Meteorological noise in crustal gas emission and relevance to geochemical exploration

The emission of gas from the earth's crust is a complex process influenced by meteorological and seasonal processes which must be understood for effective application of gas emission to geochemical exploration. Free mercury vapor emission and radon emanation are being measured in a shallow instrument vault at a single nonmineralized site in order to evaluate these influences on gas emission.Mercury concentrations in the instrument vault average 9.5 ng/m³ and range from < 1 ng/m³ to 53 ng/m³ with a strong seasonal effect. Mercury has a direct relationship to vault temperature, air temperature, soil temperature, barometric pressure, water table, and the frozen or thawed state of the soil. Air and soil temperature, barometric pressure, and relative humidity are most important in influencing mercury emission while soil moisture is also important in radon emanation. Diurnal cycles are common but do not occur on all days. A heavy precipitation event on a dry soil seals the soil resulting in a rise in mercury concentration. Precipitation on a soil that is already wet does not increase mercury emission because of the compensation caused by lowering of the soil temperature by the precipitation event. Freezing of the soil changes the physical state of the vault-soil-soil gas-atmosphere system and emits the lowest concentrations of mercury. Phase lag effects are likely important. Stepwise multiple regression of mercury as dependent variable with meteorological and seasonal parameters as independent variables gives a cumulative R value of 0.563 and R² of 0.317. The short-term noise coupled with phase lags are an important factor.The radon measurements integrated over weekly intervals smooth out much of the short-term noise. Stepwise multiple regression of radon as dependent variable with meteorological and seasonal parameters as independent variables gives a cumulative R value of 0.967 and R² of 0.934. In this portion of the study the variation in the radon emanation is adequately predicted by meteorological and seasonal parameters.