[Vortraege] Urgent Computing: A real use case

LRZ aktuell publish at lrz.de
Mo Nov 5 16:51:45 CET 2012

Im Rahmen der Vortragsreihe
 Professioneller IT-Betrieb in mittleren und grossen Umgebungen
 sind Sie zu den folgenden Vortraegen herzlich eingeladen (Anmeldung
 nicht erforderlich).
 Large-scale on-demand computations are becoming more and more crucial
 for decision making in critical environments and situations. Urgent
 computing seems to be a promising new branch of computer science
 addressing prioritized access to modern e-infrastructures. The session
 will explore fundamental aspects of Urgent Computing using
 hydro-meteorologic urgency as an example.
 Alle Vortraege werden auf Englisch gehalten. | All talks will be given
 in English.
           Urgent Computing: Late Results are Useless Results           
                          Dr. Michael Schiffers                         
                    (Munich Network Management Team)                    
 Large-scale collaborative computations are becoming more and more
 crucial for decision making in critical medical environments and
 disaster prediction and preventing situations.
 Urgent computing is a promising new branch of computer science
 addressing algorithms, methods and tools enabling the required
 prioritized and immediate access to e-infrastructures.
 The talk will explore urgent computing in more detail using the
 EU-funded DRIHM project as use case.
 Dr. Michael Schiffers is a member of the Munich Network Management
 (MNM) team at LMU. His research interests cover Grid Computing,
 Dependable Systems, and Urgent Computing. He is a member of the
 Steering Committee of the EU-funded DRIHM project.
 [Beginn: 15:15 Uhr, Dauer: 15 Minuten]
       High-resolution meteorological simulations of catastrophic       
                          Mediterranean floods                          
                           Dr. Antonio Parodi                           
                       (CIMA Research Foundation)                       
 On November 4th, 2011, the city of Genoa, Italy, was witnessing a
 catastrophic flash flood resulting from more than 500 millimeters of
 rain in approximately 6 hours.
 In the talk we analyze the performance of model simulations related to
 this extreme event using the Advanced Research Weather and Forecasting
 Model focusing on the aspect of model quantitative precipitation
 forecasts (QPFs) for flash flood predictions. The simulations performed
 at the LRZ SuperMUC suggest the crucial role of state-of-the-art high
 performance computing resources to better understand these extreme
 hydro-meteorological events. This aspect is especially explored in the
 EU-funded DRIHM project.
 Antonio Parodi, Ph.D. is a Project leader at the Italian CIMA Research
 Foundation in Savona. His research interests cover modeling of dry and
 moist convection with special emphasis on uncertainties. He is author
 and co-author of several publications in international peer-reviewed
 journals. He is the Project Director of several Hydro-Meteo EU-funded
 projects, including DRIHMS, DRIHM and DRIHM2US.
 [Beginn: 15:45 Uhr, Dauer: 25 Minuten]
 Progress and challenges in coupling hydrological and atmospheric models
                  for Earth System prediction research                  
                            Dr. David Gochis                            
          (US National Center for Atmospheric Research (NCAR))          
 The need for improved assessments and predictions of many key
 environmental variables is driving a multitude of model development
 efforts in the geosciences highlighting the role of model coupling as a
 fundamental activity itself and, at times, a significant bottleneck in
 Earth Systems and climate impacts research.
 This talk explores some of the recent activities and progress that has
 been made in assessing the attributes of various approaches to the
 coupling of Earth Systems models. As an example we will present the 
 Weather Research and Forecasting (WRF) Hydro system. The talk will also
 discuss the opportunities for fostering open, standards-based
 approaches for code development, model interoperability and data and
 metadata structures.
 Dr. David Gochis is a member of the US National Center for Atmospheric
 Research (NCAR). His academic background is inter-disciplinary between
 the meteorological and hydrological sciences and civil and agricultural
 engineering disciplines.
 Prior to returning to academia Dr. Gochis worked as a consulting
 engineer for water resource evaluations and the design of irrigation
 His current research interests include hydrometeorology,
 hydroclimatology, atmospheric convection and land surface hydrology
 where he focuses on observation, diagnosis and modeling of warm season
 precipitation and runoff processes in complex terrains.
 Dr. Gochis participates in the DRIHM2US project jointly funded by the
 European Commission and the US NSF.
 [Beginn: 16:30 Uhr, Dauer: 25 Minuten]
       Designing a Hydro Meteo Infrastructure: The DRIHM Approach       
                        Prof. Dr. Andrea Clematis                       
     (Institute of Applied Mathematics and Information Technologies)    
 We shortly review the challenges and outline the strategy adopted in
 the DRIHM project to design a new e-science infrastructure dedicated to
 meteorological and hydrologic research and covering the whole chain of
 models from meteorological forecast via hydrological effects study to
 impact analysis.
 The core of this talk is the presentation of the DRIHM aspects that
 characterize the approach to the design of this infrastructure:
   * The strong involvement of end users
   * The learning by doing approach
   * The exploitation of European e-infrastructures
   * The global interoperability requirement
 Prof. Dr. Andrea Clematis leads as Research Director the high
 performance and Grid computing group at the Institute of Applied
 Mathematics and Information Technologies of the Italian National
 Council of Research in Genova. He chairs the Euromicro Technical
 Committee and is editor of the Journal of Systems Architecture. He is
 the Italian representative of the Managing Committee of the Open
 European Network for High Performance Computing on Complex
 His research interests include Grid and parallel computing, distributed
 visualization and image processing.
 He has co-authored more than 150 technical papers and he acted as
 project leader for more than 20 international research projects. He
 leads the DRIHM Work Package on Use Case and Application Requirements
 [Beginn: 17:05 Uhr, Dauer: 25 Minuten]
        Termin: Do. 8.11.2012 von 15:15 bis 17:40 Uhr                   
           Ort: LRZ-Gebaeude; hinsichtlich des Raums siehe das          
                Hinweis-Schild am Eingang                               
 Vortragsreihe: Weitere Informationen (u.a. das komplette Programm und  
                Kurzfassungen aller Vortraege): www.lrz.de/vr-it-betrieb

 Diese Information finden Sie im WWW unter

 Ernst Boetsch

Mehr Informationen über die Mailingliste aktuell