摘要
Publications on soil and spring gases have been examined regarding their relationships with both tectonic and seismic activities. The main sources, behaviours and uses of species detected in soils and springs are displayed, and their mode of sampling and analysing briefly described. The main patterns of degassing in soils are described and we outline the wide range of geochemical signatures as the result of both permeability and mineralogical contrasts. Because thermomineral waters have been in contact with great volumes of crustal rocks at various depths, spring gases might be more representative of the local environment than soil gases. Moreover, gas signature comparisons show that spring gases are much more enriched with deep gases and slightly contaminated by atmospheric gases. Therefore, they can be considered as better samples for identifying precursors of earthquakes. Environmental perturbations are examined, and it is shown from divergent cases that pressure, temperature, soil moisture or earth tides may generate very high perturbations of the degassing process. Such effects demonstrate that no systematic correction law can be proposed and that removing external contributions from gas concentrations must be performed case by case. This demonstrates therefore the need for the simultaneous measurement of external parameters during gas monitoring. A qualitative examination of about 150 claimed precursors proposed in the literature has been reviewed. As noted by previous authors, anomalies appear at distances sometimes much greater than typical source dimensions, and occur in the field of strain higher than 10−9, most of them being in the field of strain higher than 10−8. Taking into account the very high heterogeneity of such a set of data, we can suggest that amplitudes of gas anomalies are independent of both magnitudes and epicentral distances of related earthquakes, suggesting local conditions to control amplitudes. On the contrary, precursory time and duration of anomalies seem to increase both with magnitudes and epicentral distances. Abundant evidence demonstrates the major role of crustal fluids in the earthquake cycle. Many works have outlined the fact that crustal instabilities can appear as the result of low stress/strain perturbations during loading. It has been suggested that motion of fluids may occur at various scales, from microcrack fluid transfer up to changes of hydraulic levels of water tables. The study of subsequent anomalies is expected to supply a tool for earthquake prediction. Following previous authors, we outline the need for further methodological improvements, including the setting up of multiparameter station networks and the simultaneous recording of the main external parameters (atmospheric pressure, water and air temperature, soil moisture) for signal processing.