MR. NIKHIL MUKUND

Improving the low frequency (below 30 Hz) sensitivity will be a crucial upgrade required for Enhanced Advanced LIGO and Third Generation GW Detectors. Gravity Gradient a.k.a Newtonian Noise (NN) caused due to fluctuations in local gravitational field, is one such noise that decides the Advanced LIGO noise floor from 10 - 30 Hz. In this talk, I will discuss about how we have developed realistic estimates of this noise using seismic spectra obtained from seismometers configured around the LIGO detectors. These estimates assume surface Rayleigh waves as the major contributor to seismic spectra which is in agreement with the observational seismic data from LIGO Hanford array experiment (2012). Further we show how to optimally place seismic sensors around LIGO test mass so as to achieve coherent noise subtraction from the GW strain channel. Array optimization is done in both the idealistic case of isotropic seismic fields and also for the measured Hanford array. Obtaining global array solution becomes difficult with increased number of sensors due to the numerous local minima involved and computational complexity associated with multi-dimensional parameter space. We tackle this issue using higher dimensional sampling algorithms such as Simulated Annealing and Nested Sampling. I will briefly describe how our parallelized cascaded implementation of Simulated Annealing successfully avoids getting trapped in local minima and generates optimal solution in finite time leading to maximal Newtonian Noise suppression.