inherently non-unique, which precludes assigning true uncertainty measures to solutions. The merits and limitations of any depth estimation method can therefore only be evaluated empirically. We have used the“sweet-spot”method where locations for depth estimates are hand-selected, and that optimal data is inverted in a computationally intensive process. The resulting solutions are not proven correct, but are qualified as having been conscientiously generated as suitable for inclusion in regional and national data sets. An example depth estimate is shown in Figure 3. We provide the depth models in point format, as 3D models, in cross-section and map images, and in their individual ModelVision session files for any further development or testing. Where feasible we have also produced depth surfaces from an automated (“non-intelligent”) gridding, in some cases incorporating borehole basement intersection depths. These surfaces are far more interpretive and speculative than the individual solutions from which they are derived, and issues in assigning depth solutions vary considerably from block to block as discussed in the reports. Some of the GCAS blocks do provide a special opportunity for application of an automated magnetic source depth estimator due to the peculiar suitability of the Gairdner Dolerite dykes to analysis using AutoMag, a Naudy-based depth estimator. The dykes are well represented as thin, homogeneous magnetic sheets of large depth and strike extent, and their analysis can be applied through a profile-based vertical derivative filter to increase sensitivity to their tops and reduce influences of other sources. Most importantly, the analysis of each anomaly is individually performed in a moving window and the resulting solutions can be converted to bodies for testing by forward modelling (and subsequent inversion if required). An example AutoMag analysis of a section of profile is shown in Figure 4 and a set of model solutions is shown in Figure 5. Each profile/dyke intersection provides an anomaly opportunity, and this automated procedure generates many more solutions than can be evaluated by a manual method. Data, enhancements, source depth solutions and reports can be downloaded from the Gawler Craton Airborne Survey community information page: www. energymining.sa.gov.au/minerals/gcas with a“data available”link accessed by a mouse click in the selected map block. For blocks with data already released this opens further links to the data package and its report and, for those blocks with enhancements and depth solutions released, links to the appropriate digital package and report. The Gawler Craton Airborne Survey is one of the most comprehensive, high quality airborne surveys in South Australia’s history. The data that has been recovered will support the next generation of resource industry growth. Figure 4.  Example AutoMag depth section. Top: similarity coefficient (inverted), middle: vertical derivative (black from measured, red from model computed), bottom: TMI (black from measured, red from model computed, purple = regional). Figure 5.  Perspective view of depth models automatically generated from AutoMag depth solutions. Laszlo Katona,Geological Survey of South Australia Laz.Katona@sa.gov.au Matthew Hutchens,Geoscience Australia matthew.hutchens@ga.gov.au Clive Foss CSIRO clive.foss@csiro.au News Geophysics in the Surveys 26 PREVIEW JUNE 2019