“Lag” — a geochemical sampling medium for arid regions

Abstract

Desert land surfaces are commonly characterized by a veneer or pavement of siliceous and/or ferruginous stony material. This material can be shown in most cases to be predominantly bedrock-derived, despite often severe modification by prolonged weathering, and can therefore be selectively sampled and analyzed as an indicator of bedrock geochemistry in geochemical surveys. Most pavements probably owe their existence to the interaction of several dispersion mechanisms. However, a common factor in all cases is the concentration of coarse particles at the surface as a result of the selective removal by erosion of fine dilutant material. Hence, the preferred name when used in the geochemical context is “lag”. In lag sampling, particles in the range 2.0–6.0 mm are screened on site from the unconsolidated surface material. Material in this size range has been found to be quite uniformly distributed over a wide range of arid region environments, including areas where residual soils are severely diluted by transported alluvial and aeolian materials. Data from exploration programmes for Au, Cu-Pb-Zn-Ba, and Ni allow comparison of results for lag sampling with those for alternative sample media in a variety of arid region environments. Analysis of lag samples for Au, Cu and As clearly indicates the presence of bedrock Au mineralization in the Paterson and Eastern Goldfields Provinces of Western Australia. In these areas both lags and soils exhibit good anomaly contrast, but lags show more extensive lateral dispersion, leading to advantages in reconnaissance exploration. Strong anomalies for Ni and Cu are developed in lags, compared with subdued response in fine-fraction soils over a Ni sulphide occurrence in the Eastern Goldfields which has been subjected to deep lateritic weathering. Lag geochemistry also clearly reflects sub-economic base metal and barite occurrences in the McArthur Basin, N.T., in spite of the dilution of surface soils by sands probably related to a Mesozoic marine incursion. Orientation sampling over a Pb-Zn prospect in the Pine Creek Geosyncline has demonstrated optimum response in lag samples compared with various size fractions of the associated lithosols. Variable dilution of lag samples by coarse quartz sand can be a problem in areas with substantial transported overburden. A simple procedure to ‘correct’ trace-element values using regression analysis based on the Fe content of samples is described as a means of reducing ‘noise’ resulting from such matrix variations.