LANDSCAPE READING FOR ALPINE RIVERS: A CASE STUDY FROM THE RIVER BIYA

. Anthropogenic stressors have altered the hydromorphological characteristics of rivers worldwide. Environmental guiding principles are essential for planning sustainable river restoration measures. The alpine river Biya, located in the Russian Altai mountains, originates from Lake Teletskoye and joins the Katun near Biysk, forming the Ob. The Biya represents a hydromorphological reference system in anthropogenically ‘least-disturbed’ condition. The presented study aimed to assess the river’s undisturbed morphology in relationship with the geological history of three different river stretches based on an adapted landscape reading approach using remote sensing information (ASTER GDEM v3). The established widths of the active channel, active floodplain and morphological floodplain as well as the longitudinal section were used to explain the differences between upper, middle, and lower Biya. The results confirm differences in the geological origins between the upper Biya, which has previously been described as the least developed and narrowest, and the other two stretches based on the analyses of morphological parameters. Morphological floodplain width could best explain the differences between upper (0-86 km), middle (86-196 km), and lower Biya (196-301 km). The study further showed a clear relationship between the variations in river patterns and adjacent topographic structures (valley confinements, tributary interactions), highlighting that any assessment of river morphology must consider the wider surroundings of a river stretch. The presented morphological observations and analyses of the Biya show that easily obtainable parameters can detect differences in the morphological history of river stretches within the same catchment, supporting process understanding.


INTRODUCTION
Water covers 71% of our planet (USGS 2019) and contributes majorly to the shape and character of the Earth's surface as it travels from source to sea. River (eco-) systems are the setting of critical environmental processes, like sediment or nutrient transport (Newbold et al. 1982) and connect a wide range of different ecosystems and biotic communities (Vannote et al. 1980), acting on a fourdimensional level (Ward 1989). Connectivity throughout whole river systems is necessary for the functioning of all associated processes (Grill et al. 2015). Changing single parameters within a river's catchment can lead to drastic changes in the dynamic equilibrium (Nanson and Huang 2018). Following centuries of anthropogenic interferences, rivers are affected by a multitude of stressors (Lemm et al. 2020). Rivers are among the most heavily degraded ecosystems in the world (Tickner et al. 2020, Malmqvist and Rundle 2002, Sala et al. 2000, putting freshwater megafauna at severe threat , He et al. 2018) and reducing the capacity to further fulfil ecosystem services that human societies rely on (Feio et al. 2022, Abily et al. 2021. The natural regimes of most of the world's larger river systems were altered, and it is assumed that 48% of rivers worldwide are moderately to severely impacted by flow regulation, fragmentation, or both (Grill et al. 2015). Next to these direct anthropogenic impacts, less obvious, indirect changes, like climate or land use change, are continuously transforming hydrological boundary conditions, as well as the sediment cycle, leading to significant alterations in water temperatures and other abiotic habitat factors and consequently destabilising many riverine ecosystems (Liu et al. 2020, Sala et al. 2000. Ecologically intact, or free flowing rivers have become increasingly rare and are largely restricted to remote areas with lower degrees of human development, usually found in snow climates (Feio et al. 2022, Grill et al. 2019. There is an urgent need for action and growing demand to reorganise river management based on an international and interdisciplinary approach so that the functionality of the exposed ecosystems is improved and not compromised any further (Feio et al. 2021, Muhar 1996. It is essential that river restoration approaches build on a sound understanding of the complex ecosystem processes (Tickner et al. 2020), also taking into account channel evolution (Scorpio et al. 2020). Our understanding of river morphological processes and interactions must be continuously advanced since it provides the foundation to combat current shortcomings in river health.
When trying to understand why a river looks the way it does, the focus must be put on the temporal context and the geological history of the river's catchment area: Fluvial development is subject to changes in climate as well as tectonic processes (Vandenberghe et al. 2018) that also interact with sediment dynamics and resulting changes in channel bed elevation (Anderson and Konrad 2019). Processes associated with (past) glaciation, for example, exert influence on many rivers in the temperate climate zone. These processes are majorly responsible for reshaping the landscape by eroding older sediments and reorganising river pathways and the sedimentary system through the stages of glaciation (Sokołowski et al. 2021, Comiti et al. 2019, Fildani et al. 2018. Glacial-interglacial cycles are major drivers of river incision and aggradation (Huang et al. 2019 Malatesta andAvouac 2018). Glacial and periglacial erosion provide vast amounts of sediment (Huang et al. 2019). During the glacial retreat, sediment yields increase dramatically, often causing stark and abrupt changes in the landscape, i.e., paraglacial adjustment (Hedding et al. 2020, Williams andKoppes 2019). Nonfluvial deposits associated with previous glaciation may also play a role in the development of channel patterns in fluvially dominated stretches (Hauer and Pulg 2018). In order to account for the decisive effects of (past) glaciation on river and valley formation, the associated semi-and non-fluvial processes acting in postglacial areas must be better included in the analyses of channel evolution (Hauer and Pulg 2021).
This calls for an integrative analysis of river morphology, including a river's geological history and wider morphological context. This need is met with the development of holistic frameworks for the morphological assessment and classification of river systems that try to take into account all associated processes on a catchment scale (e.g., Rinaldi et al. 2016Rinaldi et al. , 2015. Methods like the landscape reading approach (Fryirs and Brierley 2012) aim to relate large-scale landscape features with the region's geological history. It has been adapted and applied by  to assess paraglacial adjustments and the quaternary development of the river Vjosa highlighting how combined morphological understanding of channel patterns can be used to explain glacially influenced river morphology. These approaches are usually based on large-scale parameters like channel width, or valley width, which are most efficiently assessed using remote sensing techniques.
Remote sensing methods are applied to assess river morphological characteristics across various spatial and temporal scales ( (Bechter et al. 2018) that can help to increase objectivity and comparability (Zhao et al. 2019). They are particularly useful for assessing recent river morphological changes (Langat et al. 2019). Different methodologies have been developed that allow analysing, inter alia, changes in river morphology (Shahrood et al. 2020), river morphological status (Bechter et al. 2018), physical habitat and river health (Zhao et al. 2019), and even river gauging (Hou et al. 2020) based on remote sensing data.
Having established that morphological reference data (e.g., from 'least-disturbed' sites) can serve as valuable environmental guiding principles and support efficient, effective, and sustainable restoration measures (Kujanová and Matouškova 2017, Hey 2006, Newson and Large 2006 while acknowledging that it is next to impossible to find such sites in many regions of the world, like the European Alps (Comiti 2012), we can agree that reference sites must be found elsewhere. 'Least-disturbed' reference sites are not necessarily devoid of any signs of human activity but have suffered the least anthropogenic disturbance within a group of comparable sites (Stoddard et al. 2006). Such regions still exist in less densely populated parts of the world: Anthropogenic impacts in the Altai Mountains, for example, are minimal compared to mountain regions like the European Alps (Volkov et al. 2021). Analyses of such remote areas, especially, rely largely on remotely sensed data.
This study focuses on the morphology of the mountain river Biya in the Russian Altai region (Siberia), which can be regarded as an example of a morphologically leastdisturbed river. This river has been studied regarding hydrological properties (Chalov and Ermakova 2011) as well as its geological origins ). Our study aims to analyse the relationship between the Biya catchment's geological history and its morphological characteristics, described using remotely sensed data based on clearly defined morphological parameters. Following the approach described by , it shall be illustrated how the evolutionary history of the Biya's different river stretches is reflected by their morphology.

The study area
The study at hand focuses on the river Biya, one of the two headstreams of the river Ob. The Biya's catchment lies in the Russian part of the Altai Mountains in Siberia (Schletterer et al. 2021). Siberia accounts for 80% of all global freshwater resources, as three of the world's largest and longest rivers (Lena, Yenisei, and Ob) originate there (Klubnikin et al. 2000). The Altai Mountains are among the least disturbed natural areas worldwide and can be considered a center of biological diversity (Dirin and Madry 2017). The area is particularly important for conserving unique habitats characteristic of the Central Asian Mountain System (Chlachula and Sukhova 2011). The area north of Lake Teletskoye, in particular, is reported to have served as a refuge for deciduous plant species during the last glacial maximum (LGM, Hais et al. 2015).
The Biya's origin (Fig. 1A) lies at the outflow of Lake Teletskoye (51°47'13"N, 87°14'49"E, 430 m a.s.l., near the city of Artybash), which is Russia's second deepest natural lake (Dehandschutter et al. 2002). The Biya covers a distance of 301 km until its confluence with the Katun (52°26'12"N, 85°0'47"E, 160 m a.s.l., near the city of Biysk). It has a catchment of about 36,900 km². Close to the city Biysk, about 21 km upstream of its confluence with the Katun, the Biya has an average annual discharge 1 of 476 m³/s. The chainage information in the following is presented as river kilometres (rkm) starting with 0 at the outflow of Lake Teletskoye. The Biya can be divided into the following three parts: upper Biya (Lake Teletskoye to the mouth of the river Lebed' , rkm 0-86) -middle Biya (river Lebed' to Lebyaj'e, rkm 86-196) -lower Biya (Lebyaj'e to the confluence with the Katun, rkm 196-301; sub-stretch division based on Surface water resources of the USSR 1962). The most important headstream of the Biya is the Chulyshman (Fig. 1A). Fig. 1C includes additional information on grain size distribution from a Wolman pebble count (Wolman 1954) at three sites along the Biya (Artybash, Kebezen, and Biysk; data kindly provided by Friedrich Seidl). The photographs in Fig. 1B give an impression of the conditions on-site.
In order to discuss hydromorphological processes along a river, it is important to know the river's geomorphological history, which is dependent on its geological setting. The geomorphological development of the Biya valley has, however, not been studied as thoroughly as that of other rivers in the Altai mountains, such as the Katun and the Chuya (e.g., Baryshnikov 2016, Zolnikov et al. 2016, Zolnikov et al. 2015. While the Katun still follows its ancient initial course, the Biya, which used to be a small river in the early middle Pleistocene, has grown in length since then: The Biya valley between Lake Teletskoye and the village Turochak (around rkm 75) was formed only in late middle Pleistocene following the beginning of the outflow from the lake. This uppermost part can be described as a mountain river. The lower stretch downstream of Turochak is a well-developed alluvial valley, probably dating back to pre-Quaternary times .
In the Biya valley, distinct traces of large-scale flood events can be found: Giant current ripples bear witness to past megafloods caused by glacial lake outburst (Baryshnikov 2016, referring to Baryshnikov 1979. Moraine dam failure caused the catastrophic Biya debris flow (BDF) which initiated the valley incision processes that led to abrupt morphological changes and the development of the river Biya about 37.5 ka before present (Baryshnikov et al. 2016).

Morphological Characterisation
The hydromorphological characterisation of the river Biya was based on the landscape reading approach described by . This approach aims to develop an understanding of the morphological background of the area of interest (Fryirs and Brierley 2012). It requires the following three parameters: Active Channel (AC), Active Floodplain (AF), and Morphological Floodplain (MF). While AC and MF were determined during the GIS survey, AF was established from a hydraulic 1D step-backwater model (HEC-RAS) of a flood scenario with a five-year-recurrence interval (HQ5) based on . In this approach, AF is defined as the wetted cross-section width in the HQ5 scenario. The HQ5 scenario was determined using the long-time data series (Russia ArcticNET 1 ) for four gauging stations based on a Gumbel distribution (Yue et al. 1999). Channel sinuosity (see below) was added to this set of parameters to allow for a more comprehensive evaluation of the Biya's hydromorphological characteristics. Further differentiation between hydraulic and topographic sinuosity was undertaken to visualise the interactions between the planform river pattern and the surrounding terrain. Next to a geographic characterisation, hydrological and geological boundary conditions were established based on published literature and online data sources (see details below). A supplementary catalogue of hydromorphological characteristics was established for thirty 10 km-stretches (300 km in total) along the Biya. It is presented as an overview table (Appendix: Tables A.1-3), documenting the morphological variability occurring in a least-disturbed river system. A verbal discussion of the changes in channel characteristics in relation to the surrounding topography emphasises the importance of a holistic approach in applying hydromorphological reference parameters. It is targeted that the results may be used as a basis for ecological guiding principles for river restoration planning in the European Alps.

Hydrological Information
Hydrological information for four gauging stations along the river Biya was taken from the platform Russia ArcticNET 2 , where monthly discharge values are given for at least three decades. The gauging stations are located at Artybash (rkm 2), Kebezen (rkm 30), Turochak (rkm 81), and Biysk (rkm 280).

GIS Survey
As a first step, the required geometric information was generated and compiled: For this study, remote sensing data represented the most valuable input. Topographic information from ASTER GDEM v3 (Advanced Spaceborne Thermal Emission and Reflectance Radiometer Global Digital Elevation Model, Version 3 3 ) was used. The geometric attributes needed for this study were established in ArcMap (ESRI: ArcGIS Desktop 10.8.1) based on ASTER GDEM. Inputs for the analysis of planform morphological parameters (i.e., river axis, active channel width, valley cross-sections and valley width) were largely delineated manually, combining topographic information with recent satellite images. A longitudinal section of the river axis was extracted. Special focus in this study was placed on channel sinuosity.

Sinuosity Calculations
Channel sinuosity is an indicator of the meandering intensity of a river, which is commonly evaluated by a sinuosity index (Mueller 1968). The standard sinuosity index (SSI) quantifies the ratio of the channel length to valley length: where CL = channel length, and VL = valley length (here: the average between the length of the left and right valley flank).
The total sinuosity of a river is always a combination of hydraulic (i.e., controlled by the river's flow through its flood plain) and topographic (i.e., determined by the shape of the surrounding landscape) sinuosity (Mueller 1968). These two parameters are quantified by the hydraulic (Eq. 2: HSI) and topographic (Eq. 3: TSI) sinuosity index (calculated based on CL, VL, and linear distance, LD, between starting and end point of each subsection). Per definition, HSI and TSI can be added to a total of 1.
All three sinuosity indices were calculated based on the manually delineated river axis (CL) and visually obtained valley borders. The river axis was divided into subsections of 10 km length for which the calculations were performed.

Statistical Testing
All statistical tests were done in R-4.1.1 (R core team, 2021). Differences between the river stretches regarding the landscape reading elements (AC, AF, MF) were assessed with Wilcoxon rank sum tests. Comparisons between all observed parameters (incl. SSI, HSI/TSI, and slope) were based on regression analyses using average values for stretches of 10 km length.

RESULTS
For the morphological analyses, the established river stretches (upper Biya: rkm 0-86, n = 85 -middle Biya: rkm 86-196, n = 109 -lower Biya: rkm 196-301, n = 108) were kept. AC, AF, and MF were assessed based on a crosssectional view at 1 km increments (n =302). Sinuosity values (SSI, HSI, and TSI) were calculated for stretches of 10 km length (n = 30). The division between upper and middle Biya (and middle and lower Biya) was, in this case, set at rkm 90 (and rkm 200, respectively; n upper = 9; n middle = 11; n lower = 10). The performed analyses (landscape reading elements, hydraulic/topographic sinuosity, and longitudinal profile) show a clear distinction between the upper Biya and the other two sections (Fig. 5), reflecting the differences in the evolutionary history of the upper  Fig. 2A shows the longitudinal section of the Biya from its origin at Lake Teletskoye down to its confluence with the Katun, separated into upper, middle, and lower stretch. Over a length of about 301 km, the Biya travels roughly 270 m downstream at an average slope of 0.9‰ (compare to sectional slope in Table 2). The locations (river station and orographic side) of all 24 tributaries (from Russian Water Register 1 ) are indicated in the graph. Additionally, the mean annual discharge at the four gauging stations (Artybash, Kebezen, Turochak, and Biysk, from Russia ArcticNET 2 ) is included. Fig. 2B illustrates the development of the three landscape reading elements plus SSI values along the river. SSI is highest where MF peaks. A direct comparison of AC and AF suggests lateral dynamics in the upper Biya are restricted compared to the middle and lower stretch. In the upper stretch, the mean differences between AC and AF amount to 157.45 m, on average, which is less than the mean AC width of 226.82 m. In comparison, differences between AC and AF in the middle and lower Biya come to 532.78 m and 677.95 m, respectively, in both cases exceeding the mean AC widths. This suggests a higher influence of the surrounding topography in the upper Biya. Fig. 2C shows the ratio between HSI and TSI. Since HSI and TSI add up to a total of 1, those two values are presented as a stacked bar chart. Hydraulic sinuosity outweighs topographic sinuosity for most of the Biya's length, showing an increasing tendency in downstream direction.
In the upper section (rkm 0-86), the river's valley runs in a rather straight course from south to north (Fig. 3). Over the first 50 km, HSI and MF increase rather steadily. Along the upper Biya SSI and HSI exhibit analogue patterns that correlate visibly with MF values (Fig. 2B and C).
The middle Biya flows towards the northwest over a less mountainous terrain than the upper course. Overall, both valley and river axis show a more complex, winding pattern than in the upper stretch. Three tributary mouths are located within the first ten kilometres of the middle Biya (rkm 86-196). Interestingly, this correlates with an increase in HSI, but not in MF. Near rkm 110, the whole valley turns westward, coinciding with a visible drop in HSI, indicating topographic constraints. A local peak in TSI (rkm 100-110: 80.04%) precedes another peak in HSI (rkm 150-160: 97.91%) which is followed by a reduction in MF (between rkm 160 and rkm 180) and a decrease in HSI that continues until the start of the Biya's lower stretch.
The lower stretch of the Biya (rkm 196-301) starts out with a TSI-dominated section (rkm 200-210: 95.81%), after which SSI, alongside MF, reaches its peak between rkm 210-220 at the confluence between the Biya and its tributary Souskanikha. HSI increases after that local minimum and remains high for most of the remaining lower stretch.
One special focus of the study was the evaluation of channel sinuosity. The goal was to relate sinuosity parameters (SSI, HSI/TSI) to the other morphological key parameters. Fig. 3 shows the SSI values for each 10 kmsection along the Biya in plan view. Only four stretches (rkm 60-70: 1.55 -rkm 150-160: 1.55 -rkm 160-170: 1.53 -rkm 210-220: 1.74) lie above 1.5. These stretches occur in all three sections of the Biya, one each in the upper and lower stretch, and two adjoining 10 km sections in the middle Biya. None of the three stretches differ significantly from one another regarding sinuosity (SSI, Fig. 5D, mean values: upper: 1.21, middle: 1.18, lower:1.23). The number of observations (n between 9 and 11) is, however, quite low for making statistically meaningful statements. There is a wider range of SSI-values occurring in the lower Biya (1 to 1.74), than in the upper stretch (1 to 1.55) with SSI values in the middle stretch ranging from 1.02 to 1.55 ( Fig. 2B and Fig. 5D).
Next to the planform channel pattern (i.e., considering the x and y dimension), the longitudinal elevation of the three stretches of the Biya was analysed. In contrast to the middle and lower stretch, the upper Biya's longitudinal elevation profile can best be approximated by a linear equation (R² = 0.9827), while for the other two, a quadratic polynomial equation provides a better fit (Fig. 4). Fig. 2 shows that the upper, middle, and lower Biya exhibit different morphological characteristics. These are confirmed by the statistical differences between the observed parameters (Fig. 5). Pairwise comparisons between all three stretches were performed using the Wilcoxon rank sum test to test for significant differences MF (Fig. 5A) differs significantly between all three stretches of the Biya. The difference in MF between the upper Biya (mean: 1866.42 m) and both the middle (mean: 3521.16 m) and lower (mean: 4236.82 m) Biya can be described as highly significant (p < 0.001). MF in the middle and upper Biya still differs significantly (p = 0.002). AF (Fig. 5B) within the upper Biya (mean: 384.28 m) also shows highly significant differences (p < 0.001) from the other two stretches. There is, however, no significant difference between the middle (mean: 955.83 m) and the    1). The statistic differences in AC (Fig. 5C) behave similarly: The upper Biya (mean: 226.82 m) differs from the other two stretches at highly significant levels (p < 0.001), while there is no significant difference (p = 0.5) between AC widths within the middle (mean: 423.06 m) and lower Biya (mean: 373.36 m). Within each of the three stretches, MF, AF, and AC all show highly significant differences from one another (p < 0.001).
Regression analyses with all observed parameters (AC, AF, MF, SSI, HSI/TSI, and slope) were performed using data that had been averaged for stretches of 10 km length. Fig. 6 shows a pairwise arrangement of these values plotted against each other. In most cases, no distinct (linear) relationship could be identified.
Based on these initial analyses (Fig. 6), correlations between MF and sinuosity (SSI, and HSI/TSI) were suspected. These three relationships are presented in more detail in Fig. 7. As could be expected, higher SSI values correlate with higher HSI values (and lower TSI values, respectively). Channel 'maturity' (indicated by the ratio between HSI and TSI, with higher HSI values suggesting a more 'mature' stage of the river and a better developed floodplain) is closely related to MF width. Accordingly, higher HSI values were also associated with higher MF values. In all cases, the coefficient of determination (R²) is rather weak: The strongest linear relationship could be identified between SSI and HSI (R² = 0.37). The relationships between MF and SSI (and MF and HSI) are explained by linear regression at R² levels of 0.29 (and 0.19, respectively).

DISCUSSION
The appearance and shape of a natural river result from complex morphological processes and interactions, majorly influenced by geological factors and climatic and hydrological controls (Mahala 2020, Castro and Thorne 2019, Stepinski and Stepinski 2005. These boundary conditions define sediment dynamics and determine the interactions between a river and its surrounding topography (Ibisate et al. 2011). Therefore, a profound understanding of a river's geological history is of utmost importance. Periods of glaciation, for example, that date back several thousands of years have strong impacts on today's rivers since glacial processes have provided large quantities of moraine material that gradually enters the sediment cycle (Sokołowski et al. 2021, Huang et al. 2019. Other processes associated with glaciation include the development of glacial lakes that can accumulate large volumes of lacustrine sediments, which may be released through rivers draining these lakes (Baryshnikov 2016). The temporal scale of the associated processes ranges from long-term deposition of lake sediments to rapid onset release events, like glacial lake outburst floods (GLOFs). The after-effects of certain climatic conditions can extend over millennia, so it is essential to consider the whole geological background when assessing a river's morphology. The evaluation of channel sinuosity alongside morphologically significant width parameters (AC, AF, MF) in this study has highlighted the influence of the different geological histories on the morphological characteristics of the upper Biya compared to the middle and lower stretch. The results confirm a distinction based on morphological history and show how different morphological river characteristics result from a river's geological past.
Glaciation and climatic history of the Altai mountains have been described in great depth (Blomdin et al. 2018, Blomdin et al. 2016, Chernykh et al. 2014, Agatova et al. 2012, Surazakov et al. 2007). The Altai mountains have been the scene of some of the largest freshwater megafloods that are known in planetary history (Herget et al. 2020, Bohorquez et al. 2019, Carling et al. 2002, Rudoy 2002. The most recent glacial activity during the Altai mountains' latter half of the holocene seems to have taken place simultaneously with those in the European Alps (Chernykh et al. 2013). It is reported that the maximum extent of glaciation reached the Biya's origin at Lake Teletskoye (Lehmkuhl et al. 2004). As of 2008, only 6.9 km² of the Biya's catchment are glaciated, with a clear decreasing tendency (Narozhniy and Zemtsov 2011).
The basin formation of Lake Teletskoye was not caused by quaternary glaciation. The Teletsk graben developed very recently (starting in the Pleistocene) in the contact zone between two Paleozoic blocks as a consequence of shear zone reactivation (Dehandschutter et al. 2002, De Grave andVan den Haute 2002). The formation of the actual lake was, however, a consequence of glacial retreat (about 37.5 ka before present), leaving a moraine-dammed lake ). The complex tectonic conditions in the upper Biya valley determine the river's outflow direction (Baryshnikov 2016) thereby acting as controls on the course of the river.
Valley shape and different levels of confinement develop in relationship with various geological processes (Fryirs et al. 2016). Measures of confinement and valley width can be used to infer a river's geological history and to interpret its future development (Fryirs et al. 2016 referring to Brierley 2010 andFryirs 2015). Channel sinuosity is often strongly influenced by topographic constraints like valley margins (Woolderink et al. 2021, Fryirs and Brierley 2010, Tímar 2003, acting as a fundamental control on river character and behaviour (Fryirs et al. 2016). Including channel sinuosity and especially TSI in morphological analyses can help to identify interactions between river geometry and such topographic restraints. At the river Biya, high HSI values were associated with higher MF values (i.e., wider valley cross-sections). In this study, the dependence of channel curvature on topographic constraints (indicated by TSI) was shown to decrease in downstream direction, indicating an increasing stage of the river's 'maturity' .
The upper Biya (between Lake Teletskoye and Turochak) has been described as the least developed and narrowest section  with mountain river character (Baryshnikov 2016). It was only formed in the late middle Pleistocene when the outflow from Lake Teletskoye began. The remaining part, a well-developed valley is much older (probably dating back to the pre-Quaternary period, Baryshnikov et al. 2016). It could be shown that the upper Biya differs significantly from the other two stretches based on morphological parameters (AC, AF, MF). Glacial processes have determined the morphology of the upper Biya . The glacier that used to run through the basin of Lake Teletskoye contributed significantly to sediment production (Baryshnikov 2016). Lacustrine sediment deposits from glacial lakes in the upper Biya's tributary valleys and glacial moraine deposits along the main valley also supplied substantial amounts of sediment. A phase of rapid channel incision of the upper Biya was triggered by a cataclysmic glacial lake outburst event originating from Lake Teletskoye (about 37.5 ka before present, Baryshnikov et al. 2016). This incision into the alluvion is reflected by the limited extent of the active floodplain and restricted lateral dynamics in the upper stretch, indicating a higher sediment supply during the glacial period (Williams and Koppes 2019). The applied method already yielded similar findings at the Vjosa . After the rapid incision triggered by the BDF, valley deepening rates slowed to about 0.8-0.9 m per thousand years . The upper Biya exhibits a linear longitudinal profile, while the middle and lower Biya follow a concave shape best described by a quadratic polynomial equation. The linear profile is associated with an equilibrium between sediment supply and transport without downstream fining. In contrast, the concave longitudinal profile in the lower Biya is an indicator for downstream fining and channel aggradation (Rice andChurch 2001, referring to Mackin 1948). The elevation peaks in the longitudinal profiles (Fig. 4) are most likely caused by uncertainties in the delineation of the channel axis. These errors do, however, not alter the overall characteristic of the longitudinal profile shape of the three sub-stretches.
It could be demonstrated that the upper Biya, which is known to have a different geological background than the middle and lower part , significantly differs from the other two stretches based on morphological parameters (AC, AF, MF). The results clearly show that the relatively simple landscape reading approach, according to , which relies on only three width parameters on a cross-sectional basis, was able to account for the differences in the evolutionary history between the upper Biya and the two remaining sections. This is of particular interest since these parameters are clearly defined and comparably easy to obtain -allowing for a standardised differentiation between river stretches based on their morphological history. The fact that SSI could not be used to explain the differences in channel evolution between the upper Biya and the remaining river stretch was interpreted as an indicator that the focus should rather be placed on the determining factors behind channel sinuosity and include the wider morphological context. Both the landscape reading approach (after ) and the consideration of hydraulic and topographic sinuosity (after Mueller 1968) have proven useful in this regard.
The most striking constraint to the performed channel pattern analyses was the resolution of the input data at a cell size of 30 m by 30 m. Additional uncertainties within the lowermost 70 km must be mentioned, where catchment borders close in on the river axis and the extent of the morphological floodplain is hard to identify.
Comprehensive assessment of river landscape parameters is a key factor for developing a profound understanding of the complex multi-scale processes operating along and within a river system (see also the review of riverscape approaches by Torgersen et al. 2022). This is especially important when establishing environmental guiding principles and aiming to transfer these observations to other, more heavily disturbed systems. For this purpose, a supplementary catalogue of morphological parameters and verbal descriptions could be established for 10 km stretches along the whole Biya. This comprehensive set of parameters is summarised in Tables A.1-3 (see Appendix).

CONCLUSIONS
In this study of the alpine river Biya, the morphological parameters of a river in the least-disturbed condition were assessed and put into context with its geological history. The concept of 'reading the landscape' (Fryirs and Brierley 2012) in a hybrid approach, including measuring and hydrodynamic-numerical modelling ) was applied to differentiate between upper, middle, and lower Biya. The study could confirm the differentiation between the upper Biya and the rest of the river based on the different channel evolution history, highlighting the relationship between geological history and channel morphological parameters. Contrary to the initial expectations, channel sinuosity (SSI) could not be used to divide the Biya into morphologically meaningful sub-stretches. MF could best explain the differentiation between upper, middle, and lower Biya. The statistical analyses showed a relationship between MF and sinuosity (both SSI and HSI/TSI). The results of this study confirm that easily obtainable parameters (AC, AF, and MF) can be used to detect differences in the morphological history of river stretches, as was already shown for the Vjosa ).