b'Minerals geophysics available. However, a problem emerged. Very strong IP anomalies associated with high resistivities sometimes appeared in the inversions where only low order IP anomalism was present in the field results. These rogue inversion anomalies typically occurred where the geo-electric environment was one of shallow low resistivities overlying strongly resistive basement. While the anomalies were present in both pole-dipole and dipole-pole MIMDAS array subset inversions, they were most prominent when results from both subsets were co-inverted. Drill-testing of some of these features failed to locate significant sulphides, confirming that the strong IP anomalism was an unwanted effect from the inversion process itself.Conventional IP-resistivity inversion routines invert resistivity first, then reference the inverted resistivity model in the subsequent IP inversion. Could the logarithmic nature of resistivity distribution be a factor in rogue IP anomaly generation? The Zonge software I was using allowed intervention between the two inversion stages, so I simply replaced the resistivity inversion values with log resistivities and restarted the process to invert the IP values. The results were nothing short of dramatic. Our rogue IP anomalies disappeared from the inversions, leaving much more credible patterns (see Figure 2). Subsequent LIP inversion tests with valid IP inversion anomalies resulted in a replication of patterns and only a slight reduction in anomaly intensities. Where any inversion resistivity values were 1 ohm-metre or less, the entire set was multiplied by an appropriate constant (say 10) to ensure that all log resistivity values were positive prior to LIP inversion. Because the resistivity pattern itself remained unchanged, this did not affect the LIP inversion outcome.The log resistivity intervention procedure became a permanent addition to our inversion processing routine, giving usFigure 2.MIMDAS IP-Resistivity Pseudosections (for brevity, only dipole-pole shown) and Inversions a) resistivity, IP and LIP inversion products.DP Resistivity Pseudosection, b) PD-DP Resistivity inversion, c) DP IP Pseudosection, d) PD-DP IP Inversion, e) LRRIP (log resistivity referenced IP)PD-DP LIP Inversion.would have been a more informative but cumbersome acronym, and XIPexperience showed that applying the LIPNo doubt theres a mathematical (Xstrata IP) would have been more zippy,process where IP anomalism was directlyexplanation for all this, but I dont but LIP it is. Using a homogeneousassociated with strong low resistivityhave it. However, next time you have half-space as the resistivity referencezones resulted in an enhancement of IPan inversion with an IP anomaly of for the IP inversion (we could call it HIPanomalism in the LIP inversion, arguablyquestionable validity associated with inversion) produced similar results to LIP,going some way to compensating for thelocally higher resistivities, give LIP but without the nuances attributablediminution of IP effects with decreasing(or even HIP) a go. The results may to resistivity variations. Subsequentresistivity. surprise you.37 PREVIEW JUNE 2022'