Ziel
To quantify the predictability of the atmosphere on seasonal timescales over Europe and regions of the world where there may be European interests, and to test the skill of coupled models of the oceans, atmosphere and land surface, including data assimilation to provide initial conditions.
The first objective is being pursued through a series of coordinated experiments using four European atmosphere
models. These models are being run in ensemble mode over seasonal timescales using observed sea surface
temperature and initial conditions from the ECMWF reanalysis period 1979-1994. This coordinated
experimentation was underway at the beginning of the project, and is likely to be completed by the end of the
first year. Diagnosis of results will take another year, and involves additional participants in the project.
Complex numerical coupled models and ocean data assimilation techniques are being developed by five
participants of the projects, and the ability to run hindcast experimentation will be possible by the second year
of the project.
Testing the skill of the numerical coupled model seasonal forecasts will begin in the second year of the project,
through coordinated experimentation. Special attention will be given to the period 1990 onwards, since the ocean
(especially the El Niqo region), was relatively well observed in that period.
The skill of dynamically based seasonal forecasts will be compared with that obtainable by empirical prediction
techniques. In addition the possibility of using these techniques (eg based on singular value decomposition) to
correct, a posteriori, dynamical model biases will be investigated.
Finally, close coordination between PROVOST and the monsoon project SHIVA will take place throughout the
term of the project. As well as undertaking joint numerical exerimentation, the SHIVA and PROVOST project
will meet jointly each year.