CONTENTS -- Preface -- Predictability of Weather and Climate: From Theory to Practice - From Days to Decades T. N. Palmer -- 1. Introduction -- 2. 20th vs 21st C Perspectives on Predictability -- 3. Why are forecasts uncertain? -- 3.1. Initial Uncertainty -- 3.2 Model Uncertainty -- 4. Ensemble Forecasts: Some Examples -- 4.1. Medium Range -- 4.2. Seasonal Prediction -- 4.3. Climate Change -- 5. Conclusions -- Acknowledgements -- References -- Vector Returns: A New Supercomputer for the Met Office P. Burton -- 1 INTRODUCTION -- 1.1 The Met Office -- 1.1.1 Relocation -- 1.2 The Computers -- 1.3 The Applications -- 1.3.1 The Unified Model -- 1.3.2 Variational Assimilation -- 2 THE PROCUREMENT -- 2.1 The Process -- 2.2 The Winning Bid -- 3 THE NEC SX-6 -- 3.1 The CPU -- 3.2 The Node -- 3.3 The Cluster -- 3.4 The Front-end -- 3.5 The Met Office's System -- 4 TRANSITION TO THE NEW MACHINE -- 4.1 History of UM development -- 4.2 Moving to the SX-6 -- 4.2.1 Optimization for the SX-6 -- 4.3 Performance -- 5 LOOKING AHEAD -- 5.1 Challenges -- 5.2 Opportunities -- 6 SUMMARY -- References -- Vector Parallel Programming and Performance of a Spectral Atmospheric Model on the Earth Simulator S. Shingu, H. Fuchigami and M. Yamada -- 1 INTRODUCTION -- 2 DESCRIPTION OF AFES -- 3 PARALLELIZATION -- 3.1 Selection of Target Resolution -- 3.2 Fortran90 Coding Policy in AFES -- 3.3 Program Flow -- 3.4 Three-level Parallelization -- 3.5 Parallelism of AFES -- 3.6 Inter-node Parallel Processing by MPI -- 3.7 Intra-node Parallel Processing by microtasking -- 4 PERFORMANCE OPTIMIZATION -- 4.1 Optimization for the Legendre Transform -- 5 PERFORMANCE OF AFES -- 5.1 Input Data and Execution Condition for the T1279L96 Test Run -- 5.2 Performance Analysis -- 5.3 Throughput Performance in Data Transpose Communication -- 6 SUMMARY -- ACKNOWLEDGEMENTS -- References.

10-KM Mesh Global Atmospheric Simulations W. Ohfuchi, T. Enomoto, K. Takaya and M. K. Yoshioka -- 1 INTRODUCTION -- 2 AN OVERVIEW OF AFES -- 3 MESO-SCALE RESOLVING T1279L96 SIMULATIONS -- 3.1 Typhoon Genesis Simulations -- 3.2 Wintertime Cyclogenesis Simulations -- 3.3 Baiu-Meiyu Frontal Zone Simulations -- 4 CONCLUDING REMARKS -- 4.1 Discussions -- 4.1.1 Why 10-km Resolution? -- 4.1.2 Why Global? -- 4.2 Future Plans -- 4.3 Conclusions -- Acknowledgements -- References -- Development of Parallel Ocean General Circulation Models on the Earth Simulator Y. Tanaka, M. Tsugawa, Y. Mimura and T. Suzuki -- 1. Introduction -- 2. Basic Equations -- 3. Parallelization and Vectorization -- 4. Parallel computation method for the barotropic solver -- 5. Computational Performance -- 6. Summary -- Acknowledgment -- References -- 4D-Var Global Ocean Data Assimilation on the Earth Simulator N. Sugiura, S. Masuda, Y. Shen, J. D. Annan, T. Awaji, Y. Sasaki and Q. Jiang -- 1. Introduction -- 2. 4D-VAR Data Assimilation -- 3. Our 4D-VAR system -- 3.1. Preconditioning -- 3.2. Pamllelixation -- 4. Assimilation Results -- Acknowledgements -- Appendix A. Procedures of ID-VAR data assimilation -- Appendix B. Preconditioning Method -- Appendix C. Components of 4D-VAR system -- Appendix D. Cost Function -- References -- Implementation of the IFS on a Highly Parallel Scalar System M. Hamrud, S. Saarinen and D. Salmond -- Introduction -- Operational forecasting system -- 1. Data assimilation -- 2. Global atmospheric forecasts -- 3. Ocean-wave forecasts -- 4. Global Seasonal forecasts -- Planned improvements to forecasting system -- HPC procurement in 2001 -- Outcome of the procurement -- IFS system overview -- Migration timetable and status -- Ongoing optimization -- Error trapping -- Observation handling in IFS -- High volume of satellite data -- About ODB -- File I/O.

Recipes jor reducingfile count -- Use of OpenMP -- Design of IFS for vector, scalar, MPP and SMP systems -- Optimization -- Profiles from High Resolution Forecast,from benchmark - T799 L90 -- OpenMP + MPI in High Resolution Forecast from benchmark- T799 L90 -- IFS 40-Var run on IBM p690 with mixed OpenMP and MPI -- Summary -- Performance and Scalability of Atmospheric Models on LINUX Systems S. Lowder and T. E. Rosmond -- 1. Introduction -- 2. Our Models -- 3. First Steps -- 4. Computation - Single Processor Performance -- 5. Computation - Linux Cluster Performance -- 5.1. Model Resolution -- 5.2. Tasks per Node -- 5.3. Compiler Comparison -- 5.4. MPI Versions -- 6. Communication -- 6.1. Network Handware Bandwidth -- 6.2. Packet sizes -- 6.3. Bidirectional Bandwidth -- 6.4. Latency -- 7. I/O - The Achilles Heel -- 8. Overall Comparison -- 9. Non-Performance Considerations -- 10. Summary -- References -- The NOAA Operational Model Archive and Distribution System (NOMADS) G. K. Rutledge, J. Alpert, R. J. Stouffer and B. Lawrence -- 1.0 Introduction -- 2.0 Background -- 3.0 Benefits -- 4.0 System Architecture -- 4.1 Open source Project for a Network Data Access Protocol: OPeNDAP -- 4.2 The Live Access Server -- 4.3 GrADS-Data Server -- 4.4 LLNL's Climate Data Analysis Tool -- 4.5 The Thematic Real-time Environmental Data Distributed Services Project (THREDDS) -- 4.6 Distributed Metadata Server -- 5.0 NOMADS Data Availability -- 5.1 NCEP Model Input Data -- 5.2 The Weather Research and Forecast Model -- 5.3 GFDL GCM Data Availability -- 5.4 NCAR Data Availability -- 5.5 Climate Diagnostic Center -- 5.6 NASA's Global Change Master Directory -- 6.0 The Collaborators -- 6.1 Other Resources and Collaborator's -- 6.1.1 Grid Computing -- 7.0 Acknowledgments -- 8.0 Appendix A. Acronyms -- 9.0 References.

Data Ingest and Product Distribution for Teracomputing: A Scalable, Flexible E-Business Architecture B. R. Gritton, K. Pollak and M. A. Rennick -- 1. Background -- 1.1. Building the Business Case - Drivers for Transformation -- Modeling Roadmap - Teraflop Computinq -- Net-Centric Warfare (NCW) & Information Superiority -- METOC-Warfighter Workflow -- Web Enabled Navy (WEN) policies -- Affordable IT approach -- Enternrise Architecture perspectives -- Moore's Law -- E-Business Technologies -- Open Standards and Software -- Commodity Clusters -- 1.2. Conceptual Architecture - A Warfighter Enabled Infrastructure -- 1.3. Implications of the Accordion Architecture -- 2. Status -- Summary -- References -- PRISM and ENES: A European Approach to Earth System Modelling E. Guilyardi, R. G. Budich and S. Valcke -- 1. Introduction -- 1.1. ENES and the European climate research challenges -- 1.2. The PRISM infrastructure -- 2. Enhancing physical modularity in Earth system models -- 2.1 Earth System model components -- 2.2 Main scientific principles in Earth System Models -- 2.3 Standard PRISM physical and algorithmic interfaces between components -- 2.3.1. The Surface Layer Turbulence module -- 2.3.2. The Ocean-Surface module -- 3. Overall system architecture and Graphical User Interface -- 3.1. Experiment Configuration -- 3.2. Experiment monitoring and scheduling -- 3.3. The local system -- 3.4. The Web Service architecture -- 4. Coupling and I/O software -- 4.1. Coupled Model High Level Architecture -- 4.1.1. Definition phase -- 4.1.2. Composition phase -- 4.1.3. Deployment phase -- 4.1.3. Deployment phase -- 4.1.4. Deliverable -- 5. Visualisation and diagnostic library -- 6. Target computer platforms, software engineering and quality standards -- 7. Conclusion -- Acknowledgements -- References.

Implementation of Data Mining Techniques for Meteorological Applications A. S, Cofino, J. M. Gutierrez, B. Jakubiak and M. Melonek -- 1 INTRODUCTION -- 2 DISTRIBUTED DATABASES -- 2.1 ECMWF & NCEP reanalysis databases -- 2.2 Database of the mesoscale forecasts for Central Europe -- 2.3 Databases of local observations in Spain and Poland -- 3 DOWNSCALING AS A DATA MINING PROBLEM -- 4 IMPLEMENTED DATA MINING TECHNIQUES -- 4.1 SOM. Self-organizing Maps -- 4.2 MCMC. Markov Chain Monte Carlo -- Acknowledgements -- References -- An Integrated E-Science Environment for Environmental Science K. Kleese van Dam, S. Sufi, G. Drinkwater, L. Blanshard, A. Manandhar, R. Tyer, R. Allan, K. O'Neill, M. Doherty, M. Williams, A. Woolfand L. Sastry -- 1 INTRODUCTION -- 2 INTEGRATED E-SCIENCE ENVIRONMENT FOR CLRC (IESE) -- 2.1 The CLRC DataPortal -- 2.1.1 What can you do with it? -- 2.1.2 How does it work? -- 2.1.2.1 Metadata -- 2.1.2.2 Dataportal Architecture -- 2.2 HPCPortal -- 2.2.1 HPCPortal Architecture and Services -- 3 PROJECTS -- ENVIRONMENT FROM THE MOLECULAR LEVEL: AN E-SCIENCE PROPOSAL FOR MODELLING THE ATOMISTIC PROCESSES INVOLVED IN ENVIRONMENTAL ISSUES -- e-HTPX: A HIGH THROUGHPUT RESOURCE FOR PROTEIN CRYSTALLOGRAPHY -- THE NERC DATAGRID -- EUROPEAN SPATIO-TEMPORAL DATA INFRASTRUCTURE FOR HIGH-PERFORMANCE COMPUTING -- MSC PROJECT ON AUTOMATED DATA MANAGEMENT FOR CLIMATE SIMULATIONS -- 4 SUMMARY -- References -- Toward the Flexible Composition and Deployment of Coupled Models R. W. Ford and G. D. Riley -- 1. Introduction -- 2. Requirements -- 3. Solutions -- 3.1. Subroutine implementation -- 3.2. Only compute the science -- 3.3. No direct I/O -- 3.4. Model conformance -- 3.5. Auto-generation of control -- 3.6. Framework adaptors -- 3.7. Transformations are models -- 3.8. DCD -- 3.9. Summary -- 4. Conclusions -- References.

From Megaflops to Teraflops - The 10th ECMWF Workshop G.-R. Hoffmann.