Comparison of the results of estimating systematics parameter noise using multiple realizations of mock catalogues versus a block bootstrap approach on a single mock catalogue realization. The figure shows the variance in the estimated parameters (scaled by the area) as a function of the sky area coverage. The different line colours represent different systematic classes. The triangle points show the variance from using 100 realizations of mock catalogues, while the circular points show the variance estimated using the block bootstrap on a single realization. The error bars show the error on the mean, and the lines (varying styles) track the results from mocks. We see that the bootstrap approach can effectively estimate the parameter noise for higher sky area coverage and low noise parameters, while it fails for small surveys with high parameter noise.