Amaldi session on LISA

Scott Hughes - LISA science

- LISA accuracy comparable to moon, but 50x Hubble deep field

---

Dan Holz - standard sirens and cosmology

- Cosmology: all about getting distances
- SMBH: position determination to tens of arcsecs; luminosity distance to 1%
- But need a redshift from some EM counterpart; ~ 10 galaxies/arcmin^2 for host galaxies
	- Errorbox should narrow down with additional inspiral information
	- Also, use rough knowledge of cosmology to narrow down redshift; use galaxies that are morphologically promising
	- Or: calculate distance to all possible hosts, and demand concordance across multiple sources
	- Or: use statistical knowledge of source population
	- But best hope: look for something to go bang
- Having counterparts improves determination of other parameters from GW
	- COULD USE IN A CHALLENGE: GIVE POSITION, ORACLE YIELDS HOST GALAXY WITH EXACT POSITION
- Lensing: affects luminosity distances significantly. How to undo it?
	- Direct lens reconstruction; identify luminous objects along source direction
		- But matter is mostly dark
		- Build corrective lenses; but even good ones are smooth, and do a poor job of correcting for amplification
		- But lensing effects average away when added; 100 LISA sources compare to 3000 SNAP SNe for w and Omega_m determination (with different ellipse orientations)

---

Ulrich Johann - European LISA Mission industrial formulation study

---

Tuck Stebbins - U.S. LISA formulation and technology development

---

Chris Van Den Broeck - Higher harmonics help LISA

- Increase the mass reach
- Greatly improve parameter estimation

---

Stefano Vitale - LISA Pathfinder status report

- LIGO requires more stringent displacement-error control than LISA; but LISA requires more stringent acceleration errors
- test-mass noise: surface (electrostatics, EM, thermal), volume (magnetics, gravitational), reference frame (spacecraft coupling, crosstalk)
	- surface noises can be investigated on the ground; inertial sensor defines immediate environment of test mass

---

Bernard Schutz - SMBH position determination with LISA

- standard sirens require EM counterpart determination
- early estimations used simple 2-arm Michelson
	- now have TDI, spin precession
	- and PN amplitude corrections: higher SNR, bigger modulation and phase delay
- usually do Fisher-matrix approach, accurate for high SNR
- examine alternative null-stream locator
	- introduced by Gursel and Tinto, studied by Schutz and Wen
	- new null-stream locator
		- define signal basis on 3 TDI streams
		- for any vector, components ep, ec, eperp (perpendicular component = null stream) 
		- if polarization is known, another null stream exists
	- information seems different from Fisher-matrix
- study localization with higher harmonics
	- don't get same improvements as other groups

---

Alberto Vecchio - two MLDC2 efforts

- determine number and parameters of stellar-mass compact objects
- determine whether or not the stochastic signal contains a contribution due to GWs

- new: reversible-jump Markov-Chain Monte-Carlo; Bayesian analysis for stochastic signals

- compute posterior density for number of sources, source parameters, and noise
- 18 sources in window 2 vs. 3 sources (but don't need to recover all of them)
- problems determining (non)recovered sources

- do Bayesian stochastic-signal determination 
	- have two channels (one null), estimate also noise variances and correlation
	- very nice!
	- use information about relative levels of noise