Monsoon Intraseasonal Oscillation: Theoretical basis and practical strategty for improving prediction

The wet and dry spells of the monsoon intraseasonal oscillation (MISO) strongly influence extreme hydro-meteorological events, major driving forces of natural disaster, and thus the socio-economic activities in the World's most populous monsoon region. Understanding the origin and perpetuation of the monsoon cycle has eluded scientists for decades. Improved understanding of the physical mechanisms is a prerequisite for improvement of MISO. Compared with the Madden-Julian Oscillation (MJO) the MISO is more complex in nature, with prominent northward propagation and variability extending much further from the equator. The rainy phase of the MISO usually initiates over the western equatorial Indian Ocean and propagates northward along with an eastward-propagating component. An account of essential aspects of MISO is presented, including (a) what sustains MISO or why a new rainy phase is initiated in the western equatorial Indian Ocean, (b) How a titled ISO rain band is formed, (c) why the MISO rain bands move northward in the Asian-Pacific and North American monsoon regions, (d) how the MISO-ocean interaction can play an important role in MISO dynamics, and (e) how the MISO interact with mid-latitude wave trains. Practical useful MISO indices are proposed for monitoring and prediction purpose based on multivariate empirical orthogonal function (MV-EOF) analysis of daily anomalies of outgoing longwave radiation and zonal wind at 850 hPa in the region 10oS-40oN, 40o-160oE, for the extended boreal summer (May-October) season over the 30-year period 1981-2010. The prediction skill and predictability of the MISO are also examined in terms of predictable modes identified form observation and multi-model ensemble (MME) hindcasts obtained from ten coupled models participated in the IntraSeasonal Variability Hindcast Experiment (ISVHE) project. It is noted that the first two MV-EOF modes are predictable using the coupled models and the bivariate temporal correlation coefficient skill for the two modes reaches 0.5 at 22-day forecast lead for the best model but at 5-day for the worst model. As a phenomenon bridging synoptic weather and seasonal variability, MISO predictability is strongly affected by both the initial conditions and air-sea coupling. Some existing reanalysis datasets have poor representation of the MISO. A ‘signal-recovery' strategy has been developed to enhance MISO signal in initial conditions, which results in much improved forecasting skill.