Interpretation of a Subsurface Soil Profile Using a Limited Data Set and Geostatistics

Constrained soil investigation works bring inevitable uncertainties in estimating the true geotechnical design parameters. This uncertainty leads to misinterpretation of failure analysis which can later cause catastrophic damage to infrastructure. In order to avoid such circumstances, one possible way is to conduct additional in-situ tests. However, increase in in-situ tests results in additional project costs and produces negative environment through carbon emissions. In this research study, an approach is proposed to interpret the subsurface profile using geostatistical theories. The cone penetration test soundings (CPT) obtained from a soil exploration study conducted in Texas, USA was considered for the analysis. First, all the available CPT soundings were used to predict and map the variation of tip resistance, friction ratio and soil behavior type index over entire area using geostatistics. Later, a subset of original CPT soundings was used to predict all the three parameters over entire area. Both the maps produced from geostatistical analysis were compared to ascertain the quality of predictions. The results and analysis depicted that the geostatistics could able to accurately predict the soil properties with only a subset of total CPT soundings. This research highlights the adaptability of geostatistics in predicting the subsurface profile with limited data and its contribution towards the sustainable environment.