Applicability of measured soil shrinkage characteristic curves to in-situ land movement monitoring data for expansive soils

Normal seasonal variations in wet and dry conditions cause vertical land movement, due to shrinkage and swelling, particularly in deposits rich in expansive clay minerals or organic matter. This land movement can be up to decimeters depending on the characteristics of the deposit, causing damage to infrastructure and buildings. Furthermore, the increase in drought duration and intensity results in a lowering of the groundwater table to unprecedented levels in shallow deltaic and coastal subsurface deposits. Exposure to resulting low water contents in the unsaturated zone induces irreversible shrinkage and densification of the soil structure, contributing to land subsidence.

Understanding the relation between water content and bulk volume over time is crucial for predicting the land movement and assessing the potential for land subsidence and associated structural damage. The shrinkage behaviour is commonly characterized using soil shrinkage characteristic curves, which relate the water content to bulk volume.

In this study, we characterized the soil shrinkage characteristic curves (SSCCs) of undisturbed natural samples extracted from the capillary fringe zone of shallow, marine clay-rich deposits in the Netherlands. The sample site was selected based on the presence of an extensometer, constantly monitoring the water content and thickness of the soil layer. The SSCCs were created by measuring the water content and sample volume during air drying of the samples. The sample volume was measured with optical distance sensors. The water content decreased from 0.97 to 0.08 during the experiment and the void ratio from 0.97 to 0.34.  The SSCCs of the samples show a linear relation between water content and void ratio, for a water content between 0.97 and 0.34.

By applying the established linear relation to the monitored in-situ soil moisture content at the sample site, we predicted the change in thickness of the soil layer. The prediction overestimates the measured changes in land movement, indicating the importance of in-situ conditions. The measured land movement seems to be more responsive to the groundwater level than the water content. The and the coming results underscore the importance of measuring land movement in-situ.