The Mineral Population Boundary Problem

INTRODUCTION Great emphasis has been given throughout this text to the necessity of maintaining the integrity of mineralogical populations when operating upon sample data drawn from mixed populations. Mineral deposits rarely are completely homogeneous and often consist of many subsets separated by geological and mineralogical boundaries. Sometimes different mineral populations are difficult to distinguish, and may be inadvertently lumped together-often without serious consequences, but more often with detrimental effects upon analytical results. The geologist must accept the responsibility for recognizing and recording mineral boundaries that may include: changes in mineral suites, lithological horizons, host rock types, faulted and offset segments, zones of enrichment or depletion, and several other geological and mineralogical changes that effect mineralogical structuring and distribution within a deposit. When employing conventional methods for assigning estimated values to irregularly shaped mineral inventory blocks, it was only necessary for the geologist or engineer to con- fine the block limits to the known geologic boundaries. Thus if a fault interrupted the mineral zone, the trace of the fault could be used as a block boundary. Intelligent decisions could always be made as to which blocks to include in each mineral population. However, with the advent of the computer, which per- mits the computation of mineral inventory blocks using better mathematical estimation techniques, there now enters the "boundary problem." Unless the computer can be programmed to recognize population differences and geologic boundaries, greater estimation errors will be found in computer-generated mineral reserves rather than in less sophisticated conventional inventories. Much disillusionment has been expressed over the years by experienced mineral appraisers as a result of their study of computer-generated inventories that failed to account adequately for geologic vagaries. Many, thus disillusioned, have concluded that computer techniques are not satisfactory, or have only limited applications to simple deposits. As with all problems in the mineral industry,