Answers are grounded in this paper and cite their sources.
The paper presents an improved multistep statistical model (mHHSTT) for the analysis of monitoring data of arch dams, with a particular focus on displacement time series. Building upon the classical Hydro-Seasonal-Time (HST) model, the proposed approach progressively introduces nonlinear predictors within a constrained regression framework. This enables the model to capture delayed thermal and visco-elastic effect responses to changes in air temperature and reservoir level, while reducing the risk of overfitting and preserving the physical interpretability of the regression model. The framework is further extended to highly nonlinear time series exhibiting plateau-like responses, such as those associated with joint closure, by combining weighted regression models, each representing a distinct range of homogeneous structural behaviour. The paper also provides a stepwise guideline for the practical application of the model that can be readily implemented using a standard spreadsheet. The proposed methodology is validated using the arch dam owned by Électricité de France (EDF), which served as the benchmark case for the 2022 ICOLD International Benchmark Workshop (Theme A). The analysis of the measures of two pendulums and a short-base extensometer highlights significant improvements in both the explanatory and predictive performance of the basic HST model. Although the study focuses on a single, representative case, the generality of the proposed methodology is investigated in a companion paper through application to additional dams and monitoring instruments. Beyond enhancing predictive accuracy, the proposed strategy offers insights into the constitutive behaviour of dam-foundation systems, thereby supporting the development of more physically informed models in the context of dam safety assessment.
No takes yet. Share an insight, caveat, or question.
Lusini et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: