The impact of strong gravitational lensing on observed Lyman-break galaxy numbers at 4 ≤ z ≤ 8 in the GOODS and the XDF blank fields

Detections of Lyman-break galaxies (LBGs) at high-redshift are affected by gravitational lensing induced by foreground deflectors not only in galaxy clusters, but also in blank fields. We quantify the impact of strong magnification in the samples of B₄₃₅, V₆₀₆, i₇₇₅ and z₈₅₀ & Y₁₀₅ dropouts (4 ≤ z ≤ 8) observed in the eXtreme Deep Field (XDF) and the Cosmic Assembly Near-infrared Deep Extragalactic Legacy Survey (CANDELS) fields by investigating the proximity of dropouts to foreground objects. We find that ~6 per cent of bright z ~ 7 LBGs (mH₁₆₀ < 26) have been strongly lensed (μ > 2) by foreground objects. This fraction decreases from ~3.5 per cent at z ~ 6 to ~1.5 per cent at z ~ 4. Since the observed fraction of strongly lensed LBGs is a function of the shape of the luminosity function (LF), it can be used to derive Schechter parameters, α and M*, independently from galaxy number counts. Our magnification bias analysis yields Schechter-function parameters in close agreement with those determined from galaxy counts albeit with larger uncertainties. Extrapolation of our analysis to z ≥ 8 suggests that surveys with JWST, WFIRST and Euclid should find excess LBGs at the bright end, over an intrinsic exponential cutoff. Finally, we highlight how the magnification bias measurement near the XDF detection limit can be used to probe the population of galaxies beyond this limit. Preliminary results suggest that the magnification bias at M_UV ~ -18 is weaker than expected if α ≤ -1.7 extends well below the current detection limits. This could imply a flattening of the LF at M_UV ≥ -16.5. However, selection effects and completeness estimates are difficult to quantify precisely. Thus, we do not rule out a steep LF extending to M_UV ≥ -15.