资源环境科技发展态势分析平台

Time series of groundwater head measurements serve as a primary source of information on groundwater systems. In different groundwater systems, and across several scales, we observe a multitude of patterns in groundwater time series, resulting from complex hydrogeological setups. Unlike in surface hydrology, there is no generalized classification to categorize and quantify the dynamics in groundwater time series. This leads to a lack of tools that could help us disentangle the information contained in groundwater time series in a systematic way. To approach such a classification, we present a principle for organization to qualitatively describe and to quantify groundwater dynamics in a nonredundant and data efficient way. We devise a descriptive typology of groundwater dynamics and assign quantitative measures, mathematically expressing these dynamics. Based on an extensive data set of daily groundwater hydrographs from central Europe, we analyze the relationship between indices and typology based on principal component analysis. The principal component analysis is also used to investigate and discuss redundancy, that is, indices expressing similar information content of hydrographs. Further, the indices' sensitivity to measurement interval and length of the overall observed period is investigated. Finally, a case study demonstrates the potential of the typology and index approach to link groundwater dynamics to the underlying hydrogeological process controls. The tools provided for characterization and quantification of groundwater dynamics should improve future efforts of groundwater classification and prediction in ungauged aquifers and other applications.