Climate change, the South Asian monsoon and model mean state biases  Seminar

This talk examines the effects of anthropogenic warming on the South Asian monsoon, which is of great importance to society since it provides around 80% of annual rainfall during the months of June to September. In general, projections of mean monsoon rainfall for South Asia indicate increases, owing to an enhanced moisture source from the warmer Indian Ocean. However, there are large discrepancies in the pattern of change in CMIP3/5-class coupled GCMs, and recent research shows that large biases persist in the mean state of the Indian Ocean region, including a dry bias over India and wet bias over the western equatorial Indian Ocean. When efforts are made to select models based on their skill at simulating the present day monsoon, the general pattern of an increased mean monsoon in the future remains, albeit with considerable decadal variability. For society, however, it is the variability that is most important and we demonstrate some of the projected changes (and their uncertainties) to interannual, intraseasonal and extremes of monsoon variability in the future. We also look at the predictive nature of the seasonal monsoon via its teleconnection with ENSO and show that improving mean state biases in the tropical Indo-Pacific Ocean can lead to enhanced monsoon predictability. We also use this to examine the behaviour of the UK Met Office Global Seasonal Forecast Model, GloSea4 in predicting seasonal monsoon rainfall. Finally we look at some work examining monsoon biases in the latest versions of the UK Met Office Hadley Centre Model HadGEM3.