High-Redshift Clustering in the HDF

This paper addresses the problem of detecting high-redshift clustering in deep photometric surveys. We have used photometric redshifts to select different samples of galaxies in the HDF, in order to study their clustering properties within the 2.5<z<4.5 domain, in redshift bins of 0.5, using different techniques. A strong clustering signal is found for redshift bins including z~3.4 and leading to a present correlation length of r_0 = 4.1 +- 0.8 h^-1 Mpc (q_0=0.1) assuming linear evolution. An excess appears in the correlation function of chip 2 with respect to the fit. This excess could be associated to the main structure detected in this field, which contains 20% of the objects identified at 3.4<z<3.9. Its dimensions are 3 h^-1 Mpc x 0.5 h^-1 Mpc. The galaxies at 3.4<z<3.9 exhibit a SFR of a few solar masses per year, but their comoving density is a factor of ~50 higher than the population of star-forming galaxies reported by Steidel et al. (1996b). The resulting star formation rate density is at least 1.1 10^-2 h Msun/yr/Mpc^3 (1.8 10^-2 h Msun/yr/Mpc^3) with q_0=0.1(0.5), slightly higher than the results by Madau et al. (1996) at 2.5<z<3.5, and then incompatible with a global decrease of the star formation in this redshift domain. These results on star formation and clustering are consistent with a hierarchical scenario for galaxy formation.