CARDAPIO Research
Continental rainfall in the tropics is of central importance to a large fraction of the world’s population, and presents a major scientific challenge, both in terms of its fundamental nature and its prediction. Tropical regions have rainfall regimes that display a marked seasonal cycle but are also subject to large variations from one year to another and within a season. Understanding and predicting these variations is crucial for the basic needs of society, including water management, agriculture and power generation.
Variations in continental rainfall are influenced by the response of the tropical atmosphere to anomalies in the sea surface temperature (SST). This is a relatively long-lived component of the system, often persisting throughout the season. Persistent SSTs give potential predictability to the fast-reacting atmosphere. The atmospheric response consists of tropical waves and changes to the large-scale circulation, which in turn impact continental regions, modifying the rainfall. Slowly-varying tropical modes of the atmospheric circulation, such as the Madden-Julian Oscillation (MJO) also play a role.
Seasonal prediction is normally carried out by national meteorological services using state of the art General Circulation Models (GCMs) of the atmosphere and ocean. And yet some of the basic processes involved are amenable to investigation using simpler, more idealised methods, which are more readily adopted by academic institutions with limited resources. This complementary approach can contribute to assessment and improvement in the national prediction effort.
In CARDAPIO we aim to study rainfall regimes around the tropical band using a variety of scientific approaches. These will include dynamical modelling of the atmosphere and statistical modelling of the links between tropical SSTs and rainfall variability. We will test our ideas in a concrete way by making predictions. An example of this philosophy in action is the DREAM seasonal forecasting system, currently running in research mode at FUNCEME in Fortaleza. Historical predictions with DREAM have reached a level of skill comparable with GCM simulations that are much more computer-intensive.
We have gathered together a consortium of atmospheric scientists from tropical regions in three continents, to study regional rainfall variations in Brazil, West Africa and South-East Asia. The partners in CARDAPIO all have a history of collaboration with the French host laboratory LEGOS, but have less experience in working with one another. The partner institutions are also closely linked to education programs that have a strong association with LEGOS. So a central part of this project concerns the training of masters and PhD students, including internships, workshops and technical training sessions.
There is much to be gained in comparing methods and approaches around the tropical band. The project partners experience rainy seasons at different times of year, but are all subject to global influences from SSTs in the three major ocean basins. We intend to share our experience with national meteorological services and foster synergy between academic study and operational prediction. Our simulations will also provide a useful set of test data for impact assessments, for example using hydrological models. Impact assessment requires a proper appreciation of the entire chain of causality, from the SST to societal needs. Such an end-to-end transdisciplinary approach can provide a practical vision of the utility of predictions, and the uncertainty involved with each link in the chain.
