Large-scale inhomogeneities of the intracluster medium: improving mass estimates using the observed azimuthal scatter

Achieving a robust determination of the gas density profile in the outskirts of clusters is a crucial step for measuring their baryonic content and using them as cosmological probes. The difficulty in obtaining this measurement lies not only in the low surface brightness of the intracluster medium (ICM), but also in the inhomogeneities of the gas associated with clumps, asymmetries and accretion patterns. Using a set of hydrodynamical simulations of 62 galaxy clusters and groups we study these kinds of inhomogeneities, focusing on the ones on large scales, which, unlike clumps, are difficult to identify. For this purpose we introduce the concept of the residual clumpiness, CR, which quantifies the large-scale inhomogeneity of the ICM. After showing that this quantity can be robustly defined for relaxed systems, we characterize how it varies with radius, and with the mass and dynamical state of the halo. Most importantly, we observe that it introduces an overestimate in the determination of the density profile from the X-ray emission, which translates into a systematic overestimate of 6 (12) per cent in the measurement of Mgas at R200 for our relaxed (perturbed) cluster sample. At the same time, the increase of CR with radius introduces a ˜2 per cent systematic underestimate in the measurement of the hydrostatic-equilibrium mass (Mhe), which adds to the previous one, generating a systematic overestimate of ˜8.5 per cent in fgas in our relaxed sample. Because the residual clumpiness of the ICM is not directly observable, we study its correlation with the azimuthal scatter in the X-ray surface brightness of the halo, a quantity that is well constrained by current measurements, and in the y-parameter profiles, which will be obtained in the forthcoming Sunyaev-Zeldovich (SZ) experiments. We find that their correlation is highly significant (rS = 0.6-0.7), allowing us to define the azimuthal scatter measured in the X-ray surface brightness profile and in the y-parameter as robust proxies of CR. After providing a function that connects the two quantities, we find that correcting the observed gas density profiles using the azimuthal scatter eliminates the bias in the measurement of Mgas for relaxed objects, which becomes 0 ± 2 per cent up to 2R200, and reduces it by a factor of 3 for perturbed ones. This method also allows us to eliminate the systematics on the measurements of Mhe and fgas, although a significant halo-to-halo scatter remains.