b'Coping with ambiguityFeaturewould benefit the search for iron-oxide copper-gold (IOCG)magnetite) has about the same density as barren gabbro (with deposits. When K. Ehrig (pers com., c.1996) commented on thethe same amount of magnetite).similarity between certain gravity contours at Olympic Dam and the outline of the deposit as she then understood it, weEratosthenes, enjoying the benefit of independent evidence, launched into an attempt to refine an earlier model reportednevertheless profited from a lucky guess that the suns rays by Esdale, etal., (1987), using more arbitrarily shaped bodieswere parallel. Like him, I too had made a lucky guess (influenced and constrained by more detailed data than were availableby geological opinions) when I assumed that the lithology of to the earlier workers. This work predated the final stage ofthe Dubawnt Lake feature was mafic. A less, or more, dense development of what came to be called the MagGravJ Methodlithology would not have yielded the depth-to-top and percent (Hanneson, 2003), but it was nevertheless successful in directingmagnetite estimates from Figure 6 deemed to be correct by my drilling that led to the discovery of the Far North Mine Block atgeological colleagues after drill testing.Olympic Dam.ConclusionMy growing involvement in the search for IOCGs prompted a re-think of the Dubawnt Lake problem. This led to dividing theSo, provided a given anomaly is expressed in different second component (the barren mafic rock) into two; namely,independent data sets, acquiring data using different methods barren felsic rock and a dense non-magnetic component likeand applying a joint interpretation of both can significantly a mixture of hematite and sulphides. (Hematite and sulphidesreduce the ambiguity inherent in any given dataset considered must be lumped together because they are similarly dense andin isolation. But as we comb through the remaining multiplicity are non-magnetic and therefore have similar effects on gravityof allowed physical models, it is impossible to overstate the and magnetic responses). Thus Eq. 1, above, becomes importance of another form of independent information; namely, geological credibility. While allowing for the possibility d 1 p 1 + d 2 p 2 + d 3 p 3 = d m Eq. 3 of incorrect geological opinions, if we cannot create a physical model that both simulates the data and seems reasonable to and Eq. 2 becomes our geological colleagues, a manager is unlikely to support a proposal to expend funds drill testing it. In exploration, s 1 p 1 + s 2 p 2 + s 3 p 3 = s m Eq. 4 geophysics can be a useful servant, but beyond the control of geological credibility, it can be a cruel master.where now, subscripts 1, 2, and 3 refer to magnetite,My boyhood hero, Jonnie Bower, goalie of the Toronto Maple hematite + sulphide and the barren lithology, respectively.Leafs (sic) Hockey Club, when asked why he was so good at (I discuss revising term 1 in Eq. 4 as s 1 p 1 x in the 2003 report.)keeping the puck out of the net, said with some humility, Equations 3 and 4 permit calculating the density andYouve got to be lucky to be good, but then threw humility susceptibility of a defined mixture but are otherwise unhelpful.aside and added, but youve also got to be good to be lucky. The eureka moment that allowed me to move forward cameHe seemed to hint that with hard work and diligence, it may be with the realisation that if the physical properties of the threepossible to influence ones own luck, and, as I have suggested components span the full ranges for the rock-forming minerals,above, anyone who thinks luck is unimportant in geophysics is then the volumes of the three components would sum to theon a different wavelength from me.volume of the mixture, as inA final thought is provoked by one reviewer (CNB) who is p 1 +p 2 +p 3 =100Eq 5 uncomfortable with the notion of luck. Indeed the term is rather defeatist, admitting to a kind of randomness over which which, with Equations 3 and 4, form a 3x3 simultaneous set.we have no control as events unfold. When contemplating The density and susceptibility of the mixture (and the assumedthe vagaries and pitfalls in exploration, he prefers the more physical properties of the components) can then be used tooptimistic term risk management, part of which would amount calculate the proportions of the three components. This wasto doing everything possible to reduce ambiguity in the essentially the starting point in the original discussion of thegeophysical data through the joint interpretation of multiple MagGravJ method, which seems to reduce the ambiguity that isdata sets.inherent in data sets considered separately, because it permits calculating the gravity response of the dense non-magneticAcknowledgements and addendumcomponent separately from the gravity response of the dense magnetite. This is useful in the search for IOCGs which, in theI thank Chris G. Anderson, Terry V. Harvey and Callum N. Baxter classic scenario, exhibit a spatial separation between a shallowerfor suggesting many improvements to early versions of the dense non-magnetic component (hematite + sulphides) andmanuscript.deeper magnetic rocks. It also permits calculating how much magnetite and hematite + sulphides a felsic rock would needI owe a special debt to Ken C. Cross whose query prompted to have to exhibit a given density and susceptibility. Suchme to revisit Figure 6. Evidently in 1978, I thought 0.006 cgs estimates are more helpful to a geologist assessing the bodies(left end of upper horizontal axis) was a reasonable estimate of a model than simply stating the physical properties. for the susceptibility of barren gabbro. However, Id have done better by extrapolating the Mooney and Bleifuss (1953) power MagGravJ uses empirical rules to estimate percentages oflaw at least as far as 0.05% magnetite to get 0.0001 cgs which the three components from the density and susceptibility ofis consistent with 0.00008 cgs at the lower end of a range of a mixture but still requires assumptions about the physicalmeasurements (that dont indicate magnetite content) given properties of the barren lithology of model bodies and the hostby Telford, etal., in a later version (1980) of their book. The rocks. There remains another aspect of ambiguity in that felsiceffect of using a lower value for barren gabbro is to steepen rock with 14% hematite + sulphides (plus a given amount ofthe top end of the curve labeled MAGNETICS. If the nearby box DECEMBER 2022 PREVIEW 48'