Weather Gallery

Yarnell Hill Fire (Ignition: 6/28/13 near Yarnell, Arizona)

Principle Investigator: Janice Cohen (NCAR)
Collaborator: Wilfrid Schroeder (Univ. of Maryland)

On June 30, 2013, 19 firefighters were killed during the Yarnell Hill fire, when a gust front coming from the northeast blew across the fire, changing its direction, and making it spread rapidly across where they were sheltered.

The animation depicts a CAWFE coupled weather-wildland fire model simulation of the Yarnell Hill Fire.  The horizontal resolution was 370 m, a vector is shown each 4 model grid points.  It begins at 2 am on 6/30/13. The fire is initialized in the model using the ~3 am VIIRS active fire detection map.  Each frame is 1 minute apart, the sequence extends until 8:15 pm on 6/30. The fatality occurred around 4:45 PM. The color bar on the right indicates the heat flux (watts per square meter) from the fire, with more intensely burning areas in bright yellow and white, and less intensely burning areas in darker reds. In the simulation, solar heating stirs up the boundary layer circulations throughout the day. Convection occurs in outer domains (not shown) to the northeast, creating high-based convective clouds as air flows south/southeast over the Mogollon Rim. Rain falls into a very dry boundary layer,  creating a broad gust front that  reaches the south edge of the fire at frame 936, which is 51 minutes after the fatality, so the simulated rush through the fatality site is about an hour slow.

Unsteady Particle Traces in ERICA IOP4

Particle traces in ERICA IOP4, a severe Atlantic storm in 1989, illustrated using VAPOR unsteady flow.  The particles are colored according to the value of ETH (equivalent potential temperature).

SC13 Visualization Showcase submission

A team of researchers from the National Center for Atmospheric Research (NCAR), the National Center for Supercomputing Applications (NCSA), and Cray Inc., simulated the evolution of hurricane Sandy as it approached and made landfall, with catastrophic impacts over the northeastern United States.  The simulation was performed on the Blue Waters Cray supercomputer at NCSA, using the NCAR/WRF-ARW regional prediction system. The simulation was comprised of previously unsurpassed ~4 billion computation grid points, with a horizontal grid resolution of 500 m with 150 vertical levels. The simulation is part of NCAR's research agenda to advance knowledge and predictive skill of high-impact weather hazards and the transition of research-to-operational services. It illustrates the importance and benefits of advances in super-computer capacity and visualization systems for Earth-system research and weather prediction.   The NCAR VAPOR visualization system was used to create animations, tracking the storm’s progress, enabling analysis of flow trajectories and associated dynamical processes during the life cycle of Hurricane Sandy.  Technical details of the simulation are described in the SC2013 article "Petascale WRF Simulation of Hurricane Sandy: Deployment of NCSA's Cray XE6 Blue Waters".

Alan Norton, Mel Shapiro, Tom Galarneau, Perry Domingo, NCAR
Peter Johnsen, Cray Inc.
Mark Straka, NCSA, University of Illinois at Urbana-Champaign

[Image] Cloud top temperature from Hurricane Sandy

This is a single image of cloud top temperature from an ultra-high resolution simulation of Hurricane Sandy.

Coupled weather-fire behavior simulation

Coupled weather-fire behavior simulation of the Esperanza wildfire near Cabazon, California during October 2006. The visualization shows the spread of the fire, the intensity of the fire, areas burned out by the fire, and the smoke released by the fire as it spread driven by Santa Ana winds in the San Jacinto Mountains in dry chaparral fuels. cdfdata.fire.ca.gov/incidents/incidents_details_info?incident_id=161 . Janice Coen, NCAR and Phil Riggan, USDA Forest Service

Horizontal hurricane vortices

Vortices in the eye wall of a WRF hurricane simulation, illustrated by a combined streamline and image-based flow visualization. Yongsheng Chen, MMM/NCAR