Prediction, Predictability, and Multi-Decadal Variability of the South Asian Monsoon

Abstract

Advances made in seasonal prediction and predictability of the south Asian monsoon (SAM) over the past decade is briefly reviewed. To achieve potential skill of prediction of SAM rainfall, a major targeted effort to improve biases of simulation of precipitation over the Asian monsoon region in a coupled model is described. A series of retrospective forecasts by improved versions of the model indicate that skill of SAM rainfall beyond the potential predictability limit (PPL) estimated from signal to noise (S/N such as ANOVA) is achievable, indicating that S/N method is inadequate in estimating the PPL for SAM rainfall. A new model based method of estimating PPL shows that the SAM rainfall is much more predictable than thought previously. Scientific basis for this enhanced predictability of the SAM rainfall is linked to the emergence of the extratropical SSTs associated with the AMO (in addition to the tropical SSTs associated with the ENSO) as a major contributor to the predictability and the recognition that the statistics of the high frequency fluctuations (e.g. variance of the synoptic disturbances) are tied to predictable drivers and hence partly predictable. We shall also review advances made in understanding and modeling of the SAM multi-decadal variability as it not only modulates the skill of prediction and predictability of the south Asian Monsoon but also makes the human induced green house influence on the SAM difficult to be isolated.