“…This therefore provides an underdetermined problem, with fewer constraints than potential dislocation types for which we want to determine densities. There are many routes discussed in the literature to treat this problem (Britton et al, 2009;Gardner et al, 2010;Kysar et al, 2010;Kysar et al, 2007;Sun et al, 2000) (Arsenlis and Parks, 1999;Field et al, 2010;Hardin et al, 2011;Hardin et al, 2013;Ruggles and Fullwood, 2013) (Demir et al, 2009), with varying degrees of success and merit. We choose to select a solution that physically supports the curvature measured and also minimises the total line energy, accounting for differences between screw and edges using isotropic elastic energies, of the dislocations used (this a L 1 minimisation scheme).…”