|Title||Representation of the tropical intraseasonal variability and its impact on seasonal predictability in a multi-model ensemble|
|Year of Publication||2007|
|Authors||Xavier, PK, Duvel, J-P, Doblas-Reyes, F|
|Secondary Title||Technical Memorandum|
A set of multi-model ensemble seasonal hindcasts produced within the framework of the DEMETER project over the period 1980-2001 has been used to analyze the tropical intraseasonal variability. The focus is on the spatial and seasonal variations linked to the summer monsoon intraseasonal oscillations (ISO) of outgoing longwave radiation (OLR), their large-scale organization, propagation, air-sea coupling, deterministic predictability and implications for seasonal predictability. A multi-variate method known as Local Mode Analysis (LMA, a complex EOF analysis over a 90-day moving window) has been employed to evaluate the ISO's characteristics. Most models have problems to simulate large-scale organized convection. In addition, the patterns are more variable from one intraseasonal event to another compared to observations. Models exhibit some form of northeastward propagation over the Indian Ocean. Realistic periods of the modes (25-35 days) are produced in a few models, while the rest produce shorter periods (20-25 days). Models with a poor seasonal cycle tend to have larger biases. One possible source of deficiency is identified to be the air-sea interaction. The analysis of the coupling shows that most models simulate too weak SST perturbations and systematic phase quadrature between OLR and SST. The hindcasts carried out with the same AGCM and different OGCMs tend to have similar biases, prompting at the importance of atmospheric processes in defining the nature of the intraseasonal SST perturbation. Evaluation of the predictability at the ISO time scale (10-50days) is performed using pentad mean OLR maps. Results show a better predictability in the summer (May start date) hindcasts compared to the winter hindcasts (November start date). Other intraseasonal and seasonal predictability analyses are also presented, with emphasis on the influence of large-scale organized ISO perturbations on the regional seasonal evolution of the precipitation.