Author(s): , ,
Institution(s): 1. University of Science and Technology of China
We utilize CANDELS/3D-HST photometric data to analyse the spatial clustering of massive galaxies at redshift 110 Msolar according to their high surface density and divide them into quiescent (QGs) and starforming (SFGs) subsamples with UVJ color diagram. We employ a similar technique used in Hickox et al. (2012) to measure the projected two-point crosscorrelation function and derive the linear bias and characteristic dark matter halo masses for the compact galaxies. We estimate an autocorrelation length of r0 = 8.1±1.3; 8.0±1.6 h−1 Mpc for cSFGs at z~2.5 and cQGs at z~1.5, respectively. The clustering of cSFGs at z ∼ 2.5 is consistent with measurements for submillimeter galaxies (SMGs) ,optically selected quasars (QSOs) at similar redshift range and cQGs at z~1.5, supporting evolutionary scenarios combining SMGs, cSFGs, QSOs and cQGs together. Our results support the recent arguments that SMGs or cSFGs could be the progenitors of cQGs at lower redshift. And the cQGs at z~1.5 could be the precursors of local massive early type galaxies (ETGs). As a comparison, we also measure the autocorrelation length of r0 = 5.3±1.7 for diffuse SFGs at z~2.5. The result shows that the diffuse SFGs have a different evolutional path and disfavors the model that diffuse SFGs evolve into cSFGs according to structural transformation. The Lyman break galaxies (LBGs) at z>3 could be the progenitors of diffuse SFGs at z~2.5. And diffuse SFGs will evolve into local blue galaxies instead of ETGs. We propose that the evolution of cSFGs may relate to SMGs instead of diffuse SFGs.