International Data Acquisition for Disaster Prevention and Data

International Data Acquisition for Disaster Prevention and Data Access Technology Hyunseung Choo Sungkyunkwan University http: //monet. skku. ac. kr choo@skku. ac. kr

Contents ■ Introduction ■ Earth System ü weather prediction, earthquakes, and environments ■ From Networking to Grid Computing ü ET over Grid ■ The Proposal ü ü ü Importance Current status Proposed approach Final goals Members and organizations ■ Conclusion

Technology for the Weather Forecasting ■ ■ Observations Data Collection Data Analysis Weather Forecasting

Global Infrastructure for Earth System Nature

From Networking to Grid Computing (1/2) ■ Exponential Growth of Network Technology ü Network vs. Computer Performance u Computer speed doubles every 18 months u Network speed doubles every 9 months u Difference = order of magnitude per 5 years ü 1986 to 2000 u Computers: x 500 u Networks: x 340, 000 ü 2001 to 2010 u Computers: x 60 u Networks: x 4000

From Networking to Grid Computing (2/2) ■ Definition of GRID Research infrastructure for dynamic collecting and sharing research resources connected by high-speed networks Supercomputer GRID Large Scale Data Problem solving Scientific Research Instrument Experts

Grid Based Resource Integration ■ Challenge to huge problems that require massive capability over limited resources at multiple sites ü Grid Computing u Networking of locally distributed high performance (HP) resources u HP computers, storages, experimental facilities, etc. u IT infrastructure to group systems as a single system u Classified into Access Grid, Computational Grid and Data Grid ■ Core applications based on Grid ü “IT + Application” : Environment Technology (ET), Bio Technology (BT), Nano Technology (NT), Space Technology (ST), Cultural Technology (CT), etc. ■ Investigated HPCN since late 1980 in US, JP, EU ü Concentrate to Grid Technology since 2000

Pan-Korean Distributed HPC Resources • Access Grid System • Supercomputers • Experimental Facilities • Computers Chuncheon Seoul Incheon Suwon Chonan Chongju Daejeon Pohang Jeonju Daegu Kwangju Changwon Ulsan Pusan • Databases • Mass Storage • Visualization Environment Cheju • High Speed Networks

Importance of R&D Project ■ KOREN based distributed Grid Computing environments ü Overcome limitations of the meteorology area through Grid ü ü u Limitations on computing, networking and data storing Solve earlier problems based on disaster prevention data Data format for providers ⇒ Data format for consumers Offer common collaborative environments to weather experts Substitute clusters for supercomputers ■ Disaster prevention system by forecasting natural disasters ü RT data transmission, high-speed producing and processing ü Parallel technology for numerical models on natural disasters u Low-cost RT disaster prediction ü Severe weather prediction, environments, East Sea earthquake tsunami (a tidal wave) u Currently depend on GTS in the Japan Meteorological Administration

Final Goals on R&D Project ■ KOREN based Grid system for international disaster prevention KMA NIED (JP) SKKU SWU PKNU UCLA (US) Nagoya U. (JP) ■ KOREN/APII based international joint production of disaster related prevention data ■ development of high speed data access and process system model ■ parallelization of current SWs and sharing international disaster prevention resources

Key Elements on R&D Project ■ Current year ü Sharing data for weather prediction with Japan (APII, Nagoya Univ. ) ü Sharing East Sea seismological information (National Research Institute for Earth Science and Disaster Prevention of Japan) ü Numerical models for disaster preventions in Grid environments ü Access Grid for multicasting based on KOREN ■ Current year + 1 ü Sharing and verifying weather data (UCLA) ü RT transmission of disaster prevention data ü Development of applications for disaster preventions (MPICH-G 2) ■ Current year + 2 ü Use of Data Grid for massive data transmission ü Use of international core technology for disaster preventions (KOREN, APII)

Data for Disaster Preventions ■ International data collection for weather prediction, climates, earthquakes and environments for disaster predictions ü RT and distributed collection of weather prediction, climates, earthquakes and environments over the entire earth u Access Grid ü Collaborating organizations for international data sharing u Based on APII-testbed ■ Disaster prediction modeling based on the data collection ü RT locating/allocating distributed computational resources u Computational Grid ü Data acquisition for analyzing and forecasting global weather prediction ■ RT delivery systems for disaster predictions information ü KOREN + Grid technology => constructing IT infrastructure

Members and Organizations Collaborative working environment for producing disaster prevention data (PKNU(KR), Nagoya Univ. (JP)) Communication infrastructure for disaster prevention (SKKU(KR), SWU(KR)) KOREN APII-Testbed • Protocols for monitoring seismic activity • Virtual collaboration environment Seismological observation network between Korea and Japan Production of data for weather prediction by using Grid (KMA(KR), NIED(JP)) (PKNU(KR), UCLA(US)) KOREN & APII-Testbed Based Virtual Collaboration Environment For Disaster Preventions

Approaches (2003) UCLA (US) Weekly Seminar Invited Talks RT data exchange for weather prediction Nagoya Univ. (JP) NIED (JP) Analysis & sharing of massive seismological & meteorological data KMA (KR) Policies for sharing data among parties SKKU (KR) SWU (KR) Access Grid (AG) - Data authorization - International Production of data for severe weather prediction based on AG -AG environments cooperation - Collecting individual opinions -Securing AG technique throughout six parties for KOREN 2 nd Year KOREN/APII-Testbed Based Virtual Collaboration Environments Access Grid Demo 1 st Report PKNU (KR)

Approaches (2004 -2005) PKNU (KR) UCLA (US) Nagoya Univ. (JP) NIED (JP) KMA (KR) SKKU (KR) SWU (KR) AG based environments, earthquake and severe weather prediction data KOREN, APII-Testbed based international environments, earthquake and severe weather prediction data MPICH-G 2 -Applications in Grid environments Data Grid (DG) -Integrated management -Management of distributed data International Conference 2 nd Reports Results from 2003 Weekly Seminar Invited Talks From 1 st Year Current year +1 Combining AG + MPICH-G 2 + DG Constructing IT Infrastructure KOREN, APII-Testbed based data production for international environments, earthquakes and severe weather prediction, distributed data access techniques and its applications International Conference Final Reports Results from 2004 Weekly Seminar Invited Talks Current year +2

Research Goal (2003) ■ Access Grid connecting KMA, SKKU, SWU and PKNU by KOREN ■ Seismological observation network between Korea and Japan by APII-testbed ü Extended TCP/UDP protocols for high-speed communications ■ Data production line for disaster preventions based on MPICHG 2/GQ ■ Problem detections and solutions on operations for collaborative working environment based on KOREN/APII-testbed ■ Proceeding to the next year ü Tuning among participating organizations

Example: Organization for Access Grid

Example: Applications in Access Grid ■ Establishment of collaborative working environments among parties including universities, laboratories and offices

Restrictions in R&D Project ■ Lack of techniques to establish efficient collaborative working environments ü Sharing computational resources through high speed networks u supercomputers, cluster systems, massive storages ü Policies through the tuning of management strategies ■ Lack of human resources in Grid environments ü Need re-education programs ü Invited talks and international technical forums

Applications and Effects “ET over IT” Construction of co-work environments based on KOREN and APII-testbed ü Improving National Security by establishing international networks for disaster preventions ü Radical pervasive effect through a technology transfer ü Create applications using KOREN and APII-testbed “α over IT” → Unification Effect on various areas

Conclusion ■ Production and practical use of data for disaster preventions in temporally and spatially distributed computing environments ü combining KOREN with Grid computing technology ■ Raising human resources for Grid technology ü Producing detailed user guide books/manuals and sharing knowledge through them ■ The synergy effect caused by clustering supercomputers through KOREN ■ Efficient use of world best IT infrastructure