The pessimistic case occurs when the success rate in estimating the redshifts of Halpha emitters is only 45% and the scatter in the estimated photometric redshifts of background sources is 0.05(1+z). We assume that the individual redshifts are Gaussian distributed.
The idealistic case occurs when the success rate is 100% and there is no scatter in the photometric redshifts.
The Poisson shot noise dominates the uncertainty budget on small separations where the number of source-lens pairs in a given angular separation is small. On large scales however, the uncertainy is dominated by the sample covariance.
As expected the signal-to-noise ratio drops significantly as we measure the lensing around Halpha emitters at higher redshifts. It is worth noting that even at the highest redshift bin the signal-to-noise remains high ~O(10).
Since the Halpha sample is a flux-limitted sample, we expect the distant objects to be intrinsically brighter and as a result have a higher linear bias. This is in line with our findings from constriants obtained by MCMC sampling:
We note that magnification in the source catalog barely adds any new object beyond mvis>24.5.
Lensing magnification adds a significant number of objects with f<2*10**-16 to the sample of Halpha emitters.
Mohammadjavad Vakili (Leiden Observatory) & Eric Jullo (Laboratoire d'Astrophysique de Marseille)