On-Demand Severe Weather Forecasts Using TeraGrid via the LEAD Portal

摘要: This work examines the use of TeraGrid to produce on-demand severe weather forecasts as part an experiment being conducted within Hazardous Weather Testbed (HWT) spring program. effort is a cooperative project between members Linked Environments for Atmospheric Discovery (LEAD) [1] ‒ National Science Foundation (NSF) Information Technology Research (ITR) and Oceanic Administration (NOAA) units, most notably Storm Prediction Center (SPC) Severe Storms Lab (NSSL). Among other responsibilities, SPC issues official Thunderstorm Tornado watches across United States. The HWT program 7-week organized by NOAA meteorologists that brings in visiting scientists from operations, research institutions private sector order test evaluate new forecasting tools designed aide prediction threat hazardous [2] Several groups special ensembles highresolution numerical (NWP) products effort. These are used guidance experimental manual also evaluated their own merit [3]. Most these initialized at 4 pm or 7 CDT previous day they run on fixed schedule over large domain. It desirable have adapt utilize observation data current day, especially including time when solar heating can affect boundary layer. Thus LEAD proposed provide covering smaller, adaptable, target area be with mid-morning (10 am CDT, 15 UTC) day. location domain tailored needs HWT, based input participants. On-demand were produced 2007 currently 2008 In separate effort, similar automatically generated operational forecast products. composed submitted using Portal. Portal top tier serviceoriented architecture allows seamless integration data, analysis tools, grid computing resources intuitive web interface [4, 5, 6]. There three primary goals experiment: 1) report capabilities providing forecasters, 2) identify additional user future development Portal, 3) services resource support portal services. important highlight this two perspectives: software, automated resources, under continuous refinement real adaptable convective-scale system process relatively new. paper will summary logistical aspects process, briefly summarize ongoing verification predictions 2007, describe