Hillslope–riparian‐stream connectivity and flow directions at the Panola Mountain Research Watershed

Abstract

Abstract:To obtain a better understanding of water's journey from the hillslope to the stream, water level time series from 26 groundwater wells in a hillslope-riparian study area in the Panola Mountain Research Watershed, Georgia, USA, were examined with respect to the occurrence of perched groundwater and groundwater flow directions. A connected groundwater table was taken as an indication of a hydrologic connection between the hillslope and the stream. Flow directions were determined based on observed water levels at different combinations of three neighbouring wells. The analyses indicated that the hillslope was disconnected from the stream most of the time but that almost the entire hillslope became connected to the stream during large rainfall events. This coincided with more sustained streamflow during these events. There were clear differences in the timing of the groundwater response between the upper hillslope, lower hillslope and riparian zone, with water levels near the stream and on the upper hillslope rising earlier than on the midslope. The flow directions were generally downslope and were most variable around a bedrock depression on the midslope. On the hillslope, the direction of the water table followed the local bedrock topography when the water level was low at the start and end of an event and the overall surface topography when water levels were high. In the riparian zone, flow directions either did not change significantly during events or turned from a more downstream (i.e. downvalley) direction to a more downslope (i.e. towards the stream) direction. These results highlight the competing influence of surface and bedrock topography on hillslope flow directions and improve our understanding of hillslope-riparian-stream connectivity.

Groundwater similarity across a watershed derived from time‐warped and flow‐corrected time series

“…These differences in groundwater dynamics between footslope, midslope, and upslope locations are in agreement with previous studies in the same catchment [ Rinderer et al ., ] and hillslope studies at other sites [ Detty and McGuire , ; Penna et al ., ; van Meerveld et al ., ; Blumstock et al ., ] that showed a strong relation between the distance to the stream and the correlation between groundwater and streamflow response [ Seibert et al ., ; Haught and van Meerveld , ; von Freyberg et al ., ]. However, in our study the relation between mean groundwater similarity relative to streamflow and the distances to the stream (either nearest stream or the outlet) was weaker than for TWI (and not significant for CFC based similarity for the distance to the nearest stream; Table ).…”

“…The response behavior of these locations is most likely dominated by an interplay of factors comprising not only of static controls, such as topography (predominantly the flux of water from upslope as represented by the upslope accumulated area, and the drainage as represented by the local slope), effective porosity, and saturated hydraulic conductivity of the soil, but also dynamic controls, such as antecedent groundwater levels, rainfall characteristics, and hydrologic connectivity to upslope and downslope locations. This variability was also observed in other studies and highlights that midslope locations are most variable in terms of groundwater dynamics [ Rinderer et al ., ; Tetzlaff et al ., ] and important in terms of catchment‐scale hydrological connectivity [ Stieglitz et al ., ; McNamara et al ., ; Ocampo et al ., ; van Meerveld et al ., ; Jencso et al ., ]. The synchronicity in timing between the groundwater response at midslope locations with a TWI between 4 and 6 and the streamflow response (as highlighted by the significantly higher mean CFC‐based similarity compared to other landscape positions) is a first (but not necessarily a causal) indicator of the importance of midslope locations as runoff source areas during events and catchment connectivity [ van Meerveld et al ., ; McNamara et al ., ].…”

“…The peak‐to‐peak lag times between groundwater and streamflow and their associated standard deviation also increased as a function of distance from the stream [ Haught and van Meerveld , ; von Freyberg et al ., ; Blumstock et al ., ]. However, this behavior cannot be generalized since groundwater levels have also been observed to first respond in the upper hillslope [ Tromp‐van Meerveld and McDonnell , ; Rodhe and Seibert , ; Penna et al ., ], or that groundwater in footslope and upslope locations responded earlier than in midslope locations [ van Meerveld et al ., ]. This variability can lead to spatial patterns of runoff contributing areas that differ between the falling and the rising limb of events [ Nippgen et al ., ].…”

Throughfall enrichment and stream nutrient chemistry in small headwater catchments with different land cover in southern Chile

“…However, recent studies have addressed that the water connectivity assumption is not always accomplished, especially in catchments that have a marked dry seasonality. Hillslope connectivity with stream waters should be connected in regions with high rainfall inputs or during certain storm events during the year (Ocampo, Oldham, & Sivapalan, ; Hopp & McDonnell, ; McGuire & McDonnell, ; Meerveld, Seibert, & Peters, ). Therefore, nutrient balances estimated as in previous studies could lead to doubtful results (Kirchner, ).…”

Runoff generation in a pre-alpine catchment: A discussion between a tracer and a shallow groundwater hydrologist

“…10a, nr III (Stieglitz et al, 2003;Ocampo et al, 2006;Jencso et al, 2009;Detty and McGuire, 2010;van Meerveld et al, 2015). (Bachmair et al, 2012).…”