Search for primordial non-Gaussianity in the quasars of SDSS-III BOSS DR9

We analyse the clustering of 22,361 quasars between redshift 2.2 < z < 2.9 observed with the Sloan Digital Sky Survey (SDSS)-III Baryon Oscillation Spectroscopic Survey (BOSS), which are included in the ninth data release (DR9). We fit the clustering results with a {\Lambda}CDM model to calculate the linear bias of the quasar sample, b=3.74\pm 0.12. The measured value of bias is consistent with the findings of White et al. (2012), where they analyse almost the same quasar sample, although only in the range, s<40 Mpc/h. At large scales we observe an excess or plateau in the clustering correlation function. By fitting a model that incorporates a scale dependent additional term in the bias introduced by primordial non-Gaussianity of the local type, we calculate the amplitude of the deviation from the Gaussian initial conditions as 70<f_NL< 190 at the 95% confidence level. We correct the sample from systematics according to the methods of Ross et al. (2011); Ho et al. (2012) and Ross et al. (2011,2012b), with the f_NL measurements after the application of the two methods being consistent with each other. Finally we use cross-correlations across redshift slices to test the corrected sample for any remaining unknown sources of systematics, but the results give no indication of any such further errors. We consider as our final results on non-Gaussianity of the local type, 46<f_NL< 158 at 95% confidence, after correcting the sample with the weights method of Ross et al. (2011, 2012b). These results are consistent with previous tight constraints on non-Gaussianity from other LSS surveys, but are in tension with the latest results from the CMB.