A promising alternative to prediction of seasonal mean all India rainfall

Abstract

Prediction of seasonal mean All India Rainfall (AIR) is useful during extreme monsoon years (droughts and floods) when the rainfall anomaly is homogeneous over the country. It is, however, useless for any regional hydro-meteorological applications during 'normal' monsoon years (70 per cent of available record), when the rainfall anomaly is quite inhomogeneous within the country. Further, there exists an intrinsic limit to predict the seasonal mean monsoon. The theoretically achievable skill (with perfect model and near perfect data) for seasonal prediction of rainfall being barely useful, there is a need to explore an alternative strategy for monsoon prediction even if it is with a shorter lead time. Based on some of our previous work, we propose here that predicting the phases of the monsoon sub-seasonal oscillation (active and break spells) 3-4 weeks in advance is such an alternative strategy. We argue that such predictions would be more useful for regional hydro-meteorological applications. Potential for such extended range prediction is demonstrated. Using an empirical model, it is further demonstrated that this potential can be achieved and useful prediction of monsoon breaks three weeks in advance could be made. Future direction in improving such extended range prediction of sub-seasonal spells is discussed.