b'Minerals geophysics Figure 3.Use of 2.5D AEM inversion as a quality control checking tool. In South Australia, the Waddikee Tempest survey shows what portion of the measured signal is useable and what is below the noise floor.structure buried below the surface, with aInversion further quality control issues which were quite ambiguous TMI signal when sampledThe 2.5D inversion technique has a placenever anticipated at the outset emerge. along four lines that cross it. If a tensorto play in quality control for AEM surveys.Each of the geophysical signal types seem magnetic gradient signal is acquired, or inThe method is very exact and reports allto have their own unique issues, once this this case calculated (Holstein et. al. 2009),difficulties with the delivered data fromscenario is discussed.most of the ambiguity disappears and thea contractor with ease. Figure 3 shows an challenge then is to push the resolutionexample where the quality of the surveyGravityto the upmost. A sensor may have updata is being tested. In particular, theOlder gravity stations set out on up to a to 6 degrees of freedom in its trajectoryinversion requires correct calibration of X &grid of 11 km have issues with XY location, through space. The three rotational statesZ channels, proper compensating for birdand of more concern the height as a have not been routinely measured withmotion, validation of the noise model andbarometric pressure method was used the required accuracy nor corrected for.a lack of late time correlated noise. Whenhistorically. This means the elevation This has now become a requirementthe data are not processed correctly, any orestimate in older datasets has much for these high rate, higher dimensionalall these factors are easily detected. larger errors that typically cannot be acquired signals. Of course, the presence of noise and non-ideal flying directions alsoIf 1D inversion, or CDI production is allfixed, compared to more modern survey can be investigated once this framework isthat is attempted by the team checkingobservations. This means merging older available. the delivered AEM data, there is muchdata of this nature with a modern airborne less quality control being applied for agravity survey is almost a pointless Turning to AEM data sets, griddingcoherent and properly calibrated signal.exercise, except there may be no choice.also presents some quality controlThe characteristic pants-leg artefacts challenges. Typically, the magneticfrom 1D often are interpreted as anRadiometricsB-field component decay curves of theanticline by a geologist, when in fact they secondary response are the principaloften reflect a steeply dipping conductorIn earlier times, calibration of the crystals delivered data from an airborne survey.or an off-end effect associated with a fault. and the processing coefficients used As it is easy to do, grids of these earlymay either be lost or poorly executed. At time and late time values can beLonger Term Factors times flight lines need to be re-occupied generated. These grids are not directlywith a view to re-establishing what interpretable as a geological response,As geophysical surveys have been acquiredcoefficients might have been used to but rather an indication of localover more than 50 years, there areproduce the final multi-channel data coherence between soundings. It is thegenerations of workers and instrumentsand the standard four channel products. cross-sections of the log conductivitythat make for a non-homogeneous patchUnless this process is followed there is after an inversion that has morework of spatial coverage, which in turnalmost no chance of harmonizing the geological significance, and that is whathas an impact on regional dataset quality.old survey data with a next generation. the geologist needs. Regional surveys abut each other andFigure 4 shows a newly developed APRIL 2020 PREVIEW 38'