A U.S. National Multi-Model Ensemble ISI Prediction System

The proposal intends to use an existing National Multi-Model Ensemble (NMME-1)

that is already producing routine real-time intra-seasonal, seasonal and interannual (ISI) predictions as an NMME of opportunity (understood as a system where each member contributing to the multi-model has developed independently their prediction protocol). The objective is to develop it further and produce a rigorous reforecast experiment and evaluation towards an operational NMME (NMME-2) protocol. The objectives are:

- Build on existing state-of-the-art US climate prediction models and data

assimilation systems that are already in use in NMME-1 and ensure

interoperability so as to easily incorporate future model developments.

- Take into account operational forecast requirements (forecast frequency, lead

time, duration, number of ensemble members, etc.) and regional/user specific

needs. A focus of this aspect of the work will be the hydrology of various

regions in the US and elsewhere in order to address drought and extreme event

prediction.

- Utilize the NMME-2 system experimentally in a near-operational mode to

demonstrate the feasibility and advantages of running such a system as part of

NOAA?s operations.

- Enable rapid sharing of quality-controlled reforecast data among the NMME

team members, and develop procedures for timely and open access to the data,

including documentation of models and forecast procedures, by the broader

climate research and applications communities.

- Undertake the evaluation and optimization of the NMME system in hindcast

mode (e.g., assessing the optimal number of ensemble members from each

model, how to best combine the multi-model forecasts, sources of

complementary prediction skill, etc.).

- Use the NMME forecasts for regional downscaling and hydrological prediction.

The partners consider undertaking research on combination methods, model

independence, prediction of droughts as extremes of societal relevance, development of an optimal sub-seasonal prediction system (with a focus on phenomena like the MJO), and impact of the initialization of different components of the system (land and ocean).