In\u00a0the\u00a0first (review) part of\u00a0the\u00a0paper, the\u00a0articles were divided into several groups according to whether they were based on investigations of\u00a0groundwater level (GWL) fluctuations, monitoring of\u00a0wetland-influenced streamflow, tree transpiration measurements, or a combination of\u00a0different methods. Thus, we can see where current research has moved since the\u00a0original observations of\u00a0GWL fluctuations, for example in\u00a0White\u2019s 1932 paper\u00a0[1]. In\u00a0the\u00a0second (practical) part of\u00a0the\u00a0paper, GWL fluctuations were monitored in\u00a0a wetland in\u00a0the\u00a0upper part of\u00a0the\u00a0Lib\u011bchovka catchment (moderate climate, western part of\u00a0the\u00a0Bohemian Cretaceous Basin) in\u00a0the\u00a0summer of\u00a02024. Four piezometers representing different parts of\u00a0the\u00a0wetland were installed in\u00a0the\u00a0wetland. From the\u00a0measured data, periods in\u00a0which significant diurnal GWL fluctuations occurred for several days were selected; these were 8 periods of\u00a03 to 14 days. Fluctuations were evident in\u00a0all parts of\u00a0the\u00a0wetland surveyed, with GWL maxima and minima occurring at similar times in\u00a0different parts of\u00a0the\u00a0wetland; only the\u00a0amplitudes of\u00a0the\u00a0fluctuations differed. Diurnal GWL fluctuation was most evident in\u00a0the\u00a0central part of\u00a0the\u00a0wetland (amplitude up to 14.5 cm in\u00a0peak summer). As additional information on wetland conditions, soil moisture was measured at different depths in\u00a0autumn and summer. It was observed that the\u00a0soil profile changes between different locations, even several meters apart, due to the\u00a0dynamic action of\u00a0the\u00a0flowing stream. The\u00a0moisture content of\u00a0sandy layers (around 40\u00a0%) differed significantly from that of\u00a0clay-loam layers (where it was mostly between 70 and 80\u00a0%). Soil in\u00a0the\u00a0wetland was also found to be very close to saturated throughout the\u00a0profile, both in\u00a0autumn, when no significant ET takes place, and during summer. This means that the\u00a0small amount of\u00a0water added\/consumed is sufficient to cause significant vertical movement of\u00a0GWL and therefore diurnal variation of\u00a0GWL may be more apparent. For the\u00a08 selected periods in\u00a0which significant diurnal GWL fluctuations occurred for several days, the\u00a0average ET of\u00a0the\u00a0monitored wetland in\u00a0the\u00a0upper catchment of\u00a0the\u00a0Lib\u011bchovka stream was calculated as 20\u00a0l\u00a0\u2219\u00a0s-1<\/sup>\u00a0\u2219\u00a0km-2<\/sup> by White\u2019s method\u00a0[1] based on GWL fluctuation. The\u00a0two parts of\u00a0the\u00a0paper together show that the\u00a0topic of\u00a0wetland ET is important and relevant and demonstrate that wetlands need to be seen as environments where water is intensively used by vegetation.<\/span><\/p>\n Evapotranspiration (ET) is generally an important component of\u00a0the\u00a0water cycle. A wetland, as a waterlogged area with a groundwater level (GWL) close to the\u00a0surface, provides an environment where water is readily available to vegetation. Actual ET is therefore very high and approaches potential evapotranspiration (PET), which is the\u00a0maximum possible rate limited only by the\u00a0amount of\u00a0energy available for evaporation.<\/span><\/p>\n While primary ET occurs in\u00a0the\u00a0soil, in\u00a0wetlands, emerging groundwater or incoming surface water evaporates that has already undergone primary ET. The\u00a0influence of\u00a0wetlands on GWL and streamflow is therefore particularly significant during dry periods, when there is a general lack of\u00a0water in\u00a0the\u00a0landscape, but water remains available in\u00a0the\u00a0wetland due to groundwater inflow, allowing intensive ET to continue.<\/span><\/p>\n It is therefore necessary to consider the\u00a0influence of\u00a0ET not only in\u00a0planning restoration projects, calculating water balance, and in\u00a0hydrological models, but it should also be considered in\u00a0legislation \u2013 for example, in\u00a0relation to the\u00a0concept of\u00a0minimum residual flow or minimum groundwater level. According to current legislation, artificial water abstraction is restricted during dry periods in\u00a0order to maintain\u00a0the\u00a0minimum residual flow. However, under high temperatures, water consumption through ET may become so high that it will not be possible to maintain\u00a0the\u00a0minimum residual flow (or minimum GWL), even if all other artificial water abstraction is halted, because wetland and riparian vegetation is capable of\u00a0evaporating groundwater near the\u00a0watercourse\u00a0[2]. This issue is becoming particularly relevant in\u00a0the\u00a0context of\u00a0ongoing climate change, which is leading to rising temperatures and, consequently, an even more pronounced influence of\u00a0ET.<\/span><\/p>\n The\u00a0aim of\u00a0this paper is to summarise the\u00a0main\u00a0research conducted to date in\u00a0the\u00a0field of\u00a0wetland ET and to demonstrate that wetland ET is an important topic that must be taken into account in\u00a0Central Europe, along with the\u00a0latest findings. The\u00a0first, theoretical part presents a review of\u00a0articles focused on the\u00a0influence of\u00a0ET in\u00a0wetlands. The\u00a0second part provides a practical demonstration by monitoring the\u00a0influence of\u00a0ET on fluctuations in\u00a0GWL and streamflow of\u00a0a small watercourse in\u00a0a wetland in\u00a0the\u00a0upper catchment of\u00a0the\u00a0Lib\u011bchovka.<\/span><\/p>\n Wetland ET was intensively studied as early as the last century, yet it remains a current and actively researched topic today. Published articles can be divided into several groups, which are discussed in more detail in the following paragraphs. The first group of articles examines in detail the influence of ET on GWL fluctuations. Another group focuses on the relationship between ET and streamflow. A different group combines observed GWL fluctuations with fluctuations in the flow of a stream running through the wetland. A further\u00a0<\/span>group of\u00a0articles addresses the\u00a0issue in\u00a0a highly comprehensive manner, linking the\u00a0influence of\u00a0ET, GWL fluctuations, and streamflow observations with measurements of\u00a0vegetation transpiration.<\/span><\/p>\n The\u00a0relationship between wetland ET and GWL fluctuations was studied by White as early as 1932\u00a0[1]. He observed that, due to ET, GWL is lower during the\u00a0day than at night. This results in\u00a0a regular within-day GWL fluctuation (diurnal fluctuation). A typical pattern of\u00a0diurnal fluctuation is well described, for example, in\u00a0articles by Gribovszki from 2006\u00a0[3] and 2008\u00a0[4], and is illustrated in\u00a0Fig.\u00a01<\/em>.<\/span><\/span><\/p>\n White\u00a0[1] developed a method to estimate the\u00a0magnitude of\u00a0ET on a daily scale based on diurnal fluctuations. The\u00a0principle of\u00a0the\u00a0method is illustrated in\u00a0Fig.\u00a02<\/span><\/em>. The\u00a0vertical axis shows the\u00a0GWL, and the\u00a0horizontal axis represents time. For a given day, Grec presents a linear extrapolation of\u00a0expected GWL development in\u00a0the\u00a0absence of\u00a0ET-induced decline. The\u00a0extrapolation is based on water level development during the\u00a0night between midnight and 4 a.m., as ET can be considered negligible during this part of\u00a0the\u00a0day. \u2206S represents the\u00a0change in\u00a0groundwater storage during the\u00a0day (i.e., the\u00a0change in\u00a0water level between midnight of\u00a0the\u00a0current and previous day). Subsequently, ET is calculated as the\u00a0sum of\u00a0Grec and \u2206S, multiplied by the\u00a0specific yield value.<\/span><\/p>\n The\u00a0limitations and behaviour of\u00a0White\u2019s method were studied in\u00a0detail, for example, using numerical simulations in\u00a0Loheide\u2019s 2005 article\u00a0[6]. The\u00a0most significant source of\u00a0error was identified as uncertainty in\u00a0determining specific yield. A specific yield of\u00a02\u00a0% means that if the\u00a0water level in\u00a0rock environment drops by 1\u00a0m, the\u00a0amount of\u00a0water removed from the\u00a0rock environment is equivalent to removing a 2cm water column from a water volume (e.g., a water tank).<\/span><\/p>\n Some authors have proposed alternative methods for estimating ET from GWL fluctuations. In\u00a02008, Gribovszki\u00a0[4] developed a modification of\u00a0White\u2019s original method to improve the\u00a0accuracy of\u00a0the\u00a0calculation. Another method was offered by Loheide in\u00a0an article also published in\u00a02008\u00a0[7]. Carlson Mazur, in\u00a02014\u00a0[5], subsequently created a modification allowing its application across a wider range of\u00a0natural conditions. Various methods for calculating ET from GWL fluctuations are discussed in\u00a0the\u00a0article by Fahle and Dietrich from 2014\u00a0[8]. The\u00a0values were compared against reference measurements using a lysimeter; the\u00a0highest correlation with the\u00a0reference value was obtained using Gribovszki\u2019s method\u00a0[4].<\/span><\/p>\n A significant advance was also made by Malama in\u00a02010\u00a0[9]. He derived an analytical solution describing how ET (specified as an input \u201ccontrol function\u201d) manifests in\u00a0GWL fluctuations (Fig.\u00a0<\/em>3<\/span>). Malama\u2019s solution\u00a0[9] also explains the\u00a0several-hour delay of\u00a0GWL maxima and minima relative to the\u00a0solar cycle, which had previously been observed in\u00a0the\u00a0field, for example, by Gribovszki\u00a0[4]. The\u00a0derived analytical solution can be applied in\u00a0two ways: to model ET from measured diurnal GWL fluctuations given knowledge of\u00a0the\u00a0hydraulic parameters, or, inversely, to infer hydraulic parameters of\u00a0the\u00a0environment from diurnal GWL fluctuations given knowledge of\u00a0ET. In\u00a0Malama\u2019s article, the\u00a0developed model was applied to measured diurnal GWL fluctuations, yielding information on ET, hydraulic conductivity, and river stage changes (Fig.\u00a03<\/span><\/em>). This demonstrated that, in\u00a0principle, both ET and hydraulic parameters can be derived simultaneously from diurnal GWL fluctuations.<\/span><\/p>\n The\u00a0following group of\u00a0articles focuses directly on the\u00a0relationship between ET and streamflow fluctuations. Sometimes this relationship was assessed on a regional scale; an example is 1976 Daniel\u2019s article\u00a0[10]. This paper addressed the\u00a0relationship between ET from the\u00a0aquifer and streamflow in\u00a0a nearby watercourse, and described an analytical solution that can be used in\u00a0rainfall-runoff models and GWL models. Another example is Wittenberg\u2019s 1999 article\u00a0[11]. It\u00a0determined hydraulic parameters from curves describing the\u00a0long-term annual decline in\u00a0streamflow, while observing how recession curves are influenced by\u00a0ET.<\/span><\/p>\n Other articles approached the\u00a0relationship between ET and streamflow at the\u00a0catchment scale. An example is Zecharias\u00a0[12] in\u00a01988 which used a conceptual model to describe the\u00a0relationship between runoff and aquifer. A significant advance was later made by Fonley\u00a0[13], who in\u00a02019 derived an analytical relationship for back-calculating ET from records of\u00a0diurnal streamflow fluctuations (Fig.\u00a04<\/span><\/em>).<\/span><\/p>\n Another group of\u00a0articles addresses GWL fluctuations caused by ET, while also including observations of\u00a0the\u00a0stream flowing through the\u00a0wetland. Partially, the\u00a0previously mentioned article by Gribovszki\u00a0[4] could be included here, whose main\u00a0contribution was improving methods for calculating ET from GWL fluctuations, but also addressing the\u00a0relationship with stream level fluctuations. Another example is Yeh\u00a0[14], whose 2008 article examined the\u00a0regional-scale soil water balance using 19 years of\u00a0monthly runoff observations, GWL, and soil moisture. In\u00a0addition, changes were monitored in\u00a0the\u00a0thickness of\u00a0the\u00a0subsurface layer influenced by ET.<\/p>\n The\u00a0relationship between soil moisture and streamflow fluctuations was examined in\u00a0detail by Moore\u2019s 2011 study\u00a0[15], conducted in\u00a0the\u00a0HJ Andrews experimental forest in\u00a0western Oregon. He reached the\u00a0interesting conclusion that, across all time scales, soil moisture correlates very well with the\u00a0amount of\u00a0water currently flowing in\u00a0the\u00a0stream. The\u00a0correlation was strongest at high soil moisture levels.<\/p>\n The\u00a0most comprehensive studies of\u00a0wetland ET are found in\u00a0articles that monitor not only streamflow and possibly GWL fluctuations but also independently measure vegetation transpiration, for example through sap flow measurements in\u00a0trees. These articles can be divided into two groups based on location. One group originates from the\u00a0semi-arid climate of\u00a0Arizona, the\u00a0other from the\u00a0Mediterranean climate of\u00a0Oregon.<\/p>\n The\u00a0first group consists of\u00a0articles from Arizona. They form part of\u00a0the\u00a0extensive international SALSA (Semi-Arid Land\u2013Surface\u2013Atmosphere) programme, which focused on human-induced environmental changes in\u00a0semi-arid regions\u00a0[16]. ET from waterlogged areas along watercourses (riparian zones) represents an important component of\u00a0the\u00a0water balance and was therefore intensively studied within\u00a0the\u00a0project using various methods\u00a0[17]. For example, transpiration of\u00a0willows and poplars was measured using the\u00a0sap flow method, and the\u00a0results were compared with ET values indirectly derived from the\u00a0water balance\u00a0[18]. Canopy transpiration derived from sap flow measurements corresponded to values obtained using Raman lidar and was used to calibrate coefficients in\u00a0the\u00a0Penman\u2013Monteith method for calculating ET. ET of\u00a0grasses was also determined using the\u00a0Bowen ratio method. A summary of\u00a0these results, together with an estimate of\u00a0the\u00a0uncertainty in\u00a0determining water balance components and a comparison with values obtained using a model, is discussed in\u00a0detail in\u00a0the\u00a0paper by Goodrich (2000)\u00a0[17].<\/p>\n The\u00a0second group comprises studies from Oregon (HJ Andrews experimental forest). Important insights are provided in\u00a0the\u00a0paper by Bond (2002)\u00a0[19], where streamflow was measured and correlated with sap flow data from the\u00a0surrounding vegetation. The\u00a0measurements were used to estimate the\u00a0volume of\u00a0water consumed by ET and to determine the\u00a0width of\u00a0the\u00a0riparian zone from which ET occurred. The\u00a0study also described in\u00a0detail the\u00a0times of\u00a0year when diurnal fluctuations in\u00a0streamflow caused by ET were observed.<\/p>\n An interesting contribution from Oregon is the\u00a02007 paper by Wondzell\u00a0[20], which examined in\u00a0detail the\u00a0delay between the\u00a0daily ET cycle and streamflow fluctuations.\u00a0 It was found that the\u00a0delay depends on streamflow velocity, more precisely on the\u00a0speed at which the\u00a0hydraulic pulse propagates through the\u00a0channel. When streamflow was fast, the\u00a0influence of\u00a0vegetation from different parts of\u00a0the\u00a0wetland combined constructively, amplifying the\u00a0original diurnal signal. In\u00a0contrast, when streamflow was slow, the\u00a0influence of\u00a0vegetation from different parts of\u00a0the\u00a0wetland combined destructively (i.e. cancelled each other out), and the\u00a0original diurnal signal was dampened. Wondzell\u00a0[20] also mentions diurnal fluctuations in\u00a0chemical indicators and suggests that they could be subject to a similar delay depending on streamflow velocity.<\/p>\n Another contribution from Oregon is the\u00a0article by Barnard from 2010\u00a0[21]. The\u00a0aim was to describe the\u00a0relationship between transpiration and subsurface runoff. To this end, subsurface runoff from the\u00a0soil, soil moisture, and tree transpiration were measured on a selected small plot. Artificial irrigation was also carried out, and the\u00a0resulting development of\u00a0changes was monitored. It was found that the\u00a0delay of\u00a0diurnal fluctuation relative to ET depends on soil moisture. In\u00a0this case, however, Barnard\u00a0[21] considers it unlikely that the\u00a0signal from different parts of\u00a0the\u00a0catchment could synchronise in\u00a0such a way that diurnal fluctuations would be observable.<\/p>\n An interesting article from Oregon is also that of\u00a0Graham from 2013\u00a0[22], discussing the\u00a0frequency of\u00a0diurnal fluctuations caused by ET. Fifteen different catchments were monitored, and the\u00a0results were compared with sap flow measurements. Diurnal fluctuations in\u00a0streamflow caused by ET were observed in\u00a0all years and across all fifteen catchments, suggesting that this is not merely a special phenomenon limited to a small number of\u00a0sites.<\/p>\n ET in wetland environments remains a highly researched topic today, leading to significant advances in understanding over the past 20 years. GWL diurnal fluctuations caused by ET have been shown to occur across many catchments, as demonstrated by Graham in 2013 [22].\u00a0 Modifications of the original classic White method have been developed for calculating ET from GWL fluctuations, such as Gribovszki\u2019s method from 2008 [4]. Since the original observations and descriptions, research has progressed to modelling and deriving analytical relationships. An example in the relationship between ET and GWL fluctuations is the article by Malama from 2010 [9], and for the relationship between ET and streamflow fluctuations, the article published by Fonley in 2019 [13]. In the literature, there are also comprehensive studies combining observations of diurnal\u00a0fluctuations with independent ET determinations by other methods, for example through sap flow measurements in\u00a0trees. The\u00a0main\u00a0groups of\u00a0articles adopting such a comprehensive approach come from the\u00a0semi-arid climate of\u00a0Arizona and the\u00a0Mediterranean climate of\u00a0Oregon.<\/p>\n Generally, the\u00a0current state can be summarised in\u00a0the\u00a0following three points. Firstly, since the\u00a0original observations of\u00a0changes in\u00a0water level or flow in\u00a0wetlands, there has been a tremendous advance in\u00a0knowledge. Secondly, despite this great progress, it remains necessary to conduct measurements in\u00a0various locations. These measurements, employing ever-evolving technology, will help verify or supplement hydrological models\u00a0[13]. The\u00a0third and very important point is the\u00a0need to transfer these already acquired insights into the\u00a0general awareness of\u00a0experts engaged in\u00a0practical activities related to hydrology and ecology, who can then apply the\u00a0new knowledge in\u00a0everyday practice. This is an important step building on the\u00a0previous research, which is essential to ensure that current findings are considered in\u00a0everyday practice. An example of\u00a0such an effort is this article on wetland ET, which combines a literature review with measurements of\u00a0GWL fluctuations in\u00a0a long-term studied wetland in\u00a0the\u00a0upper Lib\u011bchovka catchment.<\/p>\n Following the\u00a0literature review, a practical demonstration of\u00a0the\u00a0influence of\u00a0ET on a wetland in\u00a0the\u00a0western Czech Cretaceous Basin\u00a0was carried out. GWL fluctuations and runoff from a small wetland were monitored on a minor tributary in\u00a0the\u00a0upper catchment of\u00a0the\u00a0Lib\u011bchovka. It was demonstrated that even under these conditions, diurnal GWL fluctuations caused by ET can be observed. The\u00a0measurements also provide information on what can be easily used for such monitoring, which conditions must be met, and how this diurnal fluctuation appears in\u00a0practice.<\/p>\n The\u00a0measurements were conducted in\u00a0a wetland in\u00a0the\u00a0upper part of\u00a0the\u00a0Lib\u011bchovka catchment (Fig.\u00a05). It is part of\u00a0hydrological region 4522\u00a0Cretaceous of\u00a0the\u00a0Lib\u011bchovka and P\u0161ovka. Average annual temperature (calculated based on the\u00a0period 1981\u20132010) reaches 8.2 \u00b0C, and average annual precipitation totals 595\u00a0mm\u00a0[23].<\/p>\n The\u00a0site was chosen so that local conditions would not hinder the\u00a0manifestation of\u00a0ET. A small watercourse (Sov\u00ed stream) flows through the\u00a0wetland, whose main\u00a0source of\u00a0water bearing is groundwater from quaternary sandstones; rapid runoff is negligible. Therefore, the\u00a0watercourse has a stable flow, allowing the\u00a0impact of\u00a0water consumption by ET to be clearly observed. The\u00a0area of\u00a0the\u00a0wetland was determined through field survey to be 19,000\u00a0m2<\/sup>\u00a0[24].<\/p>\n As a practical demonstration of\u00a0the\u00a0influence of\u00a0ET in\u00a0the\u00a0wetland, GWL and soil moisture were monitored. GWL was measured using piezometers, and soil moisture was determined by collecting soil samples.<\/span><\/p>\n GWL measurements<\/span><\/strong><\/p>\n This current phase followed up on earlier measurements by the\u00a0same authors\u00a0[25] from the\u00a0second half of\u00a0summer 2022, which demonstrated that diurnal GWL fluctuations in\u00a0the\u00a0monitored wetland are caused by ET and occur on warm, rain-free days. The\u00a0new study covered the\u00a0entire summer period, with piezometers installed in\u00a0different parts of\u00a0the\u00a0wetland, enabling observation of\u00a0the\u00a0gradual development of\u00a0diurnal GWL fluctuations from the\u00a0spring and summer of\u00a02024.<\/span><\/p>\n Measurements from the\u00a0beginning of\u00a02024 until 9\u00a0September 2024 were used for the\u00a0analysis. The\u00a0location of\u00a0the\u00a0newly installed piezometers, which capture GWL spatial variability in\u00a0the\u00a0wetland, is shown in\u00a0Fig.\u00a06<\/span>. The\u00a0piezometer referred to as LI2\u00a0monitors the\u00a0conditions in\u00a0the\u00a0middle of\u00a0the\u00a0lower part of\u00a0the\u00a0wetland, close to the\u00a0constructed weir. The\u00a0central area of\u00a0the\u00a0wetland is represented by the\u00a0piezometer called St\u0159edn\u00ed mok\u0159ad (\u201cCentral Wetland\u201d). The\u00a0piezometer called Bo\u010dn\u00ed mok\u0159ad (\u201cLateral Wetland\u201d) describes the\u00a0edge of\u00a0the\u00a0wetland, where the\u00a0terrain\u00a0is 2\u00a0metres above the\u00a0valley floor. However, based on the\u00a0vegetation characteristics, this area is still considered a wetland with GWL close to the\u00a0surface, which indicates that groundwater flows into the\u00a0wetland from this side. This is also consistent with the\u00a0clay layer found at a depth of\u00a0about 60\u00a0cm below the\u00a0surface, which may form a low-permeability layer along which water flows. Another inflow area to the\u00a0wetland is represented by the\u00a0piezometer designated U Stud\u00e1nky (\u201cThe\u00a0Spring\u201d). It was installed at the\u00a0upper edge of\u00a0the\u00a0wetland, near the\u00a0spring of\u00a0one of\u00a0the\u00a0branches of\u00a0the\u00a0watercourse flowing through the\u00a0wetland.<\/span><\/p>\n The\u00a0construction of\u00a0the\u00a0piezometers is described in\u00a0detail in\u00a0[24].They consisted of\u00a0pipes with a perforated lower section, in\u00a0which the\u00a0GWL was measured using Solinst Logger pressure sensors. To convert the\u00a0measured pressure to water column height, it was first necessary to subtract the\u00a0atmospheric pressure from the\u00a0measured values. Atmospheric pressure was monitored using a Solinst Logger pressure sensor located in\u00a0the\u00a0same area as the\u00a0wetland. For measurement verification, the\u00a0GWL was also manually measured with a water level meter at the\u00a0beginning and end of\u00a0the\u00a0measuring period. The\u00a0working names of\u00a0the\u00a0piezometers and the\u00a0depth of\u00a0their bases below the\u00a0surface are shown in\u00a0Tab.\u00a01<\/span><\/em>.<\/span><\/p>\n Soil moisture measurement<\/span><\/strong><\/p>\n Soil moisture was measured around all installed piezometers, thus covering different parts of\u00a0the\u00a0wetland. To avoid disturbance of\u00a0the\u00a0soil immediately surrounding the\u00a0piezometer tube, sampling points were selected approximately 1.5\u00a0m from each piezometer. The\u00a0first sampling took place First, a profile was excavated to a depth where water began to visibly seep from the\u00a0walls, and the\u00a0characteristics of\u00a0the\u00a0individual layers were roughly described. Subsequently, samples were taken. The\u00a0sampling depths were adjusted to ensure that prominent soil layers were captured. During the\u00a0second sampling, it was necessary to select a location near the\u00a0piezometer that had not been disturbed during the\u00a0previous sampling; therefore, the\u00a0profiles for the\u00a0same piezometer slightly differ between the\u00a0first and second samplings. Kopeck\u00fd cylinders with a volume of\u00a0100 cm\u00b3 were used for sampling, driven in\u00a0horizontally. Within\u00a0each profile, one sample was taken from each observed layer to determine specific yield. Immediately after sampling, the\u00a0cylinders were sealed with lids, wrapped in\u00a0foil, and placed in\u00a0plastic bags to prevent moisture loss. As soon as the\u00a0samples returned from the\u00a0field, they were weighed on scales with an accuracy of\u00a00.1 g. Subsequent weighing was conducted on the\u00a0same scales.<\/span><\/p>\n The\u00a0first step in\u00a0sample processing was to determine saturated moisture content. Cylinders containing the\u00a0samples were placed individually in\u00a0a container, filled with water from the\u00a0bottom, and left submerged for three days. They were then removed from the\u00a0container and weighed. In\u00a0two exceptional cases, a non-negligible loss of\u00a0material occurred from the\u00a0sample, which distorted the\u00a0moisture content in\u00a0the\u00a0saturated state (a note has been added in\u00a0the\u00a0resulting graphs for the\u00a0respective samples). The\u00a0second step was to measure the\u00a0weight of\u00a0the\u00a0dried soil. First, the\u00a0sample was dried for several days at room temperature, then for 1.5 weeks in\u00a0an oven at 105\u202f\u00b0C, and finally it was weighed immediately after removal from the\u00a0oven to prevent it from absorbing atmospheric moisture.<\/span><\/p>\n From the\u00a0measured data, the\u00a0volumetric moisture content of\u00a0the\u00a0sample at the\u00a0time of\u00a0collection (\u03b8odb<\/sub>) and the\u00a0saturated (volumetric) moisture content of\u00a0the\u00a0sample (\u03b8sat<\/sub>) were calculated as follows:<\/span><\/p>\n <\/p>\n where:<\/p>\n V\u00a0v\u00a0odb<\/sub> \u00a0\u00a0 is\u00a0\u00a0\u00a0\u00a0the\u00a0volume of\u00a0water in\u00a0the\u00a0sample at the\u00a0time of\u00a0collection<\/p>\n V\u00a0v\u00a0sat<\/sub>\u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 the\u00a0volume of\u00a0water in\u00a0the\u00a0sample in\u00a0the\u00a0saturated\u00a0state<\/p>\n V\u00a0vzorek<\/sub> \u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0the\u00a0volume of\u00a0the\u00a0collected sample (100\u202fcm\u00b3)<\/p>\n m\u00a0odb<\/sub> \u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0the\u00a0weight of\u00a0the\u00a0sample at the\u00a0time of\u00a0collection<\/p>\n m\u00a0s<\/sub>\u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 the\u00a0weight of\u00a0the\u00a0dried sample<\/p>\n m\u00a0sat<\/sub> \u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0the\u00a0weight of\u00a0the\u00a0sample in\u00a0the\u00a0saturated state<\/p>\n \u03c1\u00a0v<\/sub>\u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 the\u00a0density of\u00a0water (1,000\u202fkg \u00b7 m-3<\/sup>)<\/p>\n Subsequently, the\u00a0resulting moisture values were converted from decimal numbers to percentages.<\/p>\n Estimation of\u00a0ET from GWL fluctuations <\/span><\/strong><\/span><\/p>\n In\u00a0the\u00a0first step, the\u00a0specific yield of\u00a0the\u00a0soil layers in\u00a0which diurnal GWL fluctuations occur was estimated based on soil moisture measurements. This allowed the\u00a0measured GWL fluctuations to be converted into the\u00a0amount of\u00a0water gained or lost from the\u00a0groundwater during level changes in\u00a0subsequent data processing steps. For the\u00a0estimation of\u00a0specific yield, only those piezometers were selected where, during soil profile sampling, few heterogeneous layers were detected at depths around the\u00a0GWL.<\/span><\/p>\n The\u00a0calculation of\u00a0ET from diurnal GWL fluctuations was performed in\u00a0a simplified manner based on the\u00a0original classical method by White\u00a0[1]. The\u00a0principle of\u00a0the\u00a0method is summarised in\u00a0Fig.\u00a02<\/span><\/em> in\u00a0the\u00a0review section of\u00a0this article. First, the\u00a0average hourly change in\u00a0GWL (r) was calculated based on the\u00a0interval between midnight and 4 a.m. This value reflects the\u00a0long-term GWL trend without the\u00a0influence of\u00a0daytime ET. Subsequently, the\u00a0daily change in\u00a0soil water storage (\u2206Z) was calculated as:<\/span><\/p>\n where:<\/p>\n h1\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 is\u00a0 \u00a0 \u00a0 \u00a0the GWL at midnight of the current day<\/p>\n h2\u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 \u00a0 the GWL at midnight of the following day<\/p>\n A positive value of\u00a0\u2206Z indicates a drop in\u00a0GWL during the\u00a0day, meaning\u00a0a\u00a0loss of\u00a0water from the\u00a0environment.<\/p>\n In\u00a0the\u00a0next step, ET was calculated and expressed as the\u00a0height of\u00a0the\u00a0water column per day:<\/p>\n where:<\/p>\n S\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 is\u00a0\u00a0\u00a0\u00a0 the\u00a0specific yield<\/p>\n r\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 average change in\u00a0GWL without the\u00a0ET influence<\/p>\n \u2206Z\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0 zchange in\u00a0storage over a day<\/p>\n The\u00a0obtained ET value was converted from the\u00a0water column form to a flow rate form (volume of\u00a0water consumed per unit of\u00a0time). The\u00a0resulting wetland ET value was calculated as the\u00a0arithmetic mean of\u00a0the\u00a0individual values from each piezometer. This value was then compared with ET estimates based on streamflow fluctuations and the\u00a0Oudin\u00a0method for calculating PET, which had been conducted by the\u00a0same authors on earlier data from the\u00a0same site\u00a0[25].<\/p>\n The\u00a0average ET value calculated from the\u00a0piezometers using White\u2019s method was designated as ETpr\u016fm<\/sub>. However, some piezometers were excluded from the\u00a0ET calculation due to the\u00a0presence of\u00a0many different soil layers. Retrospective estimates were made for these piezometers to determine the\u00a0specific yield required for the\u00a0ET value calculated by the\u00a0White method from GWL fluctuation to equal ETpr\u016fm<\/sub>. This calculation was performed using the\u00a0Solver function in\u00a0MS Excel.<\/p>\n GWL measurement<\/span><\/strong><\/span><\/p>\n From the\u00a0measured data, periods were selected during which significant diurnal GWL fluctuations occurred over several days (Tab.\u00a02<\/span><\/em>). There were eight such periods, with the\u00a0duration of\u00a0individual episodes ranging from three to fourteen days, and fluctuations were evident at all measured locations within\u00a0the\u00a0wetland. During these periods, the\u00a0water level of\u00a0the\u00a0small watercourse at the\u00a0weir also fluctuated. However, the\u00a0amplitude of\u00a0stream level fluctuations was significantly smaller \u2013 up to a maximum of\u00a02.5 cm in\u00a0peak summer \u2013 compared to GWL fluctuations, which reached up to 14.5 cm in\u00a0peak summer at St\u0159edn\u00ed mok\u0159ad piezometer.<\/span><\/p>\n The\u00a0first significant signs of\u00a0diurnal GWL fluctuations were observed at the\u00a0beginning of\u00a0May (A in\u00a0Fig.\u00a07<\/span><\/em>). During the\u00a0eight-day period from 9 to 16\u00a0May, fluctuations were greater in\u00a0the\u00a0central part of\u00a0the\u00a0wetland (St\u0159edn\u00ed mok\u0159ad piezometer) compared to fluctuations at other locations within\u00a0the\u00a0wetland. Another period with clearly visible GWL fluctuations was the\u00a0first half of\u00a0June, specifically from 8 to 12\u00a0June (B in\u00a0Fig.\u00a07<\/span><\/em>) and from 16 to 18\u00a0June. The\u00a0amplitude of\u00a0fluctuations was more pronounced compared to May, especially in\u00a0the\u00a0case of\u00a0St\u0159edn\u00ed mok\u0159ad piezometer. An exception was U Stud\u00e1nky piezometer, where fluctuations were smaller compared to the\u00a0other piezometers.<\/span><\/p>\n From the\u00a0end of\u00a0July, marked diurnal fluctuations in\u00a0water level were observed in\u00a0all piezometers (C and D in\u00a0Fig.\u00a07<\/span><\/em>). These occurred during the\u00a0periods from 24 to 30\u00a0June, 7 to 10\u00a0July, 15 to 26\u00a0July, 28\u00a0July to 1\u00a0August, and 10 to 15\u00a0August. However, in\u00a0the\u00a0case of\u00a0the\u00a0U Stud\u00e1nky piezometer, the\u00a0amplitude of\u00a0the\u00a0fluctuations remained lower. The\u00a0longest continuous period of\u00a0diurnal GWL fluctuation occurred over 14 days at the\u00a0end of\u00a0summer, from 26\u00a0August to 8\u00a0September (E in\u00a0Fig.\u00a07<\/span><\/em>).<\/span><\/p>\n Soil moisture measurements<\/span><\/strong><\/span><\/p>\n The\u00a0soil moisture measurement results are presented in\u00a0Tab.\u00a03<\/span><\/em> and summarised in\u00a0Fig.\u00a08<\/span><\/em>. The\u00a0moisture content of\u00a0sandy layers during sampling was around 40\u00a0%, while the\u00a0moisture content of\u00a0clayey-silty layers was significantly higher, generally between 70 and 80\u00a0%.<\/span><\/p>\n Soil profiles sampled in\u00a0autumn and summer at the\u00a0same piezometer were only a few metres apart, yet the\u00a0layer composition often differed significantly. This indicates pronounced variation between soil profiles at different places within\u00a0the\u00a0wetland, which is a consequence of\u00a0the\u00a0watercourse dynamic activity. Simultaneously, this generally implies that it is difficult to determine representative hydraulic parameters for such environments, for example for the\u00a0purposes of\u00a0accurate modelling.<\/span><\/p>\n During autumn sampling, GWL was higher than in\u00a0summer. In\u00a0both periods, however, soil throughout the\u00a0entire profile was very close to saturation. The\u00a0difference between moisture content at the\u00a0time of\u00a0sampling and saturated moisture content was greatest in\u00a0the\u00a0surface layers, but even in\u00a0this case it did not exceed 4\u00a0%.<\/span><\/p>\n The\u00a0measurement error in\u00a0determining soil moisture was estimated to range between 1 and 2\u00a0%. This is based on the\u00a0fact that, in\u00a0some samples, the\u00a0measured moisture content at the\u00a0time of\u00a0sampling exceeded the\u00a0saturated moisture content, or that layers below the\u00a0GWL were not saturated. A difference value of\u00a00 between saturated and sampled moisture content indicates that the\u00a0sample\u2019s moisture content at the\u00a0time of\u00a0sampling was equal to or greater than the\u00a0saturated moisture content.<\/span><\/p>\n In\u00a0one case, it was found that a sample taken well below the\u00a0GWL was 5\u00a0% below saturation; in\u00a0another, the\u00a0moisture content of\u00a0the\u00a0sample significantly exceeded the\u00a0saturated moisture content (by 5\u00a0%). These values were considered to be processing errors caused by partial water loss from the\u00a0sandy sample during handling prior to weighing. In\u00a0the\u00a0case of\u00a0a third sample, it was found that just below the\u00a0GWL the\u00a0sample was 3\u00a0% below saturation. This was considered to be the\u00a0result of\u00a0GWL recorded from the\u00a0nearby piezometer not exactly matching the\u00a0GWL at the\u00a0location of\u00a0the\u00a0sampled profile, and the\u00a0respective sample was in\u00a0fact taken from above the\u00a0water level.<\/span><\/p>\n Estimation of\u00a0ET from GWL fluctuations<\/span><\/strong><\/p>\n The\u00a0first step was to determine specific yield of\u00a0the\u00a0soil layers at the\u00a0GWL depth. Summer moisture measurements were used, as diurnal GWL fluctuations were recorded during the\u00a0warm part of\u00a0the\u00a0year. From the\u00a0sampled profiles, the\u00a0Bo\u010dn\u00ed mok\u0159ad and U Stud\u00e1nky piezometers were selected for analysis, as the\u00a0soil composition at these sites showed little variation with depth. Samples were taken from layers above the\u00a0GWL, and specific yield was determined as\u00a0the\u00a0difference between the\u00a0saturated moisture content and the\u00a0moisture content at the\u00a0time of\u00a0sampling. Using this method, specific yield was estimated at 2\u00a0%, with an\u00a0absolute error of\u00a0\u00b11\u00a0%.<\/span><\/p>\n In\u00a0the\u00a0second step, White\u2019s method\u00a0[1] was applied to calculate ET based on the\u00a0measured diurnal fluctuations in\u00a0GWL at the\u00a0U Stud\u00e1nky and Bo\u010dn\u00ed mok\u0159ad piezometers. The\u00a0results are summarised in\u00a0Tab.\u00a04<\/span><\/em>. The\u00a0values presented are average ET calculated based on eight periods during which significant GWL fluctuations occurred over several days. The\u00a0final ET estimate was calculated as the\u00a0arithmetic mean of\u00a0the\u00a0values obtained from both piezometers and reached 20\u00a0l\u00a0\u2219\u00a0s-1<\/sup>\u00a0\u2219\u00a0km-2<\/sup>, with a range of\u00a0\u00b1 10\u00a0l\u00a0\u2219\u00a0s-1<\/sup>\u00a0\u2219\u00a0km-2<\/sup> considering a specific yield variation of\u00a0\u00b1 1\u00a0%.<\/span><\/p>\n In\u00a0the\u00a0final step, the\u00a0obtained ET value was compared with the\u00a0results of\u00a0ET estimation by a different method, carried out by the\u00a0same authors in\u00a0the\u00a0same wetland\u00a0[25]. In\u00a0the\u00a0referenced study, ET was calculated based on fluctuations in\u00a0streamflow, under the\u00a0assumption that the\u00a0highest daily flow represents discharge unaffected by ET. Using this method, ET was determined to be 11\u00a0l\u00a0\u2219\u00a0s-1<\/sup>\u00a0\u2219\u00a0km-2<\/sup>. Based on the\u00a0calculation of\u00a0PET using Oudin\u2019s method, the\u00a0study reported an average ET value of\u00a025\u00a0l\u00a0\u00b7\u00a0s-1<\/sup>\u00a0\u00b7\u00a0km-2<\/sup>. These results from 2022 are consistent with the\u00a0ET values obtained using White\u2019s method \u2013 20\u00a0l\u00a0\u2219\u00a0s-1<\/sup>\u00a0\u2219\u00a0km-2<\/sup> \u2013 within\u00a0the\u00a0accuracy limits of\u00a0the\u00a0specific yield estimate.<\/span><\/p>\nINTRODUCTION<\/h2>\n
Brief review of\u00a0previous research<\/span><\/h2>\n
Influence of\u00a0ET on GWL fluctuations<\/h3>\n
![\"\"](\"https:\/\/www.vtei.cz\/wp-content\/uploads\/2025\/06\/Patek-fig-1.jpg\")
<\/a>\nFig. 1. Typical diurnal fluctuation of\u00a0groundwater level and baseflow measured in\u00a0the\u00a0experimental catchment at the\u00a0foot of\u00a0the\u00a0Alps (modified after\u00a0[4])<\/h6>\n
<\/a>\nFig. 2. ET calculation using the\u00a0White method\u00a0[1] (modified after\u00a0[5])<\/h6>\n
<\/a><\/h6>\nFig. 3. Model fit of the\u00a0measured values of\u00a0diurnal groundwater fluctuation in\u00a0the\u00a0vicinity of\u00a0the\u00a0river. Model (1) uses the\u00a0same ET amplitude for the\u00a04th and 5th\u00a0day, whereas model (2) takes into account the\u00a0different ET amplitude for\u00a0the\u00a04th\u00a0and\u00a05th\u00a0day. The\u00a0dotted line represents the\u00a0effect of\u00a0fluctuating river stage<\/h6>\n
Relationship between ET and streamflow<\/h3>\n
<\/a>\nFig 4. Schematic overview of\u00a0the\u00a0Fonley method of\u00a0calculating ET from diurnal fluctuation of\u00a0river flow (modified after\u00a0[13])<\/h6>\n
ET and fluctuations of\u00a0GWL combined with streamflow fluctuations<\/h3>\n
ET, GWL fluctuations, and streamflow combined with independent measurements of\u00a0vegetation transpiration<\/h3>\n
Review summary<\/h3>\n
Measurement of\u00a0ET influence in\u00a0the\u00a0wetland on the\u00a0Lib\u011bchovka<\/h2>\n
LOCATION<\/h3>\n
<\/a>\nFig. 5. Wetland area ([24])<\/h6>\n
METODOLOGY<\/h2>\n
<\/a>\nFig. 6. Location of\u00a0piezometers in\u00a0the\u00a0wetland; for a\u00a0general overview, location of\u00a0the\u00a0weir constructed by the\u00a0same author during previous work\u00a0([25]) is also shown (background\u00a0map: Base map 1 : 10,000 from\u00a0[26])<\/h6>\n
Tab.\u00a01. Depth of\u00a0the\u00a0piezometer<\/h5>\n
<\/a>\n
\nin\u00a0autumn 2023 (12\u201313 November 2023). This was a cool and wet period, during which it was expected that the\u00a0soil profile would retain\u00a0an above-average amount of\u00a0water and the\u00a0GWL would be high. The\u00a0second sampling was conducted during hot summer days in\u00a0a rain-free period (11th and 12th August 2024),
\nwhen it was expected that the\u00a0soil profile would contain\u00a0a below-average amount of\u00a0water and the\u00a0GWL would be low.<\/span><\/p>\n<\/a>\u00a0 \u00a0 \u00a0(1)<\/p>\n<\/a>,\u2003(2)<\/p>\n<\/a>,\u2003(3)<\/p>\n<\/a>,\u2003(4)<\/p>\nRESULTS<\/h2>\n
Tab.\u00a02. Periods with diurnal fluctuation in\u00a0groundwater level in\u00a0the\u00a0wetland (2024)<\/h5>\n
<\/a>\n<\/a>\nFig. 7. Diurnal fluctuation in\u00a0groundwater level in\u00a0early spring (A), late spring (B), midsummer (C, D), and the\u00a0end of summer (E)<\/h6>\n
Tab.\u00a03. Results of\u00a0soil moisture measurements<\/h5>\n
<\/a>\nFig 8. Measuring soil moisture content in\u00a0the\u00a0wetland (in\u00a0the\u00a0profile description, soil horizons are described in\u00a0brackets according to the\u00a0soil classification)<\/h6>\n