What controls groundwater dynamics and hillslope-stream connectivity in an Alpine headwater catchment?

Abstract:

1. Introduction

Shallow groundwater variations in mountain catchments play an important role in runoff production, nutrient transport and landslide susceptibility (Zhang et al., 2011; Lanni et al., 2012; Penna et al., 2014). Groundwater responses are highly variable in time and space, even at small scales and are therefore difficult to predict (Haught and van Meerveld, 2011; Bachmair and Weiler, 2012). Non-linear processes, such as thresholds and hysteresis, often lead to additional complexity. Hysteresis between groundwater levels and streamflow has been observed in various catchments and has been related to differences in runoff timing across the landscape (McGlynn et al., 2004). Understanding what controls the differences in the magnitude of the hysteretic loops can give insight in the main mechanisms with which catchments retain and release water under different conditions.

The possibility for water to reach the stream and to carry solutes through subsurface flow pathways is enhanced when transient groundwater develops across the catchment and connects different landscape parts, ultimately establishing a hydrological subsurface continuum between the stream, riparian zone and hillslopes (Jencso et al., 2009). This has important implications for runoff and biogeochemical processes, since it facilitates a rapid displacement of water and nutrients from the upper zones of the catchment to the stream, possible yielding substantial contributions to peak flows (Bachmair and Weiler, 2014). Due to practical limitations in field monitoring, the heterogeneity of soil properties and the structure of subsurface flow pathways, the dominant factors that control the spatial and temporal variability of transient subsurface saturation patterns, the degree of groundwater-streamflow hysteresis and the development of hillslope-stream hydrological connectivity in mountain headwater catchments are still not well known.

In this study, we use a network of spatially-distributed piezometers in a small Alpine catchment to identify the main controls on: i) the spatio-temporal variability of shallow groundwater responses to precipitation; ii) the hillslope-stream subsurface connectivity; and iii) the variability in the degree of hysteresis in the water table-streamflow relation.

2. Study area and Methodology

Field measurements were conducted in the 0.14 km² Bridge Creek Catchment, a headwater catchment in the Italian Dolomites characterized by a clear morphological distinction between steep hillslopes and a relatively flat riparian zone (Penna et al., 2011). Elevations range between 1932 and 2515 m a.s.l.; the mean slope is 24.4°. The mean annual precipitation is 1220 mm, approximately half of which falls as snow. The main land cover is alpine grassland.

Fourteen piezometers were installed to refusal on the hillslopes and in the riparian zone, along three transversal transects. Installation depth ranged between 0.60 m and 1.60 m. Fifteen soil moisture probes were installed at 5 cm, 20 cm and 40 cm along one riparian-hillslope transect.

The water table response was analysed for 74 rainfall-runoff events between June and October in 2011, 2012 and 2013. Event rainfall ranged between 2.8 mm and 84.8 mm (average: 18.6 mm). Lateral connectivity was computed by assigning a "1" where two adjacent piezometers responded (i.e., there was a measurable and increasing water level in the piezometers), and "0" for two adjacent piezometers where the water table did not respond in at least one of the two piezometers. All pairs were then summed for each transect.

Streamflow and groundwater data for each piezometer were normalized for each event, so that they both varied between 0 and 1. Then, the normalized hysteresis area (NHA) was graphically quantified to determine the degree of hysteresis.

3. Results and Discussion

Antecedent soil moisture (expressed by the antecedent soil moisture index, ASI, computed as the average soil moisture of all probes multiplied by the installation depth) and rainfall intensity were poor predictors of the number of piezometers that responded during an event but event precipitation depth (P) and the combination of ASI and P explained most of the variation in water table response, confirming previous results at BCC (Penna et al., 2014). Correlation coefficients computed between water table responses and selected characteristics of rainfall-runoff events were very similar for piezometers grouped according to the catchment topographic unit (riparian zone and hillslope zone) and grouped for each transect (Table 1). The number of piezometers that responded was also positively correlated with stormflow, suggesting the important contribution of subsurface flow to the catchment runoff (Table 1).

We identified a threshold in the relation between stormflow and ASI+P (see Fu et al., 2013). Above this threshold, a water table response was observed in all or almost all piezometers in the catchment. The water table was deeper on the hillslopes than in the riparian zone, and increasingly deeper from the lower (downstream) to the upper (upstream) transect, reflecting the effect of catchment topography on the spatial variability of subsurface flow generation (Haught and van Meerveld, 2011; Dhakal and Sullivan, 2014).

We grouped the 74 events in four classes according to their ASI+P value (Figure 1). On average, water table peaks were higher with increasing antecedent wetness conditions and precipitation depth. Overall, the degree of hillslope-stream hydrological connectivity increased as a function of ASI+P: flow paths appeared mostly disconnected during dry conditions and small events, whereas the monitored part of the catchment appeared connected during wet conditions and large rainfall events (Figure 1). For 21 events, falling in the two classes with the highest ASI+P values, connectivity was 100% and runoff coefficients were generally high (on average, 18%). Connectivity was generally higher for the lower and middle transects than for the upper transect. The variability in connectivity was larger for the upper transect than for the other transects.

The NHA was calculated for each piezometer for each of the 21 events during which the water table responded in all piezometers. Generally, the NHA was smaller during very wet conditions and for large rainfall events: catchment-average and transect-average NHA decreased when ASI+P increased. Moreover, the NHA was smaller for events with a larger runoff coefficient. This suggests that during wet conditions, when preferential flow pathways are activated and hillslope-stream connectivity is high, rapid hydrological dynamics occur, which, in turn, lead to changes in water levels and streamflow that are more in phase than during drier periods and smaller rainfall events. The median and the coefficient of variation of the NHA were negatively correlated, indicating that the piezometers that showed a smaller median NHA were characterized by a higher variability. This probably indicates that despite the fast hydrological dynamics that occur during wet periods and that result in more synchronous streamflow and groundwater responses, certain parts of the catchment still tend to have a different timing and therefore a high variability of the NHA.

These results highlight the importance of transient groundwater levels and subsurface hillslope-stream connectivity for the overall hydrological response of the catchment and shed new light on the controls on the variability of the size of the hysteretic loops between streamflow and water levels. However, further analyses are needed to better understand the complexity of the variability of hysteretic relations in this catchment and to define a more complete conceptual model of subsurface flow generation across the landscape and during different conditions.

4. Conclusions

The analysis of shallow groundwater level data collected in the small Dolomitic Bridge Creek Catchment for 74 rainfall-runoff events showed that:

-An antecedent soil moisture plus event rainfall (ASI+P) threshold controls stormflow and subsurface flow generation.

-The ASI+P threshold reflects the fraction of piezometers for which the water table responds in different parts of the catchment. For large events during wet conditions, water table responses were observed at all piezometers.

-Subsurface hydrological connectivity varies as a function of antecedent soil moisture and rainfall depth, extending upwards from the riparian zone and the lower part of the catchment.

-The area of the hysteretic loop between streamflow and water table is typically smaller during wet conditions and large events, when connectivity and runoff are highest, suggesting more rapid and more synchronous hydrological responses.

References

Bachmair S., Weiler M., 2012. Hillslope characteristics as controls of subsurface flow variability. Hydrology and Earth System Sciences, 16, 3699-3715. doi: 10.5194/hess-16-3699-2012.

Bachmair S., Weiler M., 2014. Interactions and connectivity between runoff generation processes of different spatial scales. Hydrological Processes, 28, 1916-1930. doi: 10.1002/hyp.9705.

Dhakal A.S., Sullivan K., 2014. Shallow groundwater response to rainfall on a forested headwater catchment in northern coastal California: implications of topography, rainfall, and throughfall intensities on peak pressure head generation. Hydrological Processes, 28, 446-463. doi:10.1002/hyp.9542.

Fu C., Cheng J., Jiang H., Dong L., 2013. Threshold behavior in a fissured granitic catchment in southern China: (1) analysis of field monitoring results. Water Resources Research, 49, 1-17. doi:10.1002/wrcr.20191.

Haught D.R.W., van Meerveld H.J., 2011. Spatial variation in transient water table responses: differences between an upper and lower hillslope zone. Hydrological Processes, 25, 3866-3877. doi:10.1002/hyp.8354.

Jencso K.G., McGlynn B.L., Gooseff M.N., Wondzell S.M., Bencala K.E., Marshall L.A., 2009. Hydrologic connectivity between landscapes and streams: transferring reach- and plot-scale understanding to the catchment scale. Water Resources Research, 45, W04428, doi:10.1029/2008WR007225.

Lanni C., Borga M., Rigon R., Tarolli P., 2012. Modelling shallow landslide susceptibility by means of a subsurface flow path connectivity index and estimates of soil depth spatial distribution. Hydrological Earth Systems Science, 16, 3959-3971, 2012 doi:10.5194/hess-16-3959-2012.

McGlynn B.L., McDonnell J.J., Seibert J., Kendall C., 2004. Scale effects on headwater catchment runoff timing, flow sources, and groundwater-streamflow relations. Water Resources Research, 40, W07504, doi:10.1029/2003WR002494.

Penna D., Mantese N., Hopp L., Dalla Fontana G., Borga M., 2014. Spatio-temporal variability of piezometric response on two steep alpine hillslopes. Hydrological Processes. doi: 10.1002/hyp.10140.

Penna D., Tromp-van Meerveld H.J., Gobbi A., Borga M., Dalla Fontana G., 2011. The influence of soil moisture on threshold runoff generation processes in an alpine headwater catchment. Hydrology and Earth System Sciences, 15, 689-702. doi: 10.5194/hess-15-689-2011.

Zhang B., Tang J.L., Gao C., Zepp H., 2011. Subsurface lateral flow from hillslope and its contribution to nitrate loading in streams through an agricultural catchment during subtropical rainstorm events. Hydrology and Earth System Sciences, 15, 3153-3170. doi: 10.5194/hess-15-3153-2011.

Table 1. Spearman rank correlation coefficients between water table responses and characteristics of rainfall-runoff events (n=74). Piezometers were grouped according to the their position in the riparian zone or the hillslope zone and according to each transect. Significant correlations (p<0.05) are marked in bold.

	Frequency of water table response during rainfall events (%)
	Entire catchment	Riparian zone	Hillslope zone	Lower transect	Middle transect	Upper transect
ASI (mm)	0.19	0.31	0.06	0.19	0.21	0.10
P (mm)	0.82	0.78	0.78	0.79	0.77	0.73
ASI + P (mm)	0.83	0.82	0.78	0.81	0.79	0.72
Mean rainfall intensity (mm/15 min)	0.05	0.01	0.08	0.01	0.02	0.04
Stormflow (mm)	0.86	0.87	0.78	0.82	0.83	0.75

<< Previous Abstract | Next Abstract >>