SEARCH-ORIENTED TECHNIQUE OF THE NUMERICAL FORECASTING OF THE RARE-METAL PROXIMAL (NEAR DRIFT) PLACERS: EVIDENCE FROM THE LOVOZERO PLACER DISTRICT

Abstract

Rare metals (including rare earth metals) are significant raw materials which determine technical and scientific level of development of the industry; first of all – innovative technologies. An important source of rare metals can be a group of unique in genesis and mineral composition of loparite placers associated with the Lovozero massif of nepheline syenites (Cola peninsula). In the present work Within the Northern periphery of the Lovozero massif, which includes the Revda placer deposit, the authors have identified the main factors controlling the formation of proximal (near drift) rare-metal loparite placers. They have performed the formalization (numerical evaluation) of this factors and developed multiplicative indicators that reflect the placer potential of territories. To assess the effectiveness of the proposed technique based on the developed multiplicative coefficients, independent forecasts of placer metal content for the western and eastern placer-bearing areas of the Revda group of placers were calculated. For additional verification of the technique, the paper compares the obtained results with similar forecasts obtained using the methods of Linear Regression and Random forest of Regression trees. The developed method allows automating the process of forecasting of placer parameters, which is especially important for assessing deep-seated placers that are difficult to search and exploration and are not shown on the surface.