BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Europe/Stockholm
X-LIC-LOCATION:Europe/Stockholm
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20190719T085744Z
LOCATION:HG EO Nord
DTSTART;TZID=Europe/Stockholm:20190613T195000
DTEND;TZID=Europe/Stockholm:20190613T215000
UID:submissions.pasc-conference.org_PASC19_sess179_post143@linklings.com
SUMMARY:CLW05 - Simulating Hurricane Harvey Using Two Dimensional Flood Mo
 del on Titan and Summit at ORNL
DESCRIPTION:Poster\n\n\nCLW05 - Simulating Hurricane Harvey Using Two Dime
 nsional Flood Model on Titan and Summit at ORNL\n\nGhafoor, Evans, Kalyana
 pu, Kao, Sharif...\n\nOne dimensional flood models do not represent flood 
 wave propagation correctly, especially in urban flood-prone areas, due to 
 inaccuracies in cross-section discretization and the inability to simulate
  lateral diffusion. In 2009, the National Research Council recommended tha
 t the Federal Emergency Management Agency (FEMA) promote "greater use of t
 wo-dimensional (2D) hydraulic models" to model floodplain topography, incl
 uding preferential flood pathways and existing and planned structures. Wit
 h effective use of modern HPC platforms, it is now possible to execute 2D 
 high-resolution hydrodynamic modeling of large-scale regional flood events
 . As a demonstration, we present a hindcast of flooding from Hurricane Har
 vey. Harvey made landfall in southern Texas and was one of the most destru
 ctive hurricanes in U.S. history. Performance improvements in the 2D flood
  model via hybrid distributed parallel technologies including MPI+OpenMP a
 nd MPI+CUDA and modifications to support architecture-agnostic HPC environ
 ments are detailed. At 10 meter spatial resolution, the Hurricane Harvey s
 imulation achieved eight fold speed up on Summit compared to Titan, both m
 ulti-petascale heterogeneous HPC systems at Oak Ridge National Laboratory 
 (ORNL).
END:VEVENT
END:VCALENDAR