Variability of Water Exchange in the Hyporheic Zone of a Lowland River in Poland Based on Gradientometric Studies

The Moszczenica, a 55-km-long third-order river, flows into the Bzura River on the northern slopes of the Łódź Hills. The river flows into the main stream on the Łowicko-Błońska Plain (Fig. 1). According to the river typology in Poland (Błachuta et al. 2010), the Moszczenica represents Poland's most frequent type of lowland sandy stream. It is represented by almost 1800 Surface Water Bodies (SWB) out of a total of 4504 identified. Such water body type shows a catchment area developed in periglacial plains; it is found in wide box valleys. The riverbed has a winding course, meandering, with rock riffles separated with long sections with a calm water stream.

Fig. 1

In the Holocene (Boreal-Atlantic), in the middle course of the Moszczenica there occurred a meander runoff, as suggested by an approximately 100-m-wide meander alluvial bar zone. The Holocene valley bottom which is 100–300 m in width includes numerous paleomeanders with a longer abandoned section of the river. The nature of the Holocene runoff was evidently affected by Preboreal and Boreal intensive aeolian processes, which activated the nearby deposited Late Vistulian aeolian fields which, while entering the valley, partitioned it and blocked the river flow. It resulted in valley flooding and forced further river meandering (Kamiński 1993). The present shape of the middle Moszczenica River channel course covered by the research demonstrates a clear river winding and length reduction, which is much due to the human impact dating back to the 14th century. The construction of forges and mills resulted in an artificial river flow division into several channels and in a development of reservoirs, and thus an additional bottom sediment deposition before the weirs and the river's further incision into the sediments when the barriers were eliminated. Thus, the longitudinal profile of the middle Moszczenica section became irregular and straight channel sections are separated with strongly meandering deepened fragments with steep erosional banks. In the valley bottom, numerous former so-called dead channels have been preserved, which are today partially used for a fish farming reservoir construction.

The Moszczenica River catchment area is 519 km² (Czarnecka 2005), and it is very diverse in terms of morphology. The topographic watershed runs through the highest point of the Łódź Hills, the Dąbrowa Hill, at a height of 284 m a.s.l., whereas the mouth of the Moszczenica River to the Bzura River is found at 95 m a.s.l. The Moszczenica catchment represents the upland type, with the riverbed slope of 2–5‰, and the catchment slope of 5–20‰ (Szczepański 1995). Substantial height differences, combined with a northern and north-western exposure, led to the development of a strong precipitation gradient of approximately 70 mm P/100 m H (Dubaniewicz 1974). The morphological, geological and climate conditions contribute to an evident variation of the Moszczenica catchment in terms of runoff and its components (Table 1).

The Moszczenica River represents an abundant category of rivers with a very well-developed nival regime, specific for the rivers in the central and eastern parts of Poland. In the annual cycle for those rivers one can specify the following: high meltwater floods in spring followed by a rapid recession of river runoff and transition to a period of the low summer-autumn flow (Wrzesiński 2016, 2017).

Changes in the water levels measured in the gauging station of the Polish Institute of Meteorology and Water Management (IMWM) in Gieczno on the Moszczenica River, in the 2018 hydrological year the study was performed in, are presented in Figure 2. In that period was recorded two floods: pluvial in July and meltwater in February, and the lowest water stage was found in the second half of June.

Fig. 2

The Moszczenica River is supplied with groundwater from sediments above the clay layers, from fine sands of the flood plain and river terrace, and from sands with gravels of the adjacent alluvial plain. In the southern part of the Moszczenica River, from the west its valley adjoins the outwash plain limited with moraine hills. Groundwaters in the areas adjacent to the Moszczenica River valley occur 1.5–2.5 m deep and they show similar changes with water stages in the Moszczenica, recorded at the gauging station of the IMWM in Gieczno (Fig. 3).

Fig. 3

The Moszczenica water stage data provided by the gauging station in Gieczno (H_r) and the data from the neighbouring wells (No. 6 on the left bank and No. 3 on the right bank, Fig. 1C) facilitated the calculation of hydraulic horizontal gradients on the left bank ΔL and on the right bank ΔR in the Moszczenica bottom. The results are provided in Table 2.

The gradientmeter is a new device designed for in situ testing of the direction of water exchange in the hyporheic zone, developed at the Adam Mickiewicz University in Poznań.

The gradientmeter (Fig. 4) allows direct measurement of the difference in pressures between the surface water and groundwater 20 cm under the bottom (Marciniak, Chudziak 2015). It is the depth to which the active part (screen) of the gradientmeter is pushed into the bottom. A pipe (2) ending with a short screen (1) is welded to the lower part of the body (3) of the gradientmeter. A support (4) is screwed to the upper part of the body (3). The pipe (2) with a screen (1) is pushed into the bottom of the river or reservoir 20 cm deep. Water from the hyporheic zone flows through the screen (1) inside the pipe (2) with a hose attached with two quick release plates (5). The other end of the hose (5) is connected to the inverted-U-tube measurement system. The measurement system consists of a groundwater level measurement pipe (6), a surface water level measurement pipe (7) and a valve (8). The pipe (7) is connected to a hose with a quick release plate (9) with the other end of the hose being placed in a river or a water reservoir.

Fig. 4

The hydraulic gradient is measured by sucking ground and surface water into the measuring system. For the measurement to be performed, a vacuum must be created by sucking air through the valve (8), which will raise the water level to both measuring tubes (6) and (7). When the water level in both pipes reaches the middle of the scale, the valve (8) should be closed. The negative pressure generated in the measuring system maintains the level of ground and surface water in the measuring tubes (6) and (7). During downwelling, the water level in the tube (7) is higher than in the tube (6). The hydraulic gradient i_d is calculated by dividing the pressure difference Δh_d by the depth Δl_d of penetrating the gradient meter into the bottom sediments as: (1) id=ΔhdΔld {i_d} = {{\Delta {h_d}} \over {\Delta {l_d}}}

As for upwelling, the water level in tube (7) is lower than that in tube (6). Hydraulic gradient i_u is calculated by dividing the pressure difference Δh_u by depth Δl_u penetration of the gradientmeter into the bottom sediments as: (2) iu=ΔhuΔlu {i_u} = {{\Delta {h_u}} \over {\Delta {l_u}}}

The depth the gradientmeter is pushed into the bottom sediments is always the same: Δl_d = Δl_u = 20 cm, while the ± sign of the hydraulic gradient is conventional. In hydrology, groundwater drainage is considered to feed surface waters and, therefore, the hydraulic gradient sign is positive. An opposite convention applies in hydrogeology. Groundwater drainage reduces the groundwater resources and it is marked with a negative sign. In this paper, the gradient sign applied is compliant with the hydrological convention.

For the first study objective, a straight section of the Moszczenica River was selected, which was found in an immediate vicinity of the gauging station of the IMWM in Gieczno, without water obstacles (sunken trunks, aquatic plant clusters) that could additionally stimulate a hyporheic exchange. The bottom is sandy with specific bottom accumulation landforms, that is, ripple marks. In order to more fully determine the filtration properties in the hyporheic zone, six samples of bottom sediments were collected and subjected to granulometric analyses. Two samples were collected for each of the profiles: P1, P4 and P9 (Fig. 1C). The grain size composition results are presented in Figure 5 and in Table 3.

Fig. 5

The lithological characteristics of bottom sediments for particular profiles correspond to the range of variation between the zone of the highest water flow rate in the current and the riverbank zone with a considerably lower water flow rate in a given section of the river channel.

The measurements covered a 10-m section with the profiles evenly distributed at every 2.5 m (Fig. 1B). The measurement cycle was monitored with water level reading from the gauging station and the river flow measurement – with a current meter in the first measurement profile (Section A). For a low water level, measurements were taken in all the five profiles and during a high flow, due to difficult riverbed conditions, that is, a high-water level and its strength on the measurement staff, the measurements were limited to three Sections (A, C and E). The measurement section points in which the gradientmeter was pushed into the bottom were 1 m away from one another. Starting from the left riverbank, the measurement point was to be located not <0.5 m away (Fig. 6). As a result, in each of the sections, 5–6 hydraulic gradient measurements were taken. The gradient measurements were accompanied by a water depth reading in a given point. Every time the gradientmeter was pushed 20 cm deep into the river bottom, making sure the gradientmeter body was well-pressed to the bottom, which would eliminate any hydraulic contact between the river and the gradientmeter screen. The negative pressure was generated by sucking the air, which resulted in an increase in the surface water and groundwater up to the measurement pipes of the gradientmeter. The water levels in the measurement pipes were stabilised for 10–15 min and the measurement results were read out. The results were recorded in a special form and the data were transferred to a spreadsheet, where gradients were calculated following Eqs (1) and (2). The measurements were taken twice: at low water stages (24 June 2018) and during a flood after intensive rainfall (20 July 2018) (Fig. 2).

Fig. 6

Measurements of the water flow rate in Moszczenica were performed using the Hega 2 hydrometric mill with an accuracy of ±0.005 m³ · s⁻¹.

For the second objective, the measurements covered a 2 km section of the river (Fig. 1C). The upper part of the section includes a fragment of the medieval artificially dug canal and the lower part – a natural section with mild turns, with neither meanders nor hydrotechnical infrastructure. The canal was to increase the fall of water supplied to the former mill. It was dug in a somewhat higher part of the valley floor than the natural riverbed, which is reflected in the hypsometric profile of the area (Fig. 7). The old riverbed, in the axis of the valley, is located lower than today's canal of the Moszczenica River.

Fig. 7

The hydraulic gradient measurements were taken for profiles approximately 100 m away from one another (Fig. 1C). Prior to the gradient measurements, depth measurements were performed to find the deepest point, and it was where the gradientmeter was pushed in and, in reference to that point, two more points halfway to the left and to the right river bank were selected. The measurements were taken in three points of each of the 13 measurement profiles. The measurement methodology was the same as earlier for the straight section of the river.

In the present paper, the division of the Moszczenica River into sections results from the analysis of the map (Fig. 1C), where the old riverbed and the section replacing it since the Middle Ages are marked. Between the artificial section and the natural one, one can find a transitional section where the river gradually recovers its natural features. Such division coincides with the gradient measurement results.

In the region of the local gradient measurements (Figs 1B and 6), the depth measurements and the distribution of hydraulic gradients demonstrate that, for the Moszczenica channel section under study, the main current occurs at the left riverbank (Fig. 9A). At a low water level, upwelling dominates in the bottom of the river with a clearly axial inactive zone and a limited-in-space zone of strong upwelling at the right riverbank (Fig. 9B). The profiles (Fig. 8) show that upwelling occurs at the shallower right riverbank, while downwelling is poor and occurs at the left riverbank in the deeper part of the riverbed.

The river bottom morphology changed during a flood flow (Fig. 10A). The highest depths still occurred at the left riverbank, while shallowing at the right riverbank changed in shape slightly. In these conditions, downwelling was dominant in the bottom of the river. In the middle of the river and at the right bank, two considerably reduced upwelling zones retained their activity, while at the left riverbank a new inconsiderable upwelling zone developed (Fig. 10B).

The statistical parameters of the gradient measurement research results are provided in Table 5. Importantly, the upwelling zones in the middle of the river and at the right riverbank retained their activity despite an essential change in hydrological conditions. The upwelling zones can reflect a local anisotropy of alluvial formations where very active zones of preferred groundwater supply developed, as pointed out by Amoros et al. (1996). The zones can also constitute the final fragment of the stream line from deeper aquifers (Toth 1963, Hayashi, Rosenberry 2002, Sophocleous 2002, Wörman et al. 2002, Cardenas 2007, Song et al. 2016). The latter interpretation is supported by the comparison of hydraulic gradients calculated from the water levels of the Moszczenica River and groundwater in Well Nos 3 and 6 (Table 2), as well as the gradients measured in the river channel with the gradientmeter (Fig. 8). The gradients measured in the river bottom are more than five times higher than what would result from the supply of groundwater to the Moszczenica. It suggests the occurrence of a strong supply from deeper aquifers. The vertical supply is important enough that even during a flood flow the groundwater drainage zones are clearly visible (Fig. 10B).

At low water levels on the Moszczenica, upwelling was dominant. The upwelling gradient varied from 0 to 0.075 for a mean value of 0.016, and the downwelling gradient – from 0 to −0.015 for the mean value of −0.008. Of the 33 measurements taken with the gradientmeter, 81.8% pointed to upwelling and 18.2% to downwelling. During a high flow, the image of the gradient field changed considerably. In that case, downwelling was dominant, with the gradient value from 0 to −0.110 for the mean value of −0.027, and the upwelling gradient value varied from 0 to 0.030 with the average value of 0.018. Of the 21 measurements taken with the gradientmeter, 85.7% pointed to downwelling and 14.3% to upwelling. With the gradient measurements, the range of variation in the hydraulic gradient in the bottom of the Moszczenica (Fig. 9B and 10B) confirms the existence of clear ‘patches’ of upwelling and downwelling, as reported by Wondzell (2011). An increased importance in downwelling during a flood coincides with the observations by Harvey et al. (2012), Gerecht et al. (2011), Briggs et al. (2012) and Bhaskar et al. (2012).

The Moszczenica River part analysed (Fig. 1C) was divided into three sections: –

the upper part (profiles P1–P5, Fig. 11A), with the canal dug in the Middle Ages;

The distribution of hydraulic gradients in the respective profiles is provided in Figure 11A–C. The statistical analysis results for the hydraulic gradient measurement series are presented in Table 6.

For the entire Moszczenica section under study, the range of variation in the hydraulic gradient varied from 0 to 0.055 for upwelling and from 0 to −0.085 for downwelling. The mean gradient values were 0.019 for upwelling and −0.027 for downwelling.

In the upper Moszczenica section (in the canal), out of the 18 measurements of the hydraulic gradient, 83.3% pointed to downwelling and 16.7% to upwelling. The downwelling gradient ranged from 0 to −0.085, with the mean value of −0.029, and the upwelling gradient – from 0 to 0.005, with the mean value of 0.003.

In the transitional zone between the canal and the natural riverbed, 11 gradient measurements were taken: 63.3% pointed to upwelling and 36.4% to downwelling. The upwelling gradient varied from 0 to 0.045, with the mean value of 0.018 and the downwelling gradient from 0 to −0.040, with the mean value of −0.025.

In the lower Moszczenica section (the natural riverbed), the values were the opposite. Of the 17 measurements of the hydraulic gradient, 70.6% pointed to upwelling, and 29.4% to downwelling. Upwelling gradients varied from 0 to 0.055, with the mean value of 0.023, and downwelling gradients from 0 to −0.065, with the mean value of −0.023.

The gradient measurements taken for an approximately 1400 m section of the Moszczenica River documented the ‘escape’ of water from the canal dug a few centuries ago to the historical naturally developed drainage base. Despite the passage of hundreds of years, the natural drainage base of the Moszczenica River still remained active. It is important for identifying the hyporheic zone functioning.

Verifying the conditions of the interaction of the Moszczenica River with groundwaters along the downwelling section (from P1 to P4) involved the flow rate measurements in four profiles. For the results, see Table 7 and Figure 12.

Fig. 12

The flow rate measurements only along the section from P1 to P2 pointed to the Moszczenica River downwelling. Behind the P2 profile there is water inflow from the ditch and behind the P3 profile there is water inflow from the pond (Fig. 1C). The volume of lateral tributaries is difficult to determine. Under such conditions, only the measurements of the hydraulic gradient in the hyporheic zone can document the river downwelling.