Quantification of Subsurface Hydrological Connectivity Using Wavelet Coherence Analysis
ObjectiveSubsurface hydrological connectivity governs the rainfall-runoff transformation relationship in headwater catchments and serves as a critical foundation for describing runoff generation and concentration mechanisms and simulating hillslope hydrological processes. While quantitative methods...
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Editorial Department of Journal of Sichuan University (Engineering Science Edition)
2025-01-01
|
| Series: | 工程科学与技术 |
| Subjects: | |
| Online Access: | http://jsuese.scu.edu.cn/thesisDetails#10.12454/j.jsuese.202500369 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | ObjectiveSubsurface hydrological connectivity governs the rainfall-runoff transformation relationship in headwater catchments and serves as a critical foundation for describing runoff generation and concentration mechanisms and simulating hillslope hydrological processes. While quantitative methods for subsurface structural connectivity have been developed, characterizing subsurface process connectivity remains a major challenge in hillslope hydrology research. Existing approaches face two key limitations: 1) a lack of quantitative analysis; and 2) insufficient resolution of spatial heterogeneity. This study employs wavelet coherence analysis to quantitatively investigate subsurface hydrological process connectivity and its spatiotemporal evolution patterns, providing a reference for analyzing and simulating hydrological and soil processes in mountainous small watersheds.MethodsDaily-scale soil moisture and rainfall observation data spanning three hydrological years (May 2011—May 2014) from the Shale Hills Critical Zone Observatory (Pennsylvania, USA) were utilized. One continuous slope was selected each on a sunny slope and a shaded slope. Eight monitoring sites were established covering four slope positions: ridge, upper slope, lower slope, and toe slope. Soil moisture sensor networks were installed within three genetic soil horizons: Horizon A (topsoil), Horizon B (subsoil/accumulation layer), and Horizon C (parent material/saprolite). Firstly, Continuous Wavelet Transform (CWT) was applied to extend the soil moisture time series into the time-frequency domain, using the Morlet wavelet as the mother wavelet due to its optimal time-frequency localization for hydrological signals. Secondly, wavelet coherence analysis based on CWT was employed to generate pairwise wavelet coherence spectra, quantitatively describing the multi-scale periodic synchronization between two time series for: a) vertical connectivity (between adjacent soil horizons at the same site: A-B and B-C); and b) lateral connectivity (between adjacent slope positions within the same soil horizon: ridge-upper slope, upper slope-lower slope, lower slope-toe slope). Statistical significance was assessed against a red-noise process at the 95% confidence level using Monte Carlo simulations (1 000 iterations). Thirdly, two quantitative metrics were calculated: a) the Percentage Area of Significant Coherence (PASC), computed for three operational scales – short-term (<1 week), medium-term (1 week to 3 months), and long-term (>3 months) – representing the proportion of significantly coherent regions within each scale domain; and b) the Global Coherence Coefficient (GCC), computed as the scale-averaged coherence to evaluate the persistence of connectivity across time scales. All analyses excluded the Cone of Influence (COI) boundaries to eliminate edge effects.Results and Discussions1)Vertical subsurface hydrological process connectivity was significantly higher than lateral connectivity across all slope positions and aspects (<italic>p</italic>< 0.05). At the monthly scale, wavelet coherence coefficients between soil moisture time series from different horizons were generally above 0.7, decreasing with increasing depth (GCC for A-B was higher than for B-C), indicating that soil water movement is dominated by vertical infiltration. This is likely related to rainfall event dynamics and soil structure: soil moisture in the study area exhibits a wetting-drying cycle driven by rainfall events, with shallower soils responding more sensitively to rainfall; soils with high rock fragment content facilitate the formation of vertical connectivity. 2)Pronounced aspect differentiation (shaded vs. sunny slope) was observed in lateral subsurface process connectivity. The shaded slope exhibited overall higher lateral connectivity (PASC: shaded slope 20.71%, sunny slope 13.44%), potentially due to preferential flow paths and pore networks enhanced by a thick litter layer, improving lateral flow continuity. The sunny slope showed strong lateral connectivity only locally (upper slope-lower slope transition), where slope curvature changes induced topographic convergent flow, and the development of fissures in sandy-gravelly soils enhanced lateral water transport capacity. 3)The spatial characteristics of vertical subsurface process connectivity were more complex. Global coherence analysis indicated stronger vertical connectivity in the upper slope area of the shaded slope compared to the sunny slope upper slope, possibly related to the thicker humus layer on the shaded slope, which favors surface water retention, deep percolation, and the development of vertical preferential infiltration paths. Calculation of the significant coherence area (PASC) revealed higher vertical connectivity in deep soils at the sunny slope lower slope position (52.14%) than at the shaded slope lower slope (39.78%), suggesting connectivity in lower slope areas may be influenced by topographic convergent flow. At the ridge position, deep soils on both aspects exhibited high global coherence, indicating strong soil moisture exchange between horizons B and C on both shaded and sunny ridge areas. This is associated with the high rock fragment content (>40%) in the ridge zone, facilitating the formation of continuous macropore networks. 4)Wavelet coherence analysis proved feasible for quantifying dynamic hydrological processes, effectively capturing multi-scale soil moisture resonance phenomena. The wavelet coherence coefficient matrix successfully revealed multi-scale patterns in soil moisture dynamics, providing a robust metric for quantitatively analyzing the coupling mechanisms between soil moisture and eco-hydro-meteorological factors. Future work could apply this method to other watersheds or hillslopes to validate its applicability across diverse geographical settings. Integrating it with advancements in high-resolution databases and analytical techniques could overcome limitations related to localized quantification and discrete scales, enhancing the capacity to analyze global information within complex hydrological systems.ConclusionsThe results demonstrate that wavelet coherence analysis is an effective time-frequency tool for quantifying hillslope hydrological process connectivity. Short- and medium-scale coherence are particularly indicative of precipitation-driven local water movement and exchange processes. Within the study area, vertical subsurface hydrological process connectivity was significantly higher than lateral connectivity and decreased with depth. Pronounced aspect differentiation characterized lateral connectivity (Mean PASC: shaded slope - short-term 19.25%, medium-term 17.62%; sunny slope - short-term 12.66%, medium-term 10.55%), with the sunny slope exhibiting strong lateral connectivity only locally within a concave section (upper-lower slope transition). The spatial patterns of vertical connectivity were complex: vertical connectivity in the shaded slope upper slope exceeded that in the sunny slope upper slope, likely due to preferential flow development on the shaded slope; conversely, vertical connectivity in deep soils at the sunny slope lower slope (52.14%) was higher than at the shaded slope lower slope (39.78%), potentially influenced by topographic convergent flow. These findings provide theoretical support for understanding and simulating hydrological processes in mountainous small watersheds. |
|---|---|
| ISSN: | 2096-3246 |