Skip to content

Meeting Rutgers and PSU 25 Jan 2017

Jump to bottom
Matteo Turilli edited this page Mar 8, 2017 · 1 revision

Meeting notes

  • PSU will come up with a mini app describing the workflow
  • Get access to Yellowstone and Cheyenne
  • Create specification document for the application and share [VB]

Details on the AnEn application

  • Current estimate provides a multi-variate distribution across time and space
  • Variables - time, pressure, wind speed, wind direction (100s of more variables)
  • This data along with the WARF historical data is used to estimate the distribution for the future
  • Each of the "nodes" needs to communicate with a dynamically determined number of neighboring nodes
  • This communication can possibly be a significant overhead.
  • There exists a feedback/iteration loop in the workflow. This is currently WIP. PSU is working on writing/reading the outputs in netCDF format.
  • Automate the workflow
  • No. of simulations? Approximate simulation times?

Clone this wiki locally