Km-scale Ensemble Simulations over Singapore

The UK Met Office, in partnership with MSS, has developed an ensemble system at 4.5 km resolution to help forecast extreme rainfall events around Singapore. The high-resolution ensemble is nested within the ECMWF ensemble model with 51 members and also within MOGREPS-G, the UK Met Office global ensemble model with 18 members.

In this paper, we study the performance of this high-resolution ensemble over several case studies of extreme events. The spread of the ensemble in rainfall is used to analyse the benefits of running an ensemble model over the control member. The forecast uncertainty associated with the upscaling of model errors is studied using the model’s sensitivity to its driving model, the initial and lateral boundary conditions.