Real-Time Groundwater Dynamics Verification in the Embankment’s Substrate during the Transition of a Flood Wave

Abstract

The scope of the presented research included real-time verification of groundwater dynamics in the zone of the filled erosion channel (crevasses) and in the non-transformed zone of the floodplain area during the transition of a flood wave in the river channel. The technical goal was to provide data for the calibration and verification of mathematical model of groundwater flow. For this purpose, automatic recorders of groundwater level and electrical conductivity were installed in the zones selected earlier with the use of DEM. The measurements were carried out in 3 series during the passage of the flood wave. The obtained results indicate that in the zone of the untransformed terrace, the ascension of the water level between embankments causes the immediate propagation of pressure in the aquifer, while the filtration process itself is considerably limited, whereas the filled crevasse troughs constitute paths of privileged filtration, in particular in the proximal part of the floodplain. The appearance of water with elevated conductivity in the area of the crevasse proves the cyclicality of changes in flow directions, depending on the water level between embankments. The proposed methodology can be a valuable tool in the process of the geotechnical assessment of the construction substrate in the area of flood terraces in the lowland river. The zones with increased water conductivity parameters located near the river channel are also a reasonable place for the construction of coastal water intakes of the Riverbank filtration (RBF) type.

1. Introduction

The land management of river valleys requires knowledge not just on the current status of their natural environment, but also on its sensitivity to changes in its individual components, which affect the hydrological regime. The ability to predict the course of the evolution of the fluvial environment has currently become particularly important, when changes related to the land development of catchment areas of rivers or the bottoms of their valleys are accompanied by climate changes [1,2,3,4].

The sensitivity of the fluvial environment to changes in the hydrological regime depends mainly on the geological structure of the individual valley segments. Such dependence is vividly exemplified by river valleys in the area of the Polish Lowlands. Numerous river valleys in this area are characterized by their morphodynamic immaturity [5,6,7]. Referring to the classical concept of terrain evolution by [8], the lack of maturity manifests itself by the absence of a fully developed erosional bases, which seems to be an exception for lowland river valleys. This is because the lowland nature of the catchment areas of these rivers is not only a result of the advancement of denudation processes—it results from the glacial genesis of this area and the course of the processes of transgression and deglaciation of subsequent Scandinavian ice sheets in the Pleistocene [9]. This concerns not just the area of the Polish Lowlands, but also a major part of the North European Plain.

The geological history of the valleys in the Polish Lowlands dates back to the deglaciation of the last ice sheet in a given area. In the case of central Poland, it was the end of the Odra glaciation (MIS 8), only approximately 250 ka BP, whereas in the north-western part of the Polish Lowlands, this occurred even later, after the Vistulian Glaciation (MIS 2), approximately 22 ka BP [10]. The end of the Pleistocene was marked by the filling of the valleys by braided rivers overloaded with transported material. In the Holocene, along with the gradual warming of the climate and the alignment of discharges [11], the rivers started meandering [5,12,13]. They cut into the previously created surface, starting to build a Holocene floodplain.

Under the conditions of a uniform flow over the hydrological year, small surges of the meandering rivers caused the deposition over series of channel deposits and series of loamy overbank sediments (river muds) on the floodplain surface. In the case of Vistula, the largest river of the Polish Lowlands, the thickness of such sediments reached 7 m. Another change in the nature of river processes is dated to the sub-Atlantic period of the Holocene [13,14,15,16]. The management of the catchment area, and especially forest thinning, resulted in an increase in surface runoff and an increase in the differences between the maximum and minimum flows. Under such conditions, the channels changed again from meandering to braided ones. The nature of sediments deposited by rivers has changed. Channel, as well as flood sediments, are characterized by greater granulometric and lithofacial differentiation. The changes in the hydrological regime also increased the depth of flood erosion. During floods, deep erosion began to reach the surface of the sub-alluvial bedrock, especially in places where it forms morphological protrusions [17,18]. Their presence is a direct result of the above-mentioned morphodynamic immaturity of the valleys in the Polish Lowlands. The morphology of sub-alluvial bedrock thus affects the arrangement of flood flows currents and causes its incursion into the floodplain area, each time in the same zones [17,18,19]. This caused the creation of deep erosional troughs, cutting zones of natural levees. During the ebbing of the flood wave, the erosional troughs are filled with loose, highly permeable channel material, mostly sand. The repetitiveness of this type of phenomena has led to the transformation of fragments of the floodplain [18,20]. Concentrated flows of flood water led to the formation of often extensive depositional landforms as crevasse splays on the surface of the distal floodplain. The top part of the series of loamy flood deposits from the meandering period has been locally washed out. Such a transformation has influenced the hydrogeological regime of proximal floodplain area. The significance of contemporary floods for the shaping of hydrogeological conditions in a river valley, reaching up to sub-alluvial bedrock protrusions, was a subject of previous works of the authors’ team [17,18,20,21,22,23]. The studies were performed in the middle reach of the Vistula valley. The research resulted in the identification of zones of privileged flow of groundwaters within the alluvial, near-surface aquifer. These are the ones made of erosional troughs (crevasses) filled with loose channel sediments. Their depth ranges from 1.5 to over 8 m [24]. The impact of their presence on the hydrogeological regime of the alluvial aquifer has been confirmed by modelling studies [19,25]. The results of the mathematical modelling of hydrogeological conditions confirmed the significance of the forms of flood erosion (e.g., crevasses) for the shaping of a groundwater flow area in the alluvial stratum.

In the case of geomorphological studies of the Polish Lowlands, the digital elevation model (DEM) is particularly useful. It based on airborne laser scanning (ALS) as a part of the ISOK project (IT System of the Country Protection against extreme hazards). The ISOK project is carried out in connection with the need to implement [26]. It obliges member states to prepare planning documents for flood risk management and ensure public access to their results. As part of the ISOK project, flood risk and flood hazard maps were prepared for the entire territory of Poland [27].

In the case of embanked rivers, during flooding episodes, the intensity of the groundwater flow inside the erosion troughs (crevasses) filled with loose material could lead to the development of filtration deformations (suffosion). Since such phenomena are a frequent cause of embankment failure during floods [17,28,29,30], an important practical issue is a possibility of detecting such zones, as well as monitoring the dynamics of changes in the parameters of groundwater flow, during floods.

Research works in the field of hydrogeology based on remote-sensing reconnaissance often lead to the development of a mathematical model of groundwater flow. Each developed model imitates reality in a better or worse way. In order to maintain the credibility of the model, it is necessary to provide reliable data not only on the groundwater level, but also on the flow dynamics. Such data can be provided by automated recorders. The current state of technology allows the installation of electronic recorders in previously prepared boreholes (piezometers) and conducting reliable measurements for many years with the use of one battery. These techniques can significantly increase the reliability and usefulness of the results of research conducted, in particular in the contact zone of groundwater and surface water [31,32,33]. It is also worth noting that automated measurement techniques allow the documentation of dynamical flows, such as those occurring in the floodplain aquifer during flood wave transitions.

Therefore, the main objective of the presented research was to propose a methodology for the verification of the results obtained by DEM analyses during the flood on the river in real time with the use of automated recorders.

2. Materials and Methods

2.1. Location of Study Area

The studied area covers a wide, fully embanked floodplain of the Vistula River near Magnuszew village in central Poland (Figure 1). The floodplain terrace has an area of approximately 50 km², and the length of the flood embankment is about 15 km; the average height of the embankment is 5 m. It starts in the town of Ostrów and ends where the Pilica River flows into the Vistula River. It is an area mostly occupied by orchards and agricultural crops, and less by allotment gardens. The area is not urbanized.

2.2. Methodology

The research was conducted in three stages: stage 1 and stage 2 are described in the previous article [19] concerned with hydrogeological exploration based on archival materials, hydrogeological drilling, remote-sensing methods and mathematical modeling of groundwater flow. They are summarized in this capture below.

The research presented in this manuscript (Stage 3) is an innovative approach to verification and provides a basis for the recalibration of the hydrogeological mathematic model. The scheme of the procedure is presented in Figure 2.

2.2.1. Geological and Hydrogeological Background

The conducted research concerns the temperate climate zone. It was conducted in a lowland area. The study covered the valley of the Vistula River, the largest river in Poland. The unit of alluvial sediments forming the analyzed zone can be divided into three genetic series:

(1)

The first one consists of the alluvia of a braided river, which functioned here at the end of Last Glacial Period and transition to early Holocene [12,15,34,35,36]. They are mostly medium and coarse sands (Figure 3 and Figure 4). Those deposits build up the upper terrace of Vistula valley, as well as partly substratum of Holocene alluvia.

(2)

The second series consists of meandering river alluvia deposited in the Holocene up to 300 years BP [12]. Due to the considerable decrease in the dynamics of channel processes, resulting, e.g., from the development of dense plant cover on the basin area, that limit surface runoff [14], fine-grained material was deposited during the surges on the floodplain area [37]. This resulted in the considerable thickness of surge sediments formed as loams and silty loams; their thickness reaches 7 m in the studied area [18] (Figure 3 and Figure 4).

(3)

The third series origin is a result of changes in the hydrologic regime related to basin area management starting effectively about 300 years BP (according to the concept presented by [12,15]). Uneven discharges of contemporary river caused deposition of more differentiated alluvia. The channel deposits are medium and coarse sands with gravel intercalations. The flood deposits consist of sediment ranges from sands, silty sands, silts, to silty loams, loams and clays. More often, the flood sediments of a modern braided river occur in the top part of the complex of the Holocene sediments, above loamy muds (Figure 4). However, in the zones of concentrated flood flows, loamy flood deposits were removed and replaced by the silty and sandy muds of the contemporary river.

In the analyzed zone, the main aquifer consists of channel deposits of late Pleistocene and Holocene rivers. From the bottom, the layer is limited by poorly permeable glacial sediments, as well as Neogene loam. From the top it is limited by a complex of poorly permeable flood deposits. The layer is cut into the zones of concentrated flood flows. Erosional troughs (crevasses) that are cut through the proximal floodplain belt are filled with sands deposited during the waning phase of floods. Such forms of different sizes cross the proximal floodplain transversely [19]. They occur in the bedrock of flood embankments.

The potential impact of their presence on hydrogeological conditions in the area has been indicated based on the performed modelling studies. Under the conditions of both an effluent and an influent stream, when water in the river rests against a flood embankment, these zones constitute collectors for privileged filtration (Figure 5). The methodology of establishing such zones based on geomorphological criteria with the use of remote-sensing techniques was discussed in other paper [19], and is briefly described below in this article.

Within the floodplain, in the Quaternary stage, the presence of two aquifers was found. These are:

(1)

Near-surface aquifers built of fine sands with intermittences of dust and clay sediments of flood flows of the contemporary, braided Vistula River. Sand filtration coefficient values are 6 × 10⁻⁶ m/s based on laboratory tests. These sediments do not form a continuous cover over the entire area of the floodplain within the research polygon. The level within the distal floodplain is underlined with a series of poorly permeable flood formations with a thickness of 1 m up to 4 m. Groundwater table of this level, under conditions of average rainfall, remains at a depth of 99–101 m above sea level. This level is not exploited.

(2)

The second aquifer is made of multi-grained sands and gravels with a thickness of 1.5 m to 9 m-sediments of the channel facies of both the meandering and braided Vistula. This is the main level aquifer in the floodplain area. The coefficients of permeability of these deposits (measured by laboratory method) are from 8 × 10⁻⁶ to 2 × 10⁻⁵ m/s. In the entire area of the floodplain, outside the zone’s erosive cuts in a series of flood formations (filled crevasses and overflow channels), ditches and oxbow lakes, the second-level head stabilizes at the ordinate from about 99 up to 100 m above the sea level.

Groundwater dynamics of the floodplain area are closely related to the water level in the riverbeds of the Vistula River.

2.2.2. Remote-Sensing Analyses and Groundwater Dynamics Measurements

The basic material for the identification of crevasse channels was the Digital Elevation Model (DEM). The DEM was based on LIDAR data collected in 2011–2012 as part of the ISOK project. As part of this project, the altitude data were obtained in two standards with different density of measurements. Standard 1 ALS with a density from 4 to 6 points/m² was made on the area of 271,001 km² and covered almost all river valleys in Poland. Standard 2, dedicated to urban areas, with a density of 12 points/m², was made on an area of 13,769 km². Due to the availability of data and the potential universality of the developed method, it was decided to use the data in Standard I. The point cloud density for the studied part of the Vistula valley was ≥4 points/m². The ratio of the average distance of laser points in one line and the average distance of adjacent lines in the central zone of the series ranged from 1:1.5 to 1.5:1. The lateral coverage between the rows was ≥20% and the scanning angle was ≤ ±250. In the case of Standard I, the height accuracy of the point cloud (mh) was ≤0.15 m and the situational accuracy (mp) ≤0.50 m. The points obtained in this standard were divided into 9 classes: created, newer classified (class 0), ground (grade 2), low vegetation (grade 3), medium vegetation (grade 4), high vegetation (grade 5), building (grade 6), low point (grade 7), water (grade 8) and overlap points (grade 9). DEM was used in the work, based on points from class 2 (ground). Data in the format ASCII XYZ GRID with 1-m spatial resolution was used. The visualization of this model was also used, made available with the use of WMS technology by the National Geoportal [38]. This visualization determines “dynamic hypsometry” and automatically assigns a color scale depending on the height range in the currently analyzed area. The color composition was selected so as to emphasize slight changes of a dozen or so cm in relative heights in the vicinity of the flood embankment. This was to highlight the rectilinear forms perpendicular to the embankment, which would be a testimony of past flood flows.

Remote-sensing studies resulted in identifying the location of a crevasse channel which was created still before the construction of the flood embankment. In order to examine the lithological profile and confirm the existence of anomalies in the geological structure (erosion gutter filed with soil well conductive to groundwater), two boreholes of 6 m depth below ground level were drilled within this form and in its vicinity. The boreholes were made with a hand drill set. Piezometers made of PVC tubes with a diameter of 50 mm were installed in the boreholes. The screen was installed at a depth of 3.5–4.0 m below ground level. Electronic data recorders (Solinst LTC M 10 model 3001 Georgetown, ON, Canada) of water head level (equipotential), electrolytic conductivity and temperature were installed in the piezometers (Figure 6). Two loggers were installed in the zone where the most visual erosional channel is visible in the remote-sensing image (L2-Figure 6), and in the place where the area has not been transformed (L1-Figure 6).

Starting from 6 p.m. on 17 May 2019, the recorders saved the data with a frequency of every 15 min. This time was chosen due to the hydrological situation. It was decided to install loggers while the flood wave was accumulating in the mountainous areas before it reached the lowlands, where the research area is located. This gave a picture of the entire flood episode from the low level, through the passage through the area of two floods, to the return to low levels. A temporary stream gauge was installed in the vicinity of the flood embankment, from the river side. The elevation values of the piezometers and the “0” value of the stream gauge were determined by means of a GPS-RTK receiver. The sensors remained in the holes until 12 June 2019. During that time, they recorded changes in the water table level during 2 surges with overbank flow. During the second one, water rested against the body of the flood embankment at a level of approximately 1 m. The peaks of both surges were recorded based on the readings of the temporary stream gauge. The values of the hydraulic head output by sensors installed in the piezometers were corrected using an atmospheric pressure recorder (Solinst, Georgetown, ON, Canada), installed at a distance of approximately 10 km. The tests were repeated between 24 June and 6 August 2020. The results of measurements were copied to a spreadsheet and compiled. This enabled a search for correlation between the impact of changes in filtration dynamics during surges on the individual recorded parameters.

3. Results

In the case of a point located in the crevasse trough (Figure 7A), there is a distinct relationship between the water level and electrolytic conductivity. Two surge episodes were recorded during the first measurement session. In the case of both the electrolytic conductivity, as well as the water level in the piezometer, there was a pronounced noticeable delay in relation to changes in the water level in the channel. This delay reaches about 50–80 h.

The water of the Vistula River has a higher total dissolved solids level in relation to the groundwaters of the first aquifer in the region of Magnuszew [24]. Its presence documented in piezometer A by the values of conductance becomes noticeable once more than 55 h have passed. On the other hand, for a site in which a piezometer is placed in the zone of an untransformed terrace, with the screen situated under a layer of poorly permeable fluvial muds (under artesian conditions), the reaction to a passing surge is virtually immediate (Figure 7B). The maximum value of hydraulic head is linked in time with the maximum water level in the river. On the other hand, no direct correlation with electrolytic conductivity is observed.

On 23 May 2019, an anomaly was recorded in the piezometer in the untransformed zone, involving a rapid increase in the temperature and water level, as well as a rapid drop in the electrolytic conductivity value, virtually down to “0”. This can be linked with the heavy precipitation which occurred on that day in the analyzed area. Rainwater with a considerably higher temperature and a low TDS value entered the piezometer via an opening in which the string of the recorder was placed. The mixing of rainwater and groundwater resulted in an increase in temperature and caused the value of conductance to exceed the calibratable limit for the given type of recorder, which is 50 μS/cm.

Due to the lack of an identified impact of hydrodynamics during surge episodes on the temperature of groundwaters, this parameter was excluded to avoid further complications. The studies performed in the following year (2020) allowed us to draw the conclusion that the time difference between the maximum piezometric water pressure in the aquifer and that in the channel is a permanent trend. During the surge recorded in 2020, the value in the zone of the untransformed terrace was once again related to the value in the channel with no delay, while in the case of the piezometer located in the zone of the crevasse, the delay once again reached 50–80 h (Figure 8).

The dynamics of the very process of changes in water level in the piezometer are also noteworthy. The dynamics of changes in piezometric pressure are higher in the zone of the untransformed terrace. The pressure grows much faster and then drops once the peak of the surge has passed. On the other hand, in the case of a crevasse trough, the curve has a much gentler slope. A relatively high level is maintained for a long time after the withdrawal of water from under the foot of the flood embankment.

When analyzing the electrolytic conductivity, upon omitting the deviation resulting from the flooding of the piezometer with rainwater as described above, an almost twofold increase is observed only in the piezometer located in the crevasse. This may indicate that river water has reached the recorder (Figure 9). The beginning of the increase in measurement was approximately 22 May 2019 and 31 May 2019 (Figure 9A), and 30 June 2020 (Figure 9B) constitutes evidence of the passing of a river water propagation front in the aquifer.

4. Discussion

The issue of interaction between the water level in an embanked river and the dynamics of filtration in an alluvial stratum addressed by numerous papers [39,40] is particularly noticeable during surges. This results in a change not just in the flow rate of groundwaters, but also in the direction of filtration. This also results in the ingression of river water into the aquiferous structures of the floodplain. As pointed out by Shankar et al. [41], such a phenomenon can also be induced artificially, via intense operation of riverbank water abstraction points (RBF). In numerous cases, in areas with a post-glacial origin of sediments filling river valleys, the presence of a stratum of surge, poorly permeable sediments characterized by low electrolytic conductivity compared to the neighboring sediments is a decisive element for the nature of filtration in alluvia. Numerous hydrodynamic analyses [42,43] assume the continuity of geological strata along the course of the river. The studies of the ingression of river water into the area of the floodplain in most cases also assume the continuity of alluvial strata over the entire analyzed area [44]. However, the presence of specific hydrogeological windows within poorly permeable sediments seems to be quite common [19]. Their presence is usually pronounced in the supply system of riverbank abstraction points. As indicated by numerous publications, they constitute pathways for the privileged filtration of groundwaters [45,46,47].

The origin of such forms seems to be for others authors debatable. Goldschneider et al. [48] interpret the hydrogeological windows filled with sandy sediments and gravel as elevations of older sediments, the top of which reaches above the poorly permeable series. On the other hand, the authors’ previous research proves that they can be erosional troughs, secondarily filled with sediment, which among other things are the result of evolution of the environment in the Holocene [20]. This assumption is confirmed by the possibility of establishing the location and range of such forms based on remote-sensing criteria (19). The presence of these forms can be linked with the specific structure of the substrate of alluvia, and in particular with the presence of steps, sub-alluvial bedrock protrusions with cobbles on the surface, which are made of a material highly resistant to being washed out [17,18,20]. In these zones, the direction of the mainstream during surges depends on the morphology of the bedrock protrusion, and it is repeatedly directed towards the same zones. Such a flow feature cause floodplain transformation.

The studies described in this paper prove the specificity of conditions of supply and drainage in the zone of the specific hydrogeological window, as well as in the zone of the untransformed terrace. In the case of the untransformed zone, the relationship between waters saturating the aquifer and waters in the channel takes on features characteristic of communicating vessels. The propagation of pressure occurs virtually immediately, while the physical flow of water is limited by the “resilience” of the confined alluvial stratum. This is particularly significant during a surge, limiting the diversity of filtration directions in the substratum of the embankments. On the other hand, in the zone of the erosional trough, filtration occurs under the conditions of an unconfined water table, the pressure change front corresponding to the physical flow of water. It was originally assumed that the increase in the electrolytic conductivity value in the piezometer in the crevasse was related to its flooding with surface water and with elevated organic matter content. On the other hand, this was verified during the photographic documentation of the studies. On 28 May 2019, the piezometer was situated over a meter underwater, which proves that changes in the electrolytic conductivity result from the inflow of river water via an underground path. This constitutes evidence of the cyclicality of changes in the directions of filtration in the substratum of the embankments during almost each surge episode with a magnitude allowing the water to rest against the body of the flood embankment.

Among other things, the presented diversity of the geological structure of the floodplain results from the evolution of the fluvial environment occurring in the most recent part of the Holocene. This should be linked with changes in the climate and in land development of the catchment area, resulting among other things from anthropopressure, as widely discussed in the literature [49]. Poorly permeable surge sediments, formed in the period of the climatic optimum of the Holocene, originally covered the entire floodplain area with their compact series. Currently, due to an increase in the frequency and amplitude of surges, episodes of dynamic flows are becoming more frequent. During these episodes, the reworking of channel sediments reaches deeper, accentuating the impact of the erosional base level. Flows concentrated in a specific zone cut through the originally deposited cohesive sediments, forming erosional troughs, which are then filled with subsequent series of loose sediments, contributing to the formation of the suggested hydrogeological windows.

The slight increase in the electrolytic conductivity value precisely during a peak in the surge in the piezometer located in the zone of the untransformed terrace requires further research. The relatively flat nature of the line raises a question regarding whether this increase could be caused by the dissolution of contaminants in the piezometer column during the ascension of the water level, or whether these changes can result from dispersion, diffusion or another hydrochemical process occurring in the aquifer.

During work related to the construction and maintenance of the flood protection structure, the sediments filling erosional troughs, characterized by low consolidation, are able to, in combination with their high hydraulic conductivity, cause regional changes in the geotechnical parameters of the substrate, also affecting the stability of the structures founded thereon. This may influence a local increase in the risk of the occurrence of a malfunction [19].

In times of sustainable development, in particular in densely populated areas, it is becoming increasingly important to make land available for agriculture. Due to the presence of fertile soils in river valleys, such areas are increasingly accessible around the world by securing them with levee systems. The presented works may reduce the cost of geotechnical exploration of the building base of embankments. The applied methodology may allow a reduction in the number of expensive geological drillings and their locations in precise zones. Moreover, because of the conductivity, presented geological forms constitute places for the convenient location of riverbank water abstraction points of the RBF type [41]. The discussed research results, and the conclusions presented below, on the one hand, facilitate agricultural and construction activities, and on the other hand, allow for the identification of functions supporting the availability of water for the population. Therefore, this study contributes to the presentation of the complementary activities, enabling the additional development of rural areas. The dynamic development of remote-sensing techniques has also been revealed in recent years in research related to geology and hydrogeology [50]. In this area, it is particularly important to properly process the data in order to visualize discontinuities significant in a given discipline. Appropriate data extraction with the use of developed procedures allows for the preparation of thematic maps. Such maps can then be useful in areas of crisis management, environmental engineering or civil engineering [51]. The methodology of scientific research assumes that the results obtained should be repeatable and objectively verified. The presented methodology is an example of a good practice aiming at the development of the procedures not only for the assessment of remote-sensing materials, but also for the verification of the results obtained with this aid. Wider application of such an approach increases the level of technological readiness of specific solutions and may accelerate their implementation in design and engineering practices.

5. Summary and Conclusions

The performed studies allowed for the verification of the impact of the presence of filled erosional troughs in the proximal part of the floodplain on groundwater dynamics. In particular, it has been concluded that, in the zone of the untransformed terrace, the waters flow under conditions of elevated pressure, and the manner of flow is characteristic for communicating vessels. The increase in pressure initiated in the zone between embankments propagates virtually immediately along considerable distances, while the physical flow itself is relatively slow, and it requires breaking the pressure barrier. On the other hand, in the area of the erosional trough, the flow occurs under unconfined conditions, and its rate depends in particular on the very filtration parameters of the aquifer. Additionally, the presented results of hydrogeological research confirmed that it is possible to find erosive forms filled with sediment with the use of simple remote-sensing methods. On this basis, the following detailed conclusions have been formulated:

• In the zone of the untransformed terrace, the ascension of the water level between embankments causes immediate propagation of pressure in the aquifer, while the filtration process itself is considerably limited.
• Crevasse troughs constitute paths of privileged filtration, in particular in the proximal part of the floodplain, which confirms the results of the modelling studies performed earlier.
• The appearance of water with elevated conductance in the area of the crevasse proves the cyclicality of changes in flow directions, depending on the water level between embankments.
• Due to the cyclicality of changes in the directions of filtration, the natural process of soil clogging in the zone of contact between groundwater and surface waters within the range of the occurrence of crevasse troughs is considerably limited.
• The secondarily filled erosional troughs constitute pathways of intense supply for the alluvial aquifer, forming zones with preferable conditions for the location of riverbank water abstraction points of the RBF type.
• It has been confirmed that it is possible to identify such zones with the use of remote-sensing techniques, in particular LIDAR laser scanning and high-resolution satellite images.
• The research results could be of substantial importance when determining land development conditions for the floodplain. The identification of the location of the zones with preferential groundwater flow in the buried erosion channel along the flood embankment was possible based on the results, revealing increased conductivity. This made it easier to delineate the zones where the subsoil deformation initiated and to indicate the locations of favorable conditions for RBF groundwater extraction. The changes in outflow conditions, caused by climate change, provide additional justification for the careful determination of the effects of subsurface features on groundwater dynamics in the floodplain.

As part of the future work, it is planned to develop a remote-sensing key. The developed key will allow for the creation of a procedure for identifying zones with different hydrogeological parameters within the river valley. This procedure can reduce the costs of primary geotechnical exploration of the subsoil by reducing the number of drillings and placing them precisely in the zones of potential geological changes of the subsoil.