Quantification of subsurface hydrologic connectivity in four headwater catchments using graph theory

G Zuecco; M Rinderer; D Penna; M Borga; H J van Meerveld

doi:10.1016/j.scitotenv.2018.07.269

Quantification of subsurface hydrologic connectivity in four headwater catchments using graph theory

Sci Total Environ. 2019 Jan 1:646:1265-1280. doi: 10.1016/j.scitotenv.2018.07.269. Epub 2018 Jul 20.

Authors

G Zuecco¹, M Rinderer², D Penna³, M Borga⁴, H J van Meerveld⁵

Affiliations

¹ Department of Land, Environment, Agriculture and Forestry, University of Padova, Italy. Electronic address: giulia.zuecco@unipd.it.
² Nicholas School of the Environment, Duke University, NC, USA; Chair of Hydrology, University of Freiburg, Germany.
³ Department of Agricultural, Food and Forestry Systems, University of Florence, Italy.
⁴ Department of Land, Environment, Agriculture and Forestry, University of Padova, Italy.
⁵ Department of Geography, University of Zurich, Switzerland.

PMID: 30235612
DOI: 10.1016/j.scitotenv.2018.07.269

Abstract

Hillslope-stream connectivity significantly affects streamflow and water quality responses during rainfall and snowmelt events, but is difficult to quantify. One approach to quantify subsurface hillslope-stream connectivity is graph theory, which considers linear connections between groundwater measurement sites. We quantified subsurface connectivity based on surface topography and shallow groundwater data from four small (<14 ha) headwater catchments in the Italian Dolomites and the Swiss pre-Alps, determined the relation between rainfall, antecedent wetness conditions and subsurface connectivity and assessed the sensitivity of the results to changes in the measurement network. Event total stormflow was correlated to maximum subsurface connectivity. Subsurface connectivity increased during rainfall events but maximum connectivity occurred later than peak streamflow, resulting in anti-clockwise hysteretic relations between the two. Subsurface connectivity was positively correlated to rainfall amount. Maximum subsurface connectivity was related to the sum of total rainfall plus antecedent rainfall for the Dolomitic catchments, but these relations were less clear for the pre-alpine catchments. For the pre-alpine catchments, the fractions of time that the groundwater monitoring sites were connected to the stream were significantly correlated to the upslope site characteristics, such as the Topographic Wetness Index. For the Dolomitic catchments, the fractions of time that the monitoring sites were connected to the stream were correlated to the topographic characteristics of the upslope contributing area for the catchment with the small riparian zone, and with the distance to the nearest stream for the catchment with the large riparian zone. The leave-one-out sensitivity analysis showed that small changes in the structure of the groundwater monitoring networks had a limited influence on the results, suggesting that graph-theory approaches can be used to describe subsurface hydrologic connectivity. However, the proposed graph-theory approach should be verified in other catchments with different groundwater monitoring networks.

Keywords: Graph theory; Headwater catchments; Hydrologic connectivity; Hysteresis; Shallow groundwater; Subsurface runoff.