Importance of including small body spin effects in the modelling of intermediate mass-ratio inspirals. II Accurate parameter extraction of strong sources using higher-order spin effects

We improve the numerical kludge waveform model introduced in [1] in two ways. We extend the equations of motion for spinning black hole binaries derived by Saijo et al. [2] using spin-orbit and spin-spin couplings taken from perturbative and post-Newtonian (PN) calculations at the highest order available. We also include first-order conservative self-force corrections for spin-orbit and spin-spin couplings, which are derived by comparison to PN results. We generate the inspiral evolution using fluxes that include the most recent calculations of small body spin corrections, spin-spin and spin-orbit couplings and higher-order fits to solutions of the Teukolsky equation. Using a simplified version of this model in [1], we found that small body spin effects could be measured through gravitational wave observations from intermediate-mass ratio inspirals (IMRIs) with mass ratio eta ~ 0.001, when both binary components are rapidly rotating. In this paper we study in detail how the spin of the small/big body affects parameter measurement using a variety of mass and spin combinations for typical IMRIs sources. We find that for IMRI events of a moderately rotating intermediate mass black hole (IMBH) of ten thousand solar masses, and a rapidly rotating central supermassive black hole (SMBH) of one million solar masses, gravitational wave observations made with LISA at a fixed signal-to-noise ratio (SNR) of 1000 will be able to determine the inspiralling IMBH mass, the central SMBH mass, the SMBH spin magnitude, and the IMBH spin magnitude to within fractional errors of ~0.001, 0.001, 0.0001, and 9%, respectively. LISA can also determine the location of the source in the sky and the SMBH spin orientation to within ~0.0001 steradians. We show that by including conservative corrections up to 2.5PN order, systematic errors no longer dominate over statistical errors for IMRIs with typical SNR ~1000.