Abstract
The basic correlation that determines the mechanical and hydraulic characteristics of unsaturated soils is the Soil Water Characteristic Curve (SWCC). Critical synthesis The present review brings forth the latest developments in SWCC modeling, measurement and application, with a more specific interest in determining the factors that question the financial forecasting properties of current simple models, especially with dynamic environmental circumstances. In our analysis, we have found that current empirical models are frequently ineffective because they lack the explicit ability to integrate the complex/coupled nature of microstructural properties (e.g., fractal geometry and nano-porosity) and compositional differences (e.g., organic matter). More so, the synthesis shows that there is a fundamental conflict in modeling, whereas the more intricate a conventional empirical model gets, the less accurate it becomes, whereas sophisticated data-driven methods such as Deep Learning are much more effective at making predictions. As a consequence, this review confirms that one of the main future research needs should be the creation of coherent constitutive models that will combine mechanistic controls (fractal theory) and advanced AI prediction. Most importantly, the implications in practice reveal that the neglect of such coupled considerations directly compromises the validity of long-term geotechnical performance forecasts, that is, in relation to slope stability and foundation resilience to moisture changes caused by climate change. The review gives the conclusion by recommending that laboratory results should be immediately validated at a field scale to bring together the gap between theory and engineering design.