b'FeatureDiscovery and geophysics of the Khamsin depositovercorrection causes the gravity to mirror it. For large areasPlan views of the evolved model are shown Figures 5.1 to 5.3, comprising many surface rock types, the industry-standardwhere magenta-coloured bodies are dense and non-magnetic, density of 2.67 g/cc is widely accepted, but for small areasand greenish bodies are magnetic. Long, narrow, pale pink a different value may be optimum, since an inappropriatebodies have negative density contrasts thus plotting outside correction density can yield false gravity anomalies. The gravitythe phase diagram (Hanneson 2003), and of these, unlabelled data from Khamsin was reduced using several correctionbodies 119, 115 and 111 follow the dotted line in Figure 3.2 densities in the range from 1.9 to 2.67 g/cc (not shown) and theto simulate the linear gravity low. Figure 6 is a cross-section method of Freund (1960) was used to show that the density thatalong line 6547800N, where the solid blue profile shows the minimises the gravity vs. elevation covariance for the data setsmagnetic data, and the magnetic model response (dotted blue) shown in Figures 3.1 and 3.2 is 2.27g/cc.is barely discernible because of the accuracy of the simulation. Likewise, the gravity response of the model (dotted red) is an All Bouguer correction densities showed similar robust gravityaccurate simulation of the gravity data (solid red). As a numerical anomalies (including the double peak) in the central part ofexperiment, the single-peak gravity response (also dotted red in the area, with the similarities presumably due to the blandFigure 6) arises when the Body 115 simulating the palaeochannel topographic gradient seen in that part of the image. However,is removed from the model by giving it a zero-density contrast. there were indications of minor local covariances that were both positive and negative suggesting that no single correctionThe model benefited from knowledge that Hole KH002 density would be optimum throughout the entire area. Oneintersected basement at 457 m, but the funnel shape used for final test was carried out using the method of Rimbert et al.the dense non-magnetic top is not a strict requirement of the (1987) wherein the covariance at a point being corrected isdata. Other shapes could simulate the data as well; however, the calculated using only points within a selected radius, and theshape used was inspired by a similar shape in a notional cross Bouguer correction is then applied at the given point usingsection for Olympic Dam (Haynes et al. 1995, p298). Often, when the local covariance minimizing density. The result is shown ina model is constrained only by the ambiguous geophysical Figure 3.1, which yields slightly smoother contours than imagesdata, a small dose of geological credibility can make a model using a single density, covariances are in the -0.02 to +0.02geologically more believable and may also reduce the perceived range, and covariance minimising densities generally increaserisk that the model could be invalid. Furthermore, splaying from west to east. However, it must be noted that coincidentalthe top out along the known unconformity depth allowed correlations can occur between topographic features andincorporation of the fact that minor sulphides were intersected subsurface density inhomogeneities that cannot be predicted;at the unconformity in hole KH002. See Figures5.1 and 6.consequently, none of covariance minimisation methods areA NNE linear magnetic high (body string 39 to 50) can be immune from the generation of false anomalies.seen in Figure4.2, however the cause of this feature which is Density and magnetic susceptibility modelwell-defined in the east-west aeromagnetic flight lines seems to be west of and much deeper (800 m) than the interpreted Forward modelling with the almost arbitrary body shapespalaeochannel although both could be following an earlier fault permitted by the method of Talwani (1960, 1961) was usedstructure that in some way controls the magnetic lineament. to develop a density and magnetic susceptibility modelThe density/susceptibility model contains over a hundred whose calculated responses are fair simulations of the data.bodies; most are of no perceived economic significance but The MagGravJ method (Hanneson 2003), which, amongwere included to improve the data simulation. Many bodies other things, permits calculation of the gravity response ofplot on the Magnetite Line (left margin of coloured area) of the non-magnetic material separately from the gravity responsephase/scatter diagram in Figure 7 and, in fact, these bodies of magnetic units, was used to assess the model bodies forwere chosen to plot on the Magnetite Line but only after their concentrations of three gross categories; namely, magnetite,responses were seen to simulate, and therefore to be permitted hematite+sulphides, and a barren lithology unit, in this case,by, the local magnetic and gravity data. They are interpreted felsic rock. This can be advantageous in the search for IOCGs andgeologically to represent minor accumulations of magnetite amounts to a joint interpretation of the two data sets.within a (felsic) matrix that has the same density as the country The dotted line in Figure 3.2 traces a narrow residual gravityrocks, except to say that deep body 28 in Figure 5.3 is part of low running across the image and beyond the limits of thethis group and is interpreted to represent felsic rocks with about area. It invites speculation that a palaeochannel with relatively5 percent magnetite. It suggests deeper more reduced rocks low-density infill passes over deeper dense rocks, which, inexpected for the standard IOCG model. turn, overlie a deeper magnetic source, and that the supposedOther bodies like the greenish string (bodies 50 to 39) in palaeochannel could be the cause of the double gravity peak.Figure5.3 were chosen to plot on the dashed green Gabbro With this as an overall strategy, a new model was begun usingLine, and while simulating the data, they have the properties first a shallow, low density flattened rope-like body to simulateexpected for mafic rock with minor magnetite. Magenta the supposed palaeochannel at the north and south ends andcoloured bodies plot on the zero-magnetite baseline to the passing through the central gravity high. It was then found thatright of the Gabbro Line and are judged to represent non-a single dense contiguous ovoid feature immediately overlyingmagnetic rocks that are denser than barren mafic rock; they a magnetic vertical pipe-like body (the classic IOCG scenario)are depicted in Figure 7 and Table 1 as having the properties could simulate the gravity and magnetic data. of felsic rock with some tens of percent of material like While the earlier models that used two bodies to simulate thehematite+sulphides and no magnetite.double gravity peak suggested a disappointing prognosis forBody 115, which simulates the interpreted palaeochannel, is poorly IOCG mineralisation, the single dense ovoid gave reason to testconstrained by the data. Assuming unconsolidated in-fill, it was between the peakswhich yielded significant grades. assigned a density contrast of 1.952.65 = -0.70 g/cc (compared 44 PREVIEWFEBRUARY 2024'