Stochastic Forcings Associated with MJO Initiation Over the Indian Ocean

Numerical forecast models have trouble simulating and forecasting the MJO, in terms of both discrete events and long term statistics, for many reasons. The Dynamics of the Madden-Julian Oscillation (DYNAMO) program, which focuses the community's attention on the birth of MJO events over the Indian Ocean, is showing that one reason may be our difficulty identifying when and where individual MJO events exist. Another reason is model deficiencies such as incorrectly tuned parameterizations or inadequate subgrid parameterizations.

One method of incorporating the effects of subgrid events in global models is via stochastic forcing. Previous research has shown that the MJO can be described as a linear system consisting of a determinstic part plus stochastic forcing. Linear Inverse Modeling (LIM) can provide an estimate of both a linear description of the resolved dynamics as well as time series representing the stochastic forcing required to maintain, and possibly initiate, MJO events. Such descriptions of the MJO over the Indian Ocean, obtained from global gridded reanalyses and reforecasts, will be presented. These have been derived using atmospheric descriptors which have been shown to be present during Indian Ocean convection associated with nascent MJO events but absent during non-MJO convection. Particular emphasis will be placed on the evolution of these variables during the MJO's initiation phase over the Indian Ocean and their projection onto potentially key variables such as surface heat fluxes.