1: \begin{abstract}

2: We study the choice of the regularisation parameter for linear ill-posed problems in the presence of 

3: data noise and operator perturbations, for which a bound on the operator error is known but 

4: the data noise-level is unknown.  We introduce a new family of semi-heuristic parameter choice rules 

5: that can be used in the stated  scenario. We prove convergence of the new rules and 

6: provide numerical experiments that indicate an improvement compared to standard 

7: heuristic rules. 

8: \end{abstract}

9: