Quantifying and Numerically Representing Recharge and Flow Components in a Karstified Carbonate Aquifer. Issue 11 (4th November 2020)

• Record Type: Journal Article
• Authors: Schuler, P.; Duran, L.; Johnston, P.; Gill, L.
• Is Part Of: Water resources research. Volume 56:Issue 11(2020)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2020-11-04
• Subjects: karst -- groundwater -- time series analysis -- signal decomposition method -- pipe network model -- Ireland; Hydrology -- Periodicals; 333.91
• DOI: 10.1029/2020WR027717
• Languages: English
• ISSNs: 0043-1397

Abstract

Abstract: Karstified carbonate aquifers are highly heterogeneous systems characterized by multiple recharge, flow, and discharge components. The quantification of the relative contribution of these components, as well as their numerical representation, remains a challenge. This paper identifies three recharge components in the time and frequency domain. While the analysis in the time domain follows traditional approaches, the analysis of the power spectrum allows frequencies associated with specific spectral coefficients and noise types to be distinguished more objectively. The analysis follows the presented hypothesis that the different frequency‐noise components are the result of aquifer heterogeneity transforming the random rainfall input into a sequence of non‐Gaussian signals. The distinct signals are then numerically represented in the context of a semidistributed pipe network model in order to simulate recharge, flow, and discharge of an Irish karst catchment more realistically. By linking the power spectra of the modeled recharge components with the spectra of the spring discharge, the information usually gained by classical performance indicators is significantly widened. The modeled spring discharge is well matched in the time and frequency domain, yet the different recharge dynamics explain the signal of the aquifer outlet in different noise domains across the spectrum. This study demonstrates the conjunctive use of frequency analysis in conceptualization of a hydrological system together with modeling and evaluation. Key Points: Hydrograph analysis was coupled with frequency and noise analysis to identify recharge dynamics in a karst spring discharge time series Recharge, flow, and discharge were modeled in a semidistributed pipe network model considering aquifer heterogeneities and noise domains A new model evaluation approach using noise analysis provided information on the individual contribution of the internal flows