Model (MM5) for a domain extending from the western Indian Ocean to the date line. The control run boundaries are forced using the NCEP?NCAR reanalysis (NRA) dataset for a 24-month time period. The climatology for the 24-month period is examined to establish the robustness of MM5 for this region. Results indicate good agreement in the mean winds between the model and the forcing dataset. The primary differences are an easterly bias at 850 hPa and altered flow patterns in the Indian monsoon region. Mean outgoing longwave radiation (OLR) results are good for the model interior with larger discrepancies near the western and eastern boundaries. These discrepancies lead to a reversal of the OLR gradient along the equator. Thirty- to seventy-day bandpassed data are examined to determine how MM5 reproduces the MJO. The modeled and comparison 30?70-day zonal wind and OLR data have similar MJO periodicities, exhibit eastward propagation, and possess the observed seasonal character and vertical structure of the MJO. The ?Matthews EOF technique? reveals good similarity between the model and observed OLR. Analysis of vertical profiles of 30?70-day zonal wind reveals that lower-tropospheric winds blow in the opposite direction of upper-level winds for both the model and NRA. Vertical profiles of 30?70-day moist static energy exhibit a peak near the top of the boundary layer. Differences between the model-simulated and observed MJO events have a tendency for the OLR to be relatively noisy and for peak OLR intensity to occur in the west Indian Ocean in the model, as opposed to the eastern Indian Ocean in observations. This paper establishes the groundwork for a successive paper wherein the boundary forcings will be modified to examine how this alters the modeled MJO.