Evaluation of monsoon seasonality and the tropospheric biennial oscillation transitions in observations and CMIP models

Abstract

The Indian and Australian summer monsoon systems have considerable socioeconomic and environmental importance. Here we investigate monsoon seasonality, biennial variability and the interaction with Tropical sea surface temperatures (SST) in the Indo-Pacific sector. We consider a variety of observational and reanalysis products and also assess climate models from the Coupled Model Intercomparison Project Phase 3 and 5 (CMIP3 and CMIP5). In particular, the transitions between successive Indian and Australian monsoons, that form essential parts of the Tropospheric Biennial Oscillation (TBO) have been evaluated. We use Monte Carlo statistical techniques to examine the predictive skill that is inherent in these monsoon transitions and investigate the possible teleconnections between SST anomalies in the Indo-Pacific region, particularly associated with the El Niño-Southern Oscillation (ENSO).

Most climate models reproduce enhanced rainfall in the correct seasons for both the Indian and Australian monsoons. However, there are a number of biases in wet season duration and rainfall strength. While little improvement is seen in the overall strength of the monsoon rainfall from CMIP3 to CMIP5, there is a clear improvement in the seasonality particularly in simulating low rainfall rates outside of the monsoon season.

Enhanced predictability associated with the Indian-Australian monsoon in-phase transition is present in all observational and reanalysis datasets and most CMIP climate models, i.e. we have some skills in predicting whether an Australian monsoon will be stronger/weaker than normal, given information on the strength of preceding Indian monsoon. The SST anomalies in the Niño 3.4 region in December-March (DJFM) after the Indian monsoon season appear to be important for this transition. For the Indian-Indian monsoon out-of-phase transition, enhanced predictability only occurs in long-term observations but with little consistency across models. DJFM SST anomalies in the Niño 3.4 region over successive years appear to strongly affect this transition. The enhanced predictability for the other transitions shows little consistency between observational and reanalysis datasets, climate models and time periods. Multi-decadal variability in the TBO transitions is clearly seen in both observational and reanalysis products and climate models.