SCREAM-relevant events
3:30pm-4:15pm: 45 min breakout session
Afternoon poster session
8:30-10:30am: 20 min SCREAM overview during that time
3:15-4:15pm: 1 hr breakout session
8:30-10:30am: will include 2x 15 min SCREAM plenary talks
11-12:40: speed dating… SCREAM meets with Shaocheng’s NGD and infrastructure
Breakout #1: SCREAMv0 (Nov 19 3:30-4:15)
Short intro by Peter Caldwell about what SCREAM is and how we’ve organized sessions (3 min)
Status of SCREAMv0 (combination of Ben Hillman , Chris Terai , Noel Keen , and maybe Paul Ullrich ) (20 min)
current capabilities and remaining issues
throughput (in SDPD), cost (in MPPs/SD), and outlook for improvement
assessment of model climate
Boutique configurations (combination of Peter Bogenschutz , Hassan Beydoun , Wuyin Lin , and Qi Tang ?) (15 min)
IOP-SCREAM
RCE-SCREAM
forecasts
RRM?
Simulation plans (Peter Caldwell , 1 slide) (2 min)
Breakout #2: SCREAMv1 (Nov 20th, 3:15-4:15pm)
NH Dynamics efforts (Oksana Guba , Andrew Steyer (Unlicensed) , Luca Bertagna , Mark Taylor ) 15 min?
Getting the right answer: testing, design docs, etc (James Foucar , Andrew Bradley , Kyle Pressel (Unlicensed) , etc) (20 min, or more if needed)
What does our C++/Kokkos code looks like? (James Foucar et al, <5 min?)
Atmosphere Driver (Aaron Donahue and Luca Bertagna ) 20 min?
Plenaries (Nov 21, 8:30-10:30am)
We have 2x 15 min slots. It seems like the SCREAMv0 and “Getting the Right Answer” talks from above would be ideal as plenaries, but I also like how they frame their respective breakout sessions and am unclear what we would put in those breakouts instead.