Statistical modelling and uncertainty reduction of monitoring data in geomechanics

The analysis of field data often shows unwanted trends, quasi periodical variations at the daily and/or yearly frequency and other fluctuations and noises which are not related to the phenomenon being monitored. For field instruments based on long hydraulic circuit (e.g. hydraulic settlement gages) or rods and wire gauges (e.g. extensometers), the temperature may have significant effects on the uncertainty of the measuring system. In this paper, we show how to reduce temperature drifts and the highly non Gaussian uncertainty by means of an appropriated computerized post-processing of the field data based on robust statistical methods. To do this routinely, we propose an easy-to-use software which helps the user in the data analysis and in the choice of the appropriate modelling technique.