As a sequel to successful ICTP 2002 and 2005 activities, the purpose of the 2008 Colloquium is to facilitate adoption and use of a current state-of-the-art limited area model by participants, take stock of the progress made since 2005, and address issues of active research aimed at advancing as well as understanding the predictive abilities of limited area models.

Overall purpose and specific topics:
Regional or local atmospheric models are a standard tool for prediction of weather at a region or a local area more accurately than this is done by various global models, and are accessible and useful to a wide variety of users with modest computer resources. Yet, how much and what type of improvements over the driver global models can be obtained and for how long remain questions with many unresolved issues.  
This is to be achieved through the following combination of activities:
I. A seven-day workshop on "Design and Use of Regional Weather Prediction Models".
This workshop will include tutorial lectures on the design of regional weather prediction models, emphasizing several features of active research, such as representation of
topography, the parameterization of convection and/or cloud microphysics, land-surface/ boundary layer modelling. The design of a specific model, the Eta Model, widely used by
the research and operational community, will be emphasized, and participants will receive a copy of a recently updated version of the Eta. The workshop will include lab sessions in which students will be taught to install and run the model, and will be encouraged to perform experiments relevant to a specific region or a model feature.

II. A three-day conference on "Current Efforts Toward Advancing the Skill of Regional Weather Prediction. Challenges and Outlook". 
The conference will include invited and contributed papers highlighting some of the following topics:
- Operational and data considerations: limited area systems, including limited area ensembles, current directions;
- Basic numerical design: expectations and results from various choices pursued -- grid point Eulerian/
semi-Lagrangian, conservation/finite-volume issues, nested spectral, stretched grids and variable resolution approaches, other;
- Open questions of the numerical design: nonhydrostatic, how and when; vertical
coordinate/topography, lateral boundary conditions;
- Resolution vs domain size, convergence issues; benefits from "local" vs regional models;
- Physical parameterizations: convection/ explicit cloud physics; boundary layer/subgrid-scale transports;
- Regional weather predictability: value added by a limited area model, downscaling.

"Upscaling"? Limited area ensembles. Storm-scale prediction – how successful is it? Verification issues.

Program:
- Third ICTP workshop on "The Design and Use of Regional Weather Prediction Models",
29 September – 7 October 2008;
- Third ICTP conference on "Current Efforts Toward Advancing the Skill of Regional
Weather Prediction. Challenges and Outlook", 8-10 October 2008.