Minutes by Tom Dent, Cardiff

Announcement 2

Michele's report on ILIAS. He will create a slide explaining the opportunities.

Announcement 3

18 months will soon be up, 1st week of December submission of report, 30th October deliver draft to Michele, There will be a discussion at Erice on how to go about doing this. Volunteers needed in the preparation of report. Please contact Sathya if you are able to help.

Announcement 4

Remote participation in Erice session requested. Michele is exploring if this will be possible.

ET sensitivity 1

• In xylophone case Gr-Gr and suspension thermal noise are difficult to reconcile with LF sensitivity
• Need a fit to xylophone curve: quantum noise functional form + HF + extra LF noise -- add them together (Lueck/Hild/Freise will take care of this)
• NB: given curves are strain for a single (L-shaped) 10km ifo - therefore need to consider the correct geometry for any given source
• Put analytical fit in public area? - or in 'Science team' area with easy password access? NB curves and data files are already public

ET sensitivity 2: Tradeoffs

Report session I.1 Sathya, science summary

• BBH mergers: can Earth rotation be used for angular resolution?
• Science potential for BBH vs BNS?
• ... other open problems (slide 11)
• Vision Document: should update results & plots to ET-B curve

Report session I.2 Melatos, NS signals

• r-modes & mountains, glitches ...
• Sathya: What is most pressing question? - Accreting NS, mountains; glitch mechanism

Report session I.3: Chris, Star formation rates

• Work on disentangling / correlating NS rate from total SFR
• Question: Can you use data beyond z=1, if efficiency is accurately known?

Report session I.4: Amaro-Seoane, IMBH binaries

• ET should see merger of IMBH (if they exist) in range 100-10000 M_sun

Report session II.2 Regimbau, BNS mock data

• Lower frequency cutoff is very important for duration and confusion bg
• Analyze with standard LIGO inspiral code?
• - Methods for extracting individual sources from confusion?

Report session II.2 Bosi, Computational infrastructure

• 5 TFLOPS on a single core by 2020.
• 2 hours of data on a GPU today would take 18 days to analyze 2 million templates required in ET
• By 2020 one expects a gain factor of 120 in GPU computational power, or under 4 hours