Quantifying the Soil Water Storage Capacity of Flysh Catchments Surrounded by Mixed Forests in Outer Carpathians

The objective of this paper is to evaluate the retention capacity of catchments situated in the Beskid Wyspowy region of southern Poland. To accomplish this, we employed the Soil and Water Assessment Tool (SWAT+), a newly developed hydrological model. The large heterogeneity of the catchment area and the limitations of measurement techniques necessitated the use of hydrological models. The study found that forested and pasture areas had higher soil water retention potential than agricultural areas. Furthermore, surface runoff of both catchments correlated negatively with soil water retention potential and evapotranspiration, however, positively with mean annual precipitation. Soil moisture was directly related to mean annual precipitation and electrical conductivity. The research also advocates incorporating the Landscape Hydric Potential (LHP) index into existing hydrologic models, in line with the physical-based SWAT+. Future watershed models based on water balance should be developed to increase resilience to climate change manifestations in the flysh Carpathian Mountains. Soil water retention potential varies with land use in the catchment. Surface runoff negatively correlated with soil water retention potential. Curve number and LHP index can be incorporated into hydrologic models. Forested/pasture areas had higher water retention potential. Watershed models aid climate resilience by simulating hydrological processes. Soil water retention potential varies with land use in the catchment. Surface runoff negatively correlated with soil water retention potential. Curve number and LHP index can be incorporated into hydrologic models. Forested/pasture areas had higher water retention potential. Watershed models aid climate resilience by simulating hydrological processes.


Introduction
Land-use changes are important parameters in the runoff process as they affect water storage capacity and runoff generation mechanism (Drake and Vafeidis 2004). Mountain soils, due to their occurrence conditions, are exposed to the degradative influence of various environmental and anthropogenic factors (Jaguś 2020;Loba et al. 2020) which may lead to a change in soil properties and, consequently, to a change in water retention. Soil erosion is a physical process and it is considered a worldwide environmental problem (Maliqi and Sign 2019). The destructive effects of soil water erosion are significant, for region with different land-uses . For example, the soil profile is transformed, especially with regard to the basic properties of the topsoil (Bolotov et al. 2019). This applies to changes in the grain size composition, quantity and quality of humus compounds, and as a consequence, a decrease in soil productivity (Barsukova et al. 2020). The landslides that take place in mountain areas cause the destruction of vegetation, and they break the continuity of soil cover which results in different physicochemical and biological properties of the soil (Błońska et al. 2016(Błońska et al. , 2018. Additionally, human activities cause soil degradation (Bogunovic et al. 2020). Human activities such as deforestation, agriculture, and animal husbandry have drastically altered mountain ecosystems. Specifically, the soil's diminished health has resulted in reduced agricultural yields and increased occurrences of floods and droughts (Baffaut et al. 2020). Over recent decades, there has been a significant increase in degradation processes of soil and land within the European Union and its bordering countries, and there is evidence that these processes will further increase if no action is taken (Montanarella 2007). The reports emphasize the need to act to prevent the current degradation of Europe's soils (Belozertseva and Vlasova 2020).
Currently, many attempts are made in forest areas to restore their rich hydrographic networks, to improve the quality of their environment and recreational values, as well as to augment rainwater runoff retention (Chandler et al. 2018;Bonfante et al. 2020). Certain sections of some Polish rivers and their valleys are modified to only a relatively small extent and have high ecological value (Piniewski et al. 2018). Many sections of river valleys have been strongly modified by anthropogenic activities. The progressing fragmentation of rivers and their valleys possess a justified concern regarding the ecohydrological balance of watershed (Wierzbicki et al. 2020). In previous decades, traffic requirements have led to narrowing of river valleys, to diverting watercourses into canals or backfilling them, as well as to their environmental and landscape degradation (Dai et al. 2019). A restoration of the ecohydrological potential involves restoration and restructuring activities . It is important to ensure that the way in which they are developed does not restrict the flow of flood waters, and the used materials and components of structural landscape are reliably fixed to the ground and resistant to water (Rajib et al. 2020). The tree stands and canopy restrict water erosion of soils and water evaporation from soils, especially during the summer (Montaldo et al. 2020). The conservation of such habitat corridors like rivers and their valleys not used for industrial, residential, technical infrastructure, or agricultural purposes, without embankments, or with embankments moved away from the river and close to the valley edge, not only ensures normal functioning of plant and animal communities, but also improves flood prevention in cities and villages located in river valleys (Amatya and Wałęga 2020). It also facilitates protection of river waters against non-point pollution of agricultural origin and contributes to processes of self-cleaning of waters (Qiu et al. 2020). It also supports cleaning of waters during periodic flooding of the valley areas, as well as reducing erosion of the valley soils, thus preventing river sedimentation (Fortesa et al. 2020).
The Landscape Hydrologic Potential (LHP) method is an original and effective concept for the functional use of Geographic Information System (GIS) analyses and tools in evaluating the cumulative hydrologic effect using ecosystem attributes (Lepeška 2010). The application of this method allows the determination of the potential of ecosystems to reduce and immobilize precipitation and the ability of water to penetrate deep into the ground, thus contributing to the evaluation of ecosystem properties in terms of water management (Šatalová and Kenderessy 2017;Petrovič et al. 2017). Availability of a landscape hydric potential model for mountain catchment areas is important for effective catchment management in the face of current changes and climatic anomalies (Wojkowski et al. 2019).
Watershed models are primarily used to predict the land-meteorological system, and to understand various effcte on hydrological processes due to climate change (Pulighe et al. 2021). Soil water deficiency measures are defined by a pattern of various catchment parameters (van Tol et al. 2021). The SWAT+ model is a semi-distributed hydrological model, developed to analyze and predict the impacts of land-management practices on water, sediment and chemical yields. With SWAT+, hydrological studies are conducted to produce reliable results (Chawanda et al. 2020). To better address the limitations of the Soil and Water Assessment Tool (SWAT+), more research is necessary to explore its applicability across a range of catchment types, including forests and pastures under varying environmental conditions (Harms et al. 2022). Wang and Chen (2021) recommended that there is a need for more studies on the effectiveness and limitations of the recently developed SWAT+ model in comparison to other hydrologic models. According to Dash et al. (2023), studies are needed to explore the potential of meteorological simulations and hydrological models for water and environmental resource management in different geographical regions and catchment types. The need for more research on catchment retention capacity has become increasingly urgent, particularly in the context of changing land use patterns and the effects of climate change on forests (Sun et al. 2023).
Although the limitations of current measurement techniques are widely recognized, further research is necessary to advance and enhance these methods to effectively measure the hydrological system of catchments with greater accuracy Despite some existing research, there remains a significant research gap in understanding the complex relationship between land use/cover, soil water retention, surface runoff, and evapotranspiration, particularly in different catchment types and environmental conditions. Investigation is needed to gain a deeper understanding of the impact of precipitation and soil moisture on surface runoff and evapotranspiration in varying catchment types and environmental conditions. Whereas the Landscape Hydric Potential (LHP) index exhibits promise for integration into hydrologic models, a new approach is required to evaluate its efficacy in diverse catchment types and environmental contexts (Walega et al. 2020). Moreover, it is crucial to conduct research on water balance models and their potential to bolster up resilience to climate change across different catchment types, including pasture and forested agricultural areas.
This study emphasizes the vital role of forest conservation in maintaining a healthy balance between water infiltration and retention in mountainous areas. By assessing the flysch catchment's retention capacity through hydro-physical parameter estimation, soil water availability could be estimated realistically and enhance the Carpathian Mountains' resilience to climate change. To achieve this, we have set four specific objectives, which are: 1) to determine evapotranspiration using SWAT+; 2) to quantify the landscape hydric potential of the flysch catchment; 3) to compare soil parameters at two spatial scales in the flysch catchment; and 4) to identify areas prone to water deficit in the forest catchment utilized for agricultural purposes.

Study Area
The Smugawka and Mszanka rivers originate in the hills of northern part of the Carpathian Mountains, in the Beskid Wyspowy region near the Gorce National Park (Fig. 1). Two streams were included in the study as they are severely forested and consist typical drainage basins for the Carpathian flysch structure. Moreover, both areas are prone to water erosion, and thus, to soil water scarcity. The catchment areas of both rivers are located in the southern part of Poland, ranging in latitude from 19°47-20°15′ and longitude from 49°33-49°44′. The Beskid Wyspowy region is characterized by a large number of outstanding peaks. The tops of the mountains reach up to 1,170 m above sea level and rise 400-500 m above the typical foothill plateau with rolling hills. The approximate area of the Smugawka basin is 7.69 km 2 and the Mszanka basin covers an area of 55.14 km 2 . The Smugawka River is 4.83 km long and begins at an altitude of 570 m a.s.l., while the Fig. 1 Location of the study area-Smugawka and Mszanka river basins Mszanka River is 18.83 km long and has its sources at altitude 728 m a.s.l. The elevation above sea level of the Smugawka ranges from 391 to 858 m a.s.l., while the catchment of the Mszanka ranges from 371 to 1274 m a.s.l. (Fig. 1).

Sampling and Data Handling
Our study area is dominated by meadows, farmland, and forests. The dominant slopes in the region ranged from 10 to 20%. Measurements of soil resistance were carried out in three repetitions at one point in arable lands and grasslands at the 0-80 cm layer and in forest soil at the 0-20 cm layer, during subsequent field visits. In total, 110 samples were obtained for each type of use ( Table 1). The measurements were made using an electronic penetrometer with automatic control of the penetration speed by Eijkelkamp. An electronic penetrometer with automatic control is a device used to measure the strength and compaction characteristics of the soil. The penetrometer consists of a probe that is pushed into the soil at a controlled rate, and the force required to penetrate the soil is recorded electronically.
The automatic control feature allows the device to maintain a constant rate of penetration, which helps to ensure consistent and accurate measurements. This is important because soil strength and compaction can vary depending on factors such as soil type, moisture content, and other environmental conditions. The electronic component of the device allows for the data to be collected and analyzed digitally, which can improve the accuracy and speed of the testing process. The results can be used to evaluate soil properties for a variety of applications, including engineering and construction projects, agriculture, and environmental monitoring. To calculate the value of hydraulic conductivity of the soil (filtration coefficient) in the zone of full saturation (Ks), the Beyer formula was used: where: C is a coefficient set at 0.0045 (m s −1 ), C U is the coefficient of uniformity calculated by dividing the diameter of the soil fraction d 60 by d 10 , and d 10 is the effective diameter (mm).
Soil electrical conductivity (EC), moisture and temperature samples were performed at 80 spots by means of a test probe (type-HH2 moisture measure) using the TDR (time domain reflectometry) method in Smugawka river and 160 spots in Mszanka river. This method allows for non-invasive, accurate, and fully automatic measurements of electrical conductivity, moisture and temperature in the soil. For the analysis of particle size, averaged samples were taken (45 points) at each test area. Each of the test samples was taken from the top layer of the soil (0-25 cm). The collected samples were dried, and then passed through a sieve with an aperture of 2 mm at room temperature. Determination of soil texture was made using Casagrande's aerometric method with Prószyński's modification. A modified Tiurin method was used for organic carbon determination. It consists of determining the organic carbon by oxidizing it to carbon dioxide in an acidic environment with the use of potassium dichromate. The amount of Corg is represented by the titration of the amount of oxygen consumed to oxidize the carbon in the above-described reaction (Oleksynowa and Tokaj 1991).
The area distribution of mean annual precipitation was calculated using precipitation data from Worldclim 2 with spatial resolutions ~ 1 km 2 for the 30-year period from 1970 to 2000 (Fick and Hijmans 2017). Land use was obtained from publicly available Copernicus Global Land Service land use data (Buchhorn et al. 2020). A digital elevation model (DEM) with 1 × 1 m resolution was acquired from publicly available Geoportal data (Geoportal 2021).
(1) Ks = C ⋅ log 500∕C U ⋅ d 10 2 Table 1 Statistical comparison of land cover values and characteristics relative to the study catchments mean ± standard deviation; small letters in superscript of mean values denote significant differences between land use types of a sub-catchment area (a, b, c) and between catchments (A, B); electrical conductivity (mS·m −1 ); soil moisture (%); soil temperature (℃); hydraulic conductivity ( Data on the location of each soil type and texture were obtained from a 1:50,000 scale vector soil and agricultural map (WODGiK 2018). The amount of soil organic matter content was obtained from the global soil information map (Hengl et al. 2017). CORINE program land use, land cover vector layer were used to determine landscape classes.

Landscape Hydric Potential (LHP)
The Landscape Hydric Potential (LHP) is a soil parameter that indicates the soil's ability to retain water and is used to estimate the amount of water that can be stored in the soil for plant use. It is a measure of the soil's ability to support plant growth and is used to identify areas where soil drainage may be inadequate, such as wetlands or areas with a high water table. The LHP takes into account factors such as soil texture, depth to water table, and slope of the land surface, and is typically expressed as a value between 0 and 1, with higher values indicating greater water-holding capacity (Wojkowski et al. 2019). The LHP property estimate consists of a compilation of average precipitation rate coefficients and landscape attributes affecting runoff infiltration, retardation, and retention. The LHP index is dimensionless and reflects the relative value of LHP across the landscape. Each environmental attribute contributes to the final index score based on its importance to hydrological processes and water availability. Equation (2) is used to calculate the spatial distribution of LHP, which is the sum of all environmental attributes: where: H indicates hydrogeological conditions; St is the soil type; Ss is the soil texture; Pi is the precipitation; Si is the slope inclination; F are the forest stands; N is the non-forest landscape.

Climatic Conditions (Pi)
In the temperate zone atmospheric precipitation is the main source of water, especially in mountainous zones the presence of horizontal precipitation plays a special role presented on Fig. 2. We used the QGIS 3.16 plugin in SWAT+ to generate the potential evapotranspiration layer. The potential evapotranspiration was subtracted to determine the total amount of precipitation that could be infiltrated and retained: where: Pi is the amount of water available, able to infiltrate into the soil (mm); PØ-average annual precipitation (mm); EO-average annual potential evapotranspiration (mm). The river basins were divided depending on the value of Pi coefficient according to the following criteria: + 2: > 1100 mm; + 1: 450-1100 mm; 0: 0-450 mm; -2: 0-(-450) mm; -3: < (-450) mm (Wojkowski et al. 2022).

Characteristics of Forest Stands (F)
The hydrological functions of forest stands depend on the ecological stability in which they are located (Fig. 3). The land use layer was assessed separately using data from the report on the state of forests in Poland (IBL 2017). The degree of ecological stability of forest ecosystems was estimated and assigned the following categories: + 2: very good ecological stability; + 1.75: great ecological stability; + 1.5: lowered ecological stability; 0: unstable ecological stability; -0.25: extremely disturbed ecological stability.

Characteristics of Non-Forest Landscape (N)
The land cover maps of the CORINE program were used to determine non-forest landscape classes (Fig. 2). Land use potential relative to infiltration and precipitation retention was distinguished by assigning to the following classes (Jarvis et al. 2013;Larsbo et al. 2014): + 2.5: pastures, transitional woodlands-shrubs, sparsely vegetated areas; + 2: land cover principally occupied by agriculture with significant areas of natural vegetation; + 1: complex cultivation patterns; -1: non-irrigated arable land; -4: discontinuous urban fabric, construction sites.

Modelling of Evapotranspiration and Curve Number
Spatial distribution of ET was generated using SWAT+, a completely revised version of the model (Bieger et al. 2017). This work employed SWAT+ to analyze the impact of land use landscape change on hydrological processes, and then the usefulness of SWAT+ in the evaluation of evapotranspiration and curve number was also verified. Simulations of evapotranspiration changes in 199-2020 were prepared in order to analyze the response to changes in the landscape pattern (Fig. 3). The Penmann-Monteith method was used to generate the PET values over the watershed. In SWAT+ the catchment area is divided into sub-basin in the form of flood plains and upland. A landscape unit (LSU) is defined by a set of HRUs and can be a partial subbasin. Drainage basin may be a flood or upland unit, or a multi-HRU grid cell. The water balance of each LSU is converted from representation into HRU units called hydrologic response unit (HRU) -the catchment area is determined by soil use, land use and topography. Each HRU in the sub-basin is summed up and the resulting conditions are routed through canals, ponds and/or reservoirs to the catchment mouth for each landscape type. The statistical summary of modelled values by land cover, in relation to the study catchments, highlighted the differences in land use patterns (Table 2). LHP index was compared with occupied area percentage in study catchments, and then assesed (Table 3). To evaluate the accuracy of the modeled datasets, various indices such as NSE (Nash-Sutcliffe Efficiency) and R-square (Coefficient of Determination) were employed. The calibration data for evapotranspiration were collected from 2004 to 2014, while validation was performed for the period of 2014 to 2018 (Table 4). Data for this purpose was obtained from the Climate Data Store (CDS) (https:// cds. clima te. coper nicus. eu/ cdsapp# !/ search? type= datas et) of the Copernicus program. The CDS provides access to a wide range of climate data and products, including meteorological reanalysis data, remotely sensed data products, and climate model outputs. It is a centralized platform for accessing and managing climate-related data, and it is open to the public.

Statistical Analysis
Using the ordinary kriging method (spatial interpolation), the soil resistance, electrical conductivity, volumetric moisture, hydraulic conductivity, and temperature in the soil layer from 0 to 20 cm were shown. Data collected in 2014-2018 was used for this analysis. The results were averaged and presented in the form of spatial distribution for the entire catchment area. Geostatistical analysis was aimed at optimizing spatial information by selecting the right algorithms. The method used was useful in forecasting measurement points with a similar spatial distribution of the examined features and in selecting those variables with the greatest strength of connection with surface runoff. The maps were generated using the Surfer® version 15, intended for mapping and modeling the terrain surface. A two-way ANOVA test at a significance level of 0.05 was used to identify statistically significant differences. In addition, mean values, standard deviations and average errors were provided for the factors and summary measures of the LHP model. Using QGIS 3.16 software, spatial hydrological indices were modeled and tentatively mapped. Principal Component Analysis (PCA) is a statistical method that is widely used in hydrology to analyze large datasets and identify underlying patterns or relationships between variables. In forested catchments, PCA was used to identify the most important variables that control the hydrological behavior of the system, using R-studio.

Results
Electrical conductivity reached the highest mean values for agricultural land, where it was 49.67 for small catchment (C1) and 30.61 for large catchment (C2) (Fig. 4, Table 1). Statistically significant differences were noticed for the agricultural land of the small catchment Table 2 Statistical summary of modelled values by land cover relative to the study catchments mean ± standard deviation; small letters in superscript of mean values denote significant differences between land use types of a sub-catchment area (a, b, c)  where it differed from the other land uses and the areas of the large catchment. A two-way ANOVA showed that soil moisture was significantly higher for C1, with different results for forest and agricultural land at 41.92% and 44.08%, respectively (Fig. 4, Table 1). There were no significant differences in soil temperature between catchments and land use types. A similar lack of any statistically significant differences between the studied catchments was observed for soil hydraulic conductivity. Soil resistivity for the smaller C1 catchment differed between the different land uses. Farmland and pasture had significantly higher values relative to the catchment, with scores of 2.32 and 2.31 MPa, respectively (Fig. 4, Table 1). Soil total organic carbon content for the C1 catchment was significantly different for the forested areas where it was 16.14 g·kg −1 (Fig. 4, Table 1). On the other hand, the lowest amount of total organic carbon was recorded for pastures, with a result of 5.61 g·kg −1 . Soil uses in the large C2 catchment area were not significantly different. The highest amount of total organic carbon was noticed for the land used for agriculture, where it was 10.16 g·kg −1 . The causes for the observed differences in soil characteristics between land uses and catchment areas may be complex and multi-faceted. For example, agricultural land may have higher electrical conductivity due to the application of fertilizers and other agricultural inputs that increase the salt content of the soil. Similarly, differences in soil moisture content may be due to variations in topography, drainage patterns, and vegetation cover. The higher values of soil resistivity in farmland and pasture may be related to differences  in soil texture, which can affect the electrical conductivity of the soil. The variations in total organic carbon content may be due to differences in the amount and quality of organic matter inputs from vegetation and/or differences in soil management practices. The retention potential in both study catchment cases reached significantly higher values for forest and pasture land use. For the smaller catchment C1, the retention potential of forest land was 7.51, and for pasture land it reached an average value of 11.43 (Fig. 5, Table 2). On the other hand, agricultural land had a score of only 2.81. For the larger catchment C2, forest land and pasture land had scores of 8.37 and 7.58, while agricultural land scored 5.02. The modeled runoff curve values scored statistically significantly higher for the larger C2 catchment, with 74.21 for forest land, 76.37 for pasture land, and 82.35 for agricultural land (Fig. 5, Table 2). In both cases, statistically higher values were recorded for agricultural land. For both studied catchments, significantly higher values of annual evapotranspiration were characterized by pasture land, reaching 762.87 mm per year for the larger catchment C2 and 722.80 mm for the smaller catchment C1. The lowest values were associated with agricultural areas, with values of 680.11 mm·yr −1 for C2 and 688.13 mm·yr −1 for C1. Significantly higher values of mean annual precipitation were found in the larger catchment area of C2. Especially higher values of mean annual precipitation of both catchments were characterized by forested areas, with results of 1015.06 mm·yr −1 and 1188.11 mm·yr −1 .
The comparison of the landscape hydrological potential index with the percentage of occupied area in relation to the study catchments was characterized by more favourable properties for agricultural and forest areas (Fig. 5). In the small catchment area of C1, about 31% of agricultural land achieved the most favourable parameters placing it in the second category of very good soil retention potential (Table 3). On the other hand, in the large catchment area, the most optimal result was obtained for forest soils where 12% of forest soils were placed in the second category of very good soil retention potential. The most threatened with water loss were forest areas in the case of the smaller catchment C1 and agricultural land of the larger catchment C2, where the fourth category of organic soil retention potential was covered by 41% and 20% of the area, respectively (Table 3). The catchments were shown to exhibit moderate retention potential. Forested areas have significantly higher soil retention potential than other land uses in both catchment cases studied (Fig. 5). We identified four key areas of water deficit indicators related to surface runoff patterns. The results showed that the forested area was poorly represented. Water retention and water level changes were related to land use. Potential retention values were higher in forested areas, indicating the intentional ability of this structure to vary anthropogenic stressors that may alter the hydrology of a catchment. Agricultural areas in the catchment experienced excessive water level decline compared to forested areas (Fig. 5). On the other hand, forested areas with high LHP showed stable water levels compared to agricultural areas. These differences suggest that water management in mountainous regions may cause large fluctuations in water levels.
The NSE index for the Smugawka catchment during calibration was below 0.5, indicating a poor fit of the model to the empirical ET data. However, model performance during validation improved for both studied catchments. The validation results showed that the assessment of the model's prediction quality in previously unknown conditions is adequate, although it would be beneficial to conduct it over a longer period of time (Table 4). Another objective of the study was to estimate the actual evapotranspiration in an agricultural area using the Soil and Water Assessment Tool (SWAT+). The model was calibrated using data from 2004 to 2010, and validation was conducted for the period 2014-2018. The goodness of fit was evaluated using the R-squared value, which ranged from 0.55 to 0.65. However, the model consistently overestimated the observed values, as indicated by the higher simulated values compared to the measured ones (Eini et al. 2021).
Principal component analysis considering land use types of individual catchments in relation to the studied and modeled soil parameters explained 32.1% of variables (Fig. 6). Agricultural soils were the most different from forest and pasture soils, whose characteristics overlapped. PCA showed a correlation between the soil retention potential coefficient and the modeled values of evapotranspiration and soil temperature. Higher LHP is associated with forested and pasture land, while lower soil retention potential values are correlated with agricultural land. The number of surface runoff events correlated negatively with soil retention potential values along with evapotranspiration, but positively with mean annual precipitation (Fig. 6). Soil moisture was related to electrical conductivity and inversely correlated with the amount of average annual precipitation.

Assessment of the Catchment Retention Potential
Our results indicated that evapotranspiration modelling, soil hydraulic properties assessment, evaluation of landscape hydric potential highlight the significance of forests in water retention. The optimal place for the catchment area may be the system of mutual use of forest and water management (Erlandsson Lampa et al. 2020). Forest growing on mountain area must cover the entire catchment area with high water potential ability (Lidberg et al. 2020). Precipitation data and drainage areas are key inputs to all models. Our study found that the grassland had a low level of annual evapotranspiration consistent with earlier studies. For grassland and forest areas, estimated values were found to be most realistic (Fig. 3). The enhancement of plant and pasture productivity will result from the lengthening of the growing season and the increase in atmospheric total over time (Fiorini et al. 2020). Infiltration rates were found to be ten to a hundred times higher under trees, while the wooded area remained relatively intact compared to adjacent pastures (Latorre and Fig. 6 Principal component analysis including land use types of individual catchments relative to studied and modeled soil parameters. CN -curve number; Prec -precipitation; LHP -landscape hydric potential; Evap -evapotranspiration; ST-soil temperature; TOC -total organic carbon; EC -electrical conductivity; SM -Soil moisture; HC -hydraulic conductivity, SR -soil resistance Moret-Fernández 2019). In areas where the soil has been degraded, attention must be given to available soil water. It must be reinstated here that important hydrological processes in forest and agricultural catchment may be simulated with SWAT+ (Fig. 3). Water management in forestry use may contribute to the stabilization of the water level. Due to the increasing scarcity of water, local managers are looking for new and sustainable solutions to problems related to water supply and topographic boundaries of the relief, where water is the most valuable in a landscape. Water retention capability in agricultural and forested areas can be evaluated using the Curve Number (Kristanto et al. 2021). SWAT+ model was able to accurately represent CN parameter, as evident from low level of CN parameter in forest, a medium level in grasslands, and a high level in agricultural areas (Fig. 3).
For the most effective projects in terms of adapting forests to climate change, plant-climate-resilient strategies has been implemented (Lilburne et al. 2020;Wysocka-Czubaszek and Roj-Rojewski 2020). Lack of cost-effective indicators to quantify the anthropogenic potential of hydrological changes in water area on a regional and national scale has become a concern. We have developed a classification framework according to landscape-scale water storage potential for land use. The catchment area should be managed by human activities to support rainfall infiltration and retention as land use activities are secondary factors that alter the hydrology of the drainage. LHP provided a tool for the assessment of anthropogenic retention factors (Fig. 5). In catchment areas, land productivity, soil organic carbon, and land cover and land cover change can be assessed, and in particular, the impact of soil degradation (Pena et al. 2020). The environmental impact of drought cannot be quantified with the help of the capacity of the catchment's hydrological potential (Simmons and Anderson 2016). However, this study revealed that modeling of hydrological properties, is essential to show forest area prone to drought (Fig. 5). While many such indirect indicators exist, we recommend the use of SWAT+ and LHP for assessing soil degradation and water deficiency. Any inappropriate human activity in the river basin can cause a series of irreversible changes that can completely affect the nature of water resources. The proposal of a higher demand for hydrological models is another issue.
Soil protection is one key element in the field of land degradation (Zhuo et al. 2020). The solutions proposed under this article could provide policymakers with more effective incentives to adequately protect the soil in the EU . Proposals and ways to tackle soil quality problems through a targeted and proportionate risk-based approach are still under debate. The 2013 EU Strategy on Adaptation to Climate Change recognizes the importance of combating desertification as one of the climate adaptation actions that should be supported for soil protection (Panagos et al. 2018). To regionalize point measurements to a catchment scale, statistical or geostatistical methods are typically used. These methods involve analyzing the spatial patterns and relationships between the point measurements and other environmental variables, such as elevation, soil type, and land use, to develop a model that can be used to estimate variables of interest across the entire catchment. In the context of the SWAT+ model, regionalization is important because it allows the model to be calibrated and validated using data from a limited number of locations, while still providing accurate predictions of hydrological processes at the catchment scale. We used SWAT+ and regionalization techniques, and a regionalization approach known as the LHP to regionalize point measurements of soil water retention at the catchment scale (Fig. 4). The LHP method involves comparing the variability of the point measurements to the variability of other environmental variables at different spatial scales to determine the optimal scale for regionalization. We found that using LHP in combination with the SWAT+ model improved the accuracy of sediment yield predictions at the catchment scale compared to using SWAT+ alone. This is because LHP allowed for a more robust and accurate estimation of soil water retention across the catchment, based on the spatial patterns and relationships between the point measurements and other environmental variables. Overall, the combination of regionalization techniques such as LHP with watershed scale models such as SWAT+ is important for improving our understanding and prediction of hydrological processes in large and complex catchments, particularly where data are limited or incomplete.

Waterproof Forest Ability
Forest programs and national sustainable forestry practices are beneficial to the soil , especially adaptation to changing climatic conditions (Silva and Lambers 2020). Due to the expected increase in the frequency, range and intensity of extreme climatic phenomena, it is highly advisable to prepare a soil management system in agricultural and forest areas (Kercheva et al. 2019). One main finding of the study is highlighting the importance of incorporating accurate measures of soil water retention potential, evapotranspiration, and mean annual precipitation into hydrologic models to better understand and manage surface runoff in catchments. We also recommend the potential use of alternative methods such as the curve number approach and LHP to improve hydrologic modeling accuracy (Fig. 3).
Forest has an impact on the prevention of desertification and soil degradation, and consequently, against erosion. It also increases the ability to absorb carbon (Wells et al. 2019). The course of climate change is largely unpredictable, especially on a regional and local scale, therefore, it is advisable to focus on adaptation measures (Slessarev et al. 2019). Agro-forestry activities can help to tackle soil degradation through local action on peripheral action. As forest protection against biotic and abiotic active forestry is within the main environmental policy, it is essential to have up-to-date information on the condition of forest soils (Erdozain et al. 2020). Interaction with the water sector plays an important role in adaptation measures. Unavailability of water resources may limit adaptation activities by eliminating species with increased water requirements (Mwazi et al. 2020). On the other hand, in catchments, the higher the forest cover, the higher the stability of water resources in the event of climate change and the greater the potential for natural water retention (Lee et al. 2020). The main task in forest management should be to improve the effectiveness of forest protection. Logging companies can take precautionary measures to prevent many kinds of negative environmental effects from their activities, which negatively affect soil density and water infiltration (Wei et al. 2019). As a result, many factors influence hydrological model calibration, including soil properties, vegetation, topography, and soil moisture. Specifically, we believe that the SWAT+ results should improve the model in a tangible way with LHP, especially for small agricultural catchments (Fig. 5). The midforest reservoirs and watercourses should be restored and the existing swamps and peat bogs must be preserved. It is in the best interest of the land managers to ensure that forest soils remain healthy, as incorrect identification of retention capacity can lead to reduced tree production and poor timber quality for future harvesting operations (Eichler et al. 2020). Abiodun et al. (2018) used daily timescale analyses, employing a calibration period of six years (2000)(2001)(2002)(2003)(2004)(2005) and a validation period of seven years (2007)(2008)(2009)(2010)(2011)(2012)(2013), to estimate evapotranspiration through the application of the SWAT and MOD16 methodologies. The results showed differences in ET estimation between the methods, with variations of up to 31%, 19%, 15%, 11%, and 9% observed at spatial resolutions of 1, 4, 9, 16, and 25 km 2 , respectively. The study recommends a spatial scale of confidence of 4 km 2 for estimating catchment-scale evapotranspiration in areas with complex terrain. In another investigation, the output results of the SWAT model were found to be reliable and effective for a typical watershed of the Yellow River Basin, with a recommended coefficient of R 2 greater than 0.5. The finding revealed that land use and land cover change (LUCC) in the catchment led to a significant increase in forest (21.61%) and settlement (23.52%) and a slight reduction in cropland (-1.35%), resulting in a 4.93% increase in evapotranspiration and a clear decline in surface runoff and water yield by 15.68% and 2.95%, respectively, at the whole basin scale. At the sub-basin scale, the increase in settlement led to an increase in surface runoff and water yield (Liu et al. 2023). Getachew and Manjunatha (2022) investigated impact of landuse change on Lake Tana basin and reported a significant impact on hydrological processes. Specifically, the studied basin experienced an increase in farmlands and built-up lands, coupled with a decline in shrublands, grasslands, and bare lands. Consequently, this led to an increase in the basin's mean annual water yield and surface runoff, along with a decrease in evapotranspiration and lateral flow. These results emphasize the importance of considering land use changes in future watershed and basin management to sustain hydrological processes. Monitoring should also include activities carried out as part of the mountain water retention, which will allow for the reconstruction of valuable natural ecosystems, thus creating a positive impact on slowing down the outflow of surface waters (Hou et al. 2020).
The higher LHP values observed in forested and pasture lands can be attributed to the presence of vegetation cover, which enhances water infiltration and retention in the soil (Fig. 3). The roots of trees and plants in forested and pasture lands help to bind soil particles, which increases soil porosity and promotes water infiltration. This, in turn, reduces the potential for surface runoff and increases the soil retention capacity (Fig. 5). On the other hand, the lower soil retention potential values observed in agricultural land can be attributed to the removal of vegetation cover and soil disturbance during land preparation activities such as plowing and tilling. The absence of vegetation cover increases soil erosion, which reduces soil porosity and lowers soil retention capacity. Additionally, agricultural lands are often subjected to frequent tillage practices, which further degrade soil structure and reduce soil retention potential. The negative correlation between soil retention potential and surface runoff events can be attributed to the fact that higher soil retention potential values imply higher infiltration rates and lower surface runoff potential (Fig. 6). On the other hand, areas with lower soil retention potential are more prone to surface runoff events due to the reduced capacity of the soil to retain water. A positive correlation between surface runoff events and mean annual precipitation is expected since high precipitation events are more likely to cause surface runoff in areas with lower soil retention potential.
The relationship between soil moisture and electrical conductivity can be explained by the fact that high levels of dissolved ions in the soil solution indicate high electrical conductivity. These ions can increase the osmotic potential of the soil solution, reducing the availability of water to plants and contributing to lower soil moisture levels. Moreover, higher average annual precipitation levels can lead to the leaching of soil ions, which can increase electrical conductivity.
The study's relationships between land use, soil properties, and hydrological variables can provide valuable insights into the management of forested catchments for water resources protection and sustainable land use. GIS techniques can be used to gather valuable data from many years and assess land use and soil water parameters for more effective management. LHP and SWAT+ modelling enabled the estimation of soil retention capacity using contextual indicators. Our study provides valuable insights into addressing sustainability concerns related to forestry. Effective management guidelines for catchment areas should include the dissemination of good practices and techniques for evaluating the effects of irrigation investments on agricultural land. Moreover, the study promotes the use of plant species that are better adapted and more tolerant to climate change, which can help ensure sustainable land use in the long run.
Despite the contributions of our research, there may be some limitations that need to be considered. These include potential uncertainties in data inputs and measurement errors that could affect the accuracy of our findings. Therefore, further research should focus on examining alternative approaches to estimating soil moisture and incorporating prospective land use changes into model calculations, in order to provide more robust and reliable results.

Conclusions
The aim of this study was to underscore the importance of protecting mountainous forests to ensure sustainable soil water availability. Regular evaluations of soil degradation using the SWAT+ model should be conducted. Forest areas have a high potential for water retention, making their protection and appropriate management critical. Coniferous stands, where spruce is the dominant species, are particularly vulnerable to short-term extreme weather events that cause measurable losses. Therefore, emphasis should be placed on the restoration of species adapted to the habitat to prevent harmful erosive problems in the future. Our findings support specific forestry measures that can mitigate desertification. Based on existing forest monitoring, a system for climate change adaptation can be developed by introducing a number of water and soil indicators that are significant for individual catchments. In conclusion, it can be noted that the modeling of evapotranspiration, assessment of soil hydraulic properties, and evaluation of landscape hydric potential highlight the crucial role of forests in climate change policy. However, some limitations of the current study include measurement errors and uncertainties in field measurements and the uncertainty arising from climate data inputs. Several of the points raised here, such as quantifying uncertainty in an integrated modeling framework, were beyond the scope of our research. Future studies should focus on examining the effects of alternative approaches to estimating PET and incorporating prospective land use changes into model calculations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.