Using hydro‐morphological assessment parameters to estimate the flood‐induced vulnerability of watercourses ‐ a methodological approach across three spatial scales in Germany and the Czech Republic

In addition to their ecological importance, rivers and streams have always been used in diverse ways by humans, resulting in the development of settlements and their connected built environments along many of the world's watercourses. During heavy rainfall, buildings, traffic infrastructure and water‐related infrastructure are exposed to potential hazards in the form of (flash) floods. In contrast to near‐natural watercourses, anthropogenically modified channels in urban areas are particularly susceptible to damage by flooding. Previous damage assessments have highlighted the need to forecast such damage to watercourses in order to identify critical areas and justify the selection and expansion of adaptation measures. Within the scope of the current study, we have developed a method based on the hydro‐morphological properties of watercourses to make transferable estimates of the economic damage potential based on ecologically‐relevant parameters. Using a scale‐specific cause‐effect analysis, we have identified characteristics of the watercourse type and adjacent structures as well as construction‐related properties of reinforcements that can increase the damage potential during flooding. In this way, we are able to show that several influencing factors determine the vulnerability of watercourses: in addition to the specific longitudinal gradient and size (macroscale) of various watercourse types, damage‐relevant boundary conditions in watercourse sections (mesoscale) and the resistance of typical bed and bank constructions are also important, reflecting the specific structural conditions. Taking rivers in Germany and the Czech Republic as case studies, in the following, we review the local identification of critical areas and describe the necessary data management. The presented “Hydro‐morphological based Vulnerability Assessment‐Concept (HyVAC)” can contribute to the flood damage prevention at watercourses by utilizing existing basic data to the greatest possible extent and thus is suitable for preliminary investigations according to the EC Flood Risk Management Directive.

Accordingly, this study suggests a wide-scale methodology to identify particularly vulnerable areas of watercourses in a step-by-step manner, from which it is possible to derive measures for damage prevention.
Since we adopt an application-oriented approach, the necessary data requirements to represent the spatial occurrence will also be detailed alongside the theoretical-methodological framework. Here the interpretation of existing remote-sensing data, as well as hydro-morphological assessment parameters of watercourses, are key factors (readily available information according to EC FD) in assessing the flood-induced vulnerability of watercourses. This also ensures the transferability of the method. The described steps thus characterize our so-called Hydromorphological based Vulnerability Assessment-Concept (HyVAC).

| Concept of the assessment method
In the context of vulnerability research, this paper views watercourses as parts of the built environment and draws on methodological principles of vulnerability. For this purpose, the conceptual understanding of hazard, spatial occurrence, and vulnerability are applied according to their common usage in research on natural hazards and risk. Here flood risk represents the degree of potential damage to exposed receptors under specific impact scenarios (hazards) and the associated ranges of impact magnitudes (for a more detailed discussion of terms, see Cutter, 1996, Adger, 2006, Birkmann et al., 2011, Paul, 2013, Schanze, 2016 The potential economic impact of flood events on watercourses can be described by means of three spatial scales in the case-study regions (Figure 1). At the macroscale, climatic influences (heavy and continuous rainfall) cause specific precipitation-runoff processes depending on the ecoregion and catchment size. In turn, the flow force of the water increases as a function of the flow rate. At the mesoscale, the extent of damage to the watercourse during flooding will depend on settlement patterns, adjacent infrastructure (e.g., roads) or certain forms of land use (e.g., agriculture) along the watercourse. In this respect, the presence and resistance of these damage-relevant structures must be taken into account. At the microscale, namely, the level of individual objects, the type of bank and bed constructions must be characterized by their location, whereby previous damage influences the level of vulnerability.
HyVAC combines existing sources of data on watercourses from the fields of hydrology, ecology, geomorphology (macroscale) and land use (mesoscale), as well as hydro-morphology, hydraulic engineering and channel hydraulics (microscale). Once the parameters have been specified, they can be used to assess the vulnerability of watercourses when flooded.

| Analysis of empirical base data in Germany and the Czech Republic
Within the framework of data screening, we analyzed existing German and Czech forms of assessment and classification of watercourses to derive relevant information and to examine their ease of transferability. The transnational approach is justified by the joint river basin management of the Elbe and the related cooperation of the two countries in flood risk management. The analysis showed that Germany follows the system B) according to EC-WFD (EC, 2000) to classify 25 watercourse types based on eco-regional, hydro-morphological, biological and chemical-physical parameters (Briem, 2003;Pottgiesser, 2018). In Czechia, system A) of the WFD (EC, 2000) is used to classify 11 watercourse types into seven groups according to the ecoregion, elevation, catchment size, and geology (Kujanová, Matoušková, & Kliment, 2016). Hence, the Czech approach is based on a somewhat coarser eco-morphological classification of watercourses.
In Germany, the hydro-morphology of watercourses is recorded using the so-called "hydro-morphological survey protocol" (Gellert, Pottgiesser, & Euler, 2014;LANUV NRW, 2012;LAWA, 2019). On the basis of typological references (Pottgiesser, 2018, UBA, 2014, 34 individual parameters are recorded for the watercourse, the channel and the adjacent land use of the floodplain (some parameters may be slightly adjusted depending on the federal state). In Czechia, two methods are employed for hydro-morphology assessment. The first simplified ("s") assessment of streams focuses on small-and mediumsized watercourses (Matoušková, 2008;MŽP ČR, 2009). The second extended ("e") method is hydro-ecological monitoring according to Langhammer (2014). In the simplified methodology, data on the channel and floodplain is captured separately; in each case, hydromorphological criteria are recorded as separate indicators. The assessment of the channel makes use of four criteria and 17 indicators, whereas three criteria and six indicators are used for the floodplain (MŽP ČR, 2009). Following the detailed methodology ("Metodika monitoringu hydromorfologických", HEM), data is recorded separately for the channel, extended bank and floodplain on the basis of 18 parameters (Langhammer, 2014).

Our analysis of the individual German and Czech parameters and
indicators shows that individual hydro-morphological assessment parameters can be used to identify potentially occurring damage mechanisms and vulnerable infrastructure. For example, the presence of a bank wall (D: EP 5.2, CZ: Indicator 3.4 (s) and 12 (UBR), respectively) can, on the one hand, be assessed as an artificial structure in terms of water ecology; on the other hand, we also understand the bank wall as a flood protection device which is associated with a damage potential. Therefore, in the following, we make use of the parameters identified in the German and Czech evaluation systems at the three considered spatial scales. However, the method is also transferable to other countries (see Belletti et al., 2015;Gostner, 2019;Tomšová, 2013

| Macroscale: Types of watercourses
Flood or heavy rainfall hazards generally arise in catchments with high relief energy and large areas of sealed or only slightly effective retention areas (e.g., Beckers et al., 2013;Yigzaw, Hossain, & Kalyanapu, 2013). Damage results, in particular, from dynamized flooding processes where streambeds are more steeply inclined (Wharton, 1992). Therefore, we identified the slope of the valley bottom and the channel cross-section (as a measure of discharge) as key influencing variables. According to the Gaukler-Manning-Strickler formula, these parameters also determine the flow velocity: where v m is mean flow velocity in m/s, k st the coefficient of channel roughness in m 1/3 /s, R the hydraulic radius and I the bottom line slope ($bed slope) in %.
In this context, straightened watercourse sections with high bed slope in intensively used and populated or urban landscapes have a particularly high damage potential (Bornschein & Pohl, 2018;Hartmann, Jílková, & Schanze, 2018;Jordan, Annable, Watson, & Sen, 2010). Accordingly, hydraulic stress on streams varies with the watercourse size and adjacent land use (Brierley & Fryirs, 2005;Buffington & Montgomery, 2013;Buraas et al., 2014;Knighton, 1999;Newson, Clark, Sear, & Brookes, 1998;Vocal Ferencevic & Ashmore, 2011;Wharton, 1992). While runoff dynamics are particularly pronounced in the catchments of small mountain streams albeit with lower absolute flows, enormous damage can occur on mid-mountain streams due to the higher absolute discharge rates and associated larger flow pulses (Bjerklie, 2007;Bryndal, Franczak, Kroczak, Cabaj, & Kołodziej, 2017). For this purpose, Knighton (1999), for example, described a method to estimate the erosion dynamics of streams as a function of their catchment size, stream length and width. This approach was later adopted and expanded by a number of authors to calculate the (specific) stream power (Buraas et al., 2014;MacBroom et al., 2017;Vocal Ferencevic & Ashmore, 2011).
This stream power seems to be a suitable measure of the impact intensity since the water body and catchment size are integrated via the discharge volume. The variable can be calculated as: where ω is the specific stream power in W/m 2 , γ the specific weight of water in g/cm 3 , Q the discharge in m 3 /s, S e the bed slope in m/m and w the channel width in meters. Referring to the German and T A B L E 1 Classification of potential hydraulic vulnerability based on stream types according to Pottgiesser (2018) for Germany and Kujanová et al. (2016) for the Czech Republic. Valley floor slope and size class were taken from the respective watercourse profiles (D) and method descriptions (CZ) Czech case studies, we classified the respective stream types (Kujanová et al., 2016;Kujanová & Matoušková, 2017;Matoušková, 2008;Pottgiesser, 2018) by their typical mean valley floor slope. Based on the characteristics of the mean slope, we derived five slope classes with specific threshold values. Considering their size and designation, we also divided the watercourses into five classes: small streams, streams, rivers, large rivers and main rivers.
Based on this systematization, potential type-specific impact intensities on watercourses could be derived for a large-scale overview ( Table 1). The resulting index "watercourse type-specific vulnerability" (WTSV index) represents a first evaluation step in our methodology.

| Mesoscale: Section classification
A further spatial specification of watercourses is needed to character-   to bed or bank areas (Montgomery & Buffington, 1993, MacBroom 2017. Due to these potentially occurring damage mechanisms, we integrated "transverse and crossing structures" and the "mouth of tributaries" as punctual boundary conditions.
To identify linear structures, we used event analyses as well as the results of an extended literature review of model investigations.
Thereby, the in-channel flow direction could be highlighted as a particular influencing variable (e.g., Song et al., 2018). Furthermore, several authors have shown that the condition and use of the riverbanks, in particular, influences the resistance of each section to hydraulic impacts (Bridge & Jarvis, 1977, Jin, Steffler, & Hicks, 1990, Miller, 1995, Wharton, 1995, Khatua & Patra, 2007, Terrier, Robinson, Shiono, Paquier, & Ishigaki, 2010, Buraas et al., 2014, Ghobadian, Tabar, & Koochak, 2016, MacBroom 2017. From this, we derived the linear boundary conditions "channel geometry", "special bank pressures" and "location in the channel". If stabilization structures are necessary, they must be introduced into the channel in a hydraulically favorable way. Transitional areas exist at the intersection of different forms of stabilization and when channel geometries change (e.g., from trapezoidal profile to natural profile). Such areas are subject to particularly high hydrodynamic stresses; if not properly designed or maintained, damage mechanisms can be enhanced, resulting in damage during flood events (Wharton, 1992, Hajdukiewicz et al., 2016, MacBroom 2017. Because of these relationships, we incorporate the linear boundary conditions of "construction changes" and "geometry transitions" when assessing the vulnerability of sections. Backwater areas are formed by a sudden change in bedline slopes, such as caused by transverse structures. During flooding, considerable sediment mobilization or sediment accumulation can occur here (Hajdukiewicz et al., 2016;Wicherski, Dethier, & Ouimet, 2017). As such artificially intensified flood potential can also damage the watercourse, these "backwater areas" must be taken into account as a linear damagerelevant boundary condition. In order to prevent damaging processes, it is important that regular and especially ecologically-oriented T A B L E 2 Damage-relevant punctual, linear and planar boundary conditions at watercourses identical for Germany and Czech Republic (white background); German method (light grey background); Czech method; (s), simple; (e), extended, (dark grey background); SP, single parameter; RS, remote-sensing; m, meters

Boundary condition Database
Vulnerability class of the watercourse section due to the specification of a boundary condition backwash and the resulting restrictions on the stability of infrastructures in the channel (see Figure 2). In particular, we represent these aspects by the planar boundary conditions "land use" and "harmful land features".

(B) Databases
To obtain potential parameters to assess the above-mentioned damage-relevant boundary conditions, we used general characteristics of the watercourse section (remote-sensing/watercourse network) and specific characteristics of individual parameters of the watercourses in Germany and the Czech Republic (Langhammer, 2014;LANUV NRW, 2012;MŽP CR, 2009). Next, we interpreted the characteristics of selected individual parameters to determine whether their spatial occurrence could result in damage processes during a flood event. Some features such as the "location in the channel" or the "mouth of tributaries" can be taken from remote-sensing data, geo-information systems or detected with the help of aerial photographs (see Langhammer, 2014). be mentioned that only structures whose distance is assessed as "low" have been included in the calculation, considering SP 6.3.

| Microscale: Construction types
Based on the mesoscale observation in sections, we conducted the microscale assessment of the vulnerability of watercourses at the object level. Within a watercourse section, there may be different types of bed or bank protection. While these are subject to the same hydrodynamic pressures in the event of flooding, they will present different levels of resistance due to their diverse types of construction.
In order to characterize the behavior of construction types under certain impact situations, the authors prepared (A) an overview of construction types and databases, and then (B) an evaluation of the resistance of the construction types.

(A) Construction types and databases
Based on a comprehensive literature review of construction types in channel hydraulics, we found a huge diversity of possible bank and bed protection. However, there is a lack of detailed information on susceptibility to damage during flooding. For this reason, we grouped individual types of construction and examined information on the stress limits of the subordinate groups. Basically, it is possible to distinguish between solid construction types, fill construction types,

Boundary condition Database
Vulnerability class of the watercourse section due to the specification of a boundary condition near-natural construction types, as well as combined construction types (bio-engineering). The authors pursued the idea of using previously existing parameters from the hydro-morphological survey protocol (see Section 2.2) to derive a suitable classification of construction types. The classification is thus based on the German and Czech recording methods (Langhammer, 2014, LANUV NRW, 2012, MŽP CR, 2009). On the German side, we used the individual parameters "3.3 -bed protection" and "5.2 -bank protection" for the bed and bank areas (Table 3). On the Czech side, we identified the suitable indicators "3.4 -bank protection left", "3.5 -bank protection right" and "3.6 -bed protection" (simplified methodology) as well as "6-UDN" and "12-UBR" (extended methodology). Making use of 21 classes (MŽP CR, 2009), the simplified methodology differentiates between significantly more potential types of bank and bed protection than the German approach, which uses 14 classes. Table 3 shows the classification of the construction types as derived by the authors.  Table 3). Other authors such as Zeh (2007) (2018), further information on near-natural riparian features such as reed beds, riparian scrub, or woody galleries can be found, which find application in the Czech survey methodology. We summarized these bio-engineered and nature-based construction types before assessing them in terms of their resistance (see Table 3). Compilations of the resistance of different types of structure based on the respective impact variables are also given by Florineth (1993), LfU (1996), Schiechtl and Stern (1997), Oplatka (1998), Bollrich (2013, and Patt, Jürging, and Kraus (2018). These compilations are also integrated into our derivations. In Table 3, the construction types identified on the German and Czech sides are classified in terms of their resistance.  Magilligan, 1992, LfU, 1996, Oplatka, 1998, Knighton, 1999, Gerstgraser 2000a& 2000b, Schillinger, 2001, SMUL, 2005, Julian & Torres, 2006, Nachtnebel, 2008, Feldmann, 2009, MŽP CR, 2009, Krapesch, Hauer, & Habersack, 2011, LANUV NRW, 2012, Sin, Thornton, Cox, & Abt, 2012, Vocal Ferencevic & Ashmore, 2011, Buraas et al., 2014, Langhammer, 2014, EFIB, 2015, Marchi et al., 2016, MacBroom 2017 No Based on a further literature screening (Davis & Harden, 2014;Gerstgraser, 2000aGerstgraser, & 2000bHopkinson & Wynn-Thompson, 2016;Klösch et al., 2018;Kolb, 1979;Magilligan, 1992;Park, Kim, Park, Jo, & Kang, 2016;Sin et al., 2012), we identified flow velocity and shear stress to be governing impact variables. A number of articles also rely on the impact variables of absolute and specific stream power presented earlier in Section 3.2. These studies describe how flood events affect stream hydro-morphology and ecology (Anderson, Rizzo, Huston, & Dewoolkar, 2017;Bizzi & Lerner, 2015;Hajdukiewicz et al., 2016;Hickin & Nanson, 1984;Knighton, 1999;Krapesch et al., 2011;Lague, 2014;Magilligan, 1992;Marchi et al., 2016;Miller, 1990;Thompson & Croke, 2013;Vocal Ferencevic & Ashmore, 2011). envisioned the introduction of so-called "structure damage codes" for streams. We follow these approaches by assuming that resistance decreases as a function of the structural condition of bed and bank construction types. Accordingly, vulnerability at the microscale level is assessed by classifying the watercourse condition.   Magilligan, 1992, LfU, 1996, Oplatka, 1998, Knighton, 1999, Gerstgraser 2000a& 2000b, Schillinger, 2001, SMUL, 2005, Julian & Torres, 2006, Nachtnebel, 2008, Feldmann, 2009, Krapesch et al., 2011, Sin et al., 2012, Vocal Ferencevic & Ashmore, 2011, Buraas et al., 2014, Song et al. 2018 T A B L E 5 Classification of the structural condition of construction types based on condition classes, including extent of damage and verbal description (indicator description 3.8 from Czech method in brackets); pictures: 1: Garack, 2013 Tables 4 and 5 Vulnerability of construction (V c ) Condition class of construction (Table 5) 1 2 3 4 5 Resistance class of construction (Table 4)  We determined the vulnerability of bed and bank construction types by overlaying resistance (see Table 4) and structural condition (see Table 5) in the form of an evaluation matrix ( 3.5 | HyVAC methodological framework and calculation method 3.5.1 | Methodological framework Figure 3 summarizes the methodological framework for determining the vulnerability of watercourses to flooding at the three spatial scales described. Regarding the macroscale classification of ecoregions, it is already evident that on the German side, three ecoregions influence the vulnerability assessment, whereas on the Czech side almost the entire national territory is assigned to the region "Central Highlands". Table 7 summarizes the relevant assessment parameters at the three spatial scales with the respective influences on the vulnerability of watercourses (derived by impact analysis). The assessment was initially carried out separately in the form of a qualitative assessment at the macroscale and a quantitative assessment at the mesoscale and microscale. After integrating the quantitative assessments at these last two scales, the result was combined with the qualitative assessment specific to the type of watercourse. In so doing, we are able to take these characteristics into account when deriving suitable adaptation measures with the help of the index "watercourse type-specific vulnerability" (WTSV index, see Table 1). The application of our HyVAC method to the German and Czech case studies will now be described in the following sections.

| Quantitative calculation method across scales
We determine the vulnerability of the watercourse section V S by combining the vulnerabilities of the watercourse bed and the left and right bank, taking into account the boundary conditions, according to Equation (3): F I G U R E 3 Methodological framework and visualization of the three spatial scales used to assess the vulnerability of watercourses [Color figure can be viewed at wileyonlinelibrary.com] with, V … vulnerability.
S … section of the watercourse.
IA … impact area (bed, left bank, right bank).
The respective impact of the boundary conditions is taken into account for assessing vulnerability in the watercourse section. The characteristics in Table 2 are used to evaluate a boundary condition. In addition, the boundary conditions are assigned to the area of impact in the channel, that is, whether harmful effects on the bank or bed construction are to be assessed ("bed", "respective banks"; see Table 8). Therefore, the vulnerabilities of the two influencing variables of the immediate impact area (IA) and the T A B L E 7 Comparison of watercourse characteristics for hydro-morphological assessment and assessment of flood-induced vulnerability of watercourses (impact and parameters): qualitative assessment (macroscale) and quantitative assessment (mesoscale and microscale)  (4): with, BC … boundary conditions.
The influence of the damage-relevant punctual, linear and planar boundary conditions is integrated with different weighting when determining the vulnerability (see Section 3.3, Table 2). It should be noted that not all boundary conditions influence all three impact areas (see Table 8).
For example, the backwater areas (L 5 ) only influence the bed and not the banks. This is expressed by an influential factor in Equation (5). In the case of L 5 , for example, the influential factor is equal to 1 for the calculation of the bed and zero for the calculation of the banks: with, b … boundary condition (Punctual, Planar, Linear: b = Pu 1 , Pu 2 , Pl 1 , Pl 2 , L 1 , L 2 , L 3 , L 4 , L 5 , L 6 , L 7 ).
k i … influential factor (boundary condition influences impact area: w b … weighting factor (moderate influence of boundary condition: On the watercourse banks, all boundary conditions can have a damage-relevant effect, with the exception of L 5 (backwater areas), so that k i,L5 = 0 applies here. The boundary conditions Pu 1 , L 3 , L 4 , L 5 and L 7 can potentially increase the damage to the watercourse bed so that the influential factor k i is set to 1 in these cases. The boundary conditions of transverse and crossing structures (P 1 ), the location in the water body (L 1 ) and the channel geometry (L 2 ) have a particularly strong effect and are therefore included in the above Equation (4) with a weighting factor w b of 2.
The vulnerability of the impact area of the watercourse section, which is included in Equation (3) at the mesoscale, is determined by combining the vulnerabilities of individual construction types within this section, which are determined individually at the microscale. The integration is carried out using a worst-case approach according to Equation (6). This means that the construction type with the highest vulnerability primarily determines the aggregated vulnerability of the respective impact area.
with, C n … construction type.
n … number of different construction types in the impact area.

| Integration of quantitative and qualitative assessment method
To integrate the quantitative assessment results into the qualitative assessment of the macroscale watercourse type-specific vulnerability, we combine V s (Equation (3)) and WTSV ( Consequently, the damage potential in watercourses with a higher WTSV index is also higher. The schematic comparison of V s and WTSV is shown in Figure 4, where red areas indicate major adaptation measures and blue areas represent less extensive adaptation measures.

| Case study areas in Germany and the Czech Republic
As part of the transnational INTERREG VA research project "STRIMA II" on Saxon-Czech flood risk management, a comparison was undertaken of methods and case studies in these two countries.
In each case, small and medium-sized watercourses (10-1,000 km 2 catchment area) were studied. In Germany, the focus was on a medium-sized watercourse, namely the Müglitz in the Eastern Ore Mountains (Osterzgebirge), which in its upper reaches also runs a few kilometers through Czechia. On the Czech side, the small town of Frýdlant in the Liberec district was chosen for the study due to its tendency to suffer flooding (most recently, in 2010 and 2011).
Here the Řasnice, a small watercourse, was the object of investigation ( Figure 5).  Table 2). The current vulnerability will be assessed on the basis of an exemplary section.

| Application of the HyVAC method: Qualitative vulnerability assessment of watercourses at the macroscale
To assess vulnerability at the macroscale, the Müglitz and the Řasnice are classified according to Table 1. In Table 9, in addition to a general overview, we present the assessment approaches applied to these watercourses. In order to compare the two river sections at a macroscale level of vulnerability, it is crucial to know the watercourse type of the Müglitz and Řasnice and their vulnerability classes as specified in Table 1. According to Table 1 and Table 9, the Müglitz can be assigned a "very high" watercourse type-specific vulnerability (type 5); the Řasnice, on the other hand, is a watercourse type C4 and thus has a "medium" watercourse type-specific vulnerability. To distinguish these watercourses, the vulnerability of the water body section can be specified with the help of the WTSV index (vh: very high, h: high, are determined at the individual scale levels and then integrated step by step using the calculation rule described in Section 3.5.2 before transferring them to the next scale in each case according to Section 3.5.3. Figure 6 illustrates the step-by-step determination of the vulnerability of an impact area in a watercourse section.

| Watercourse type-specific vulnerability
The assessment methodology begins with step 1, namely the quantitative determination of the watercourse type-specific vulnerability (WTSV index, see Section 3.2, Figure 3 and Section 4.4, Figure 6), which is calculated from the slope class (step 1a) and the discharge class (step 1b). The investigated section of the Müglitz has a slope class of 5 and a discharge class of 2. Using the evaluation matrix in Table 9 gives a WTSV index of "very high". For the Řasnice, a slope class of 3 and a discharge class of 2 give a WTSV index of "medium". Accordingly, the section of the Müglitz shows a very high damage potential while the section of the Řasnice has a medium damage potential.

| Vulnerability due to damage-relevant boundary conditions
In step 2, the potentially damage-relevant boundary conditions at the considered watercourse sections are classified according to Table 2 in Section 3.3 (see also Figures 3 and 6). The assessment is carried out separately for each impact area (watercourse bed and banks).  (5). Equation (7) illustrates this procedure for the left bank of the Müglitz: with, w b … weighting factor of boundary conditions. b.
(influential factor k i = 1 in all cases shown and therefore omitted here) Calculation of the vulnerabilities due to the boundary conditions gave a value of 2.15 for the right bank of the Müglitz and 2.00 for the bed, respectively. In an analogous way, the vulnerabilities due to the boundary conditions for the Řasnice were determined as 2.33 for the bed, 2.31 for the left bank and 3.15 for the right bank.

| Vulnerability of construction types in the impact areas (watercourse bed and banks)
In step 3a of the assessment methodology (see Figures 3 and 6), the resistance of the construction types in the watercourse section to be investigated is determined under the assumption that the construction is in perfect condition. Here also the assessment is carried out separately for each impact area, namely the watercourse bed and banks. The resistance classes of the construction types were assigned to the construction types found in the watercourse sections using the T A B L E 9 Vulnerability of watercourse types in Germany/Czechia and WTSV index of the case study watercourses  In order to move from resistance to an assessment of the vulnerability of the construction types, the structural condition of the protection construction types must be examined in the following step.
Therefore, in step 3b, the structural conditions are classified according to the scheme shown in Table 5 (see Section 3.4). For the example of the Müglitz, the wall of the left bank of the watercourse shows superficial damage. The masonry structure was thus assigned to condition class 2 according to Table 5. The right walled bank of the Řasnice was assigned to the same class.
Using the evaluation matrix (Table 6) Evaluations carried out in an analogous manner of the construc-  Figure 6a.
For the other impact areas of the Müglitz and also the Řasnice, the vulnerability due to the boundary conditions is higher than the originally determined vulnerability of the construction types, thus leading to a higher vulnerability when considering the extended impact area.
The vulnerability of the bed of the Müglitz in the investigated watercourse section is 2.00 and that of the right bank 2.08, taking into account the (extended) impact area. In the same manner, values of 2.58 for the right bank (see Figure 6b), 2.15 for the left bank and 1.67 for the bed were determined for the investigated section of the Řasnice.
In step 4a, we determine the vulnerability of the watercourse section. For this purpose, the calculated vulnerabilities of the different impact areas at the mesoscale are combined into an average value according to Equation (3). For the Müglitz, the vulnerability of the investigated watercourse section is calculated as follows: T A B L E 1 0 Section-specific assignment of damage-relevant boundary conditions for the Müglitz and the Řasnice (values in bold are used in the following exemplary calculations) Similarly, the vulnerability of the investigated section of the Řasnice is 2.13, a slightly lower value than for the Müglitz.

| Transferability of the HyVAC method
Based on the investigated case studies, the German and Czech methods for assessing the hydro-morphology of watercourses can be judged highly suitable for assessing flood-induced vulnerability.
Regarding the transferability of the two methods, it was found that while they use fundamentally similar approaches, there are differences in the three spatial scales. At the macroscale, the German watercourse classification appears more detailed and differentiated, a fact also reflected in the naming of the stream types (which make reference to the substrate and local geology). On the other hand, the eco-regional diversity in the Czech case (mainly "central highlands") does not compare to that on the German side ("lowland", "central highlands", "Alps"); this implies a higher diversity of watercourse types in Germany. At the level of the highlands, the two countries show comparable types of watercourses. However, the ranges of gradient and discharge classes appear somewhat coarser on the Czech side, which is also evident in the watercourse classification (Table 1).
At the mesoscale, the two national methods are comparable and well suited to assessing the vulnerability of watercourse sections (see Table 2). Here, the specificity of the "simplified" and "extended" methods on the Czech side is clearly an advantage. For the boundary condition P 1 , however, it would be useful to have a classification of bridges and culverts comparable to that of the German method. The same applies to the "harmful" parameter of the German method ("Harmful land features"), as this would enable an improved derivation of damage processes occurring in the channel (concerning L 6 ). Regarding land use, the Czech methodology could be refined by defining and recording the distances of structures from the channel (small/ medium/large) (Pl 1 and Pl 2 ). In the methodology we have presented, only structures whose distance is assessed as "small" are included in the calculation. Further, in Germany, a distinction is made between unpaved and paved traffic areas, which is also relevant to the likelihood of erosion in the case of overflow.
At the microscale, it is important to highlight the more differentiated Czech methodology, which allows a particularly detailed recording of construction types on the bed and banks, even though the types "sheet-piling", "training works" and "groyne" are missing (see Table 3). Another positive aspect of the Czech survey methodology is the obligatory assessment of the condition of bank and bed constructions, which in the context of watercourse maintenance not only shows the relevance of potential damage but also gives an estimate of the required maintenance. For more detailed comparisons of hydromorphological survey methods, please also refer to the work of Belletti et al. (2015) or Kampa and Bussettini (2018). The problem of missing data on the hydro-morphology of the Řasnice could be compensated on the Czech side by an independent subdivision into 100-m sections and on-site mapping according to the German survey methodology. Here, the authors were able to draw on their own extensive mapping experience. During the survey, it became apparent that a further subdivision of construction types, following the simplified Czech survey method, would help specify the degree of obstruction (damage potential) in urban watercourses. If necessary, these findings should be taken into account in a revision of the German survey method.

| CONCLUSIONS AND FUTURE RESEARCH
The investigations undertaken in this study confirm the suitability of deriving damage-relevant parameters from data on the hydromorphological condition of watercourses. Based on a comparison of German and Czech data acquisition methods, we were able to show which parameters are relevant at the macro-, meso-and micro-scale.
At the macroscale, the ecoregion, catchment size and associated length of watercourses are of particular importance in the international context. Studies by Knighton (1999), Vocal Ferencevic and Ashmore (2011)  ther results from field studies or laboratory investigations would be of great interest (see Bjerklie, 2007, MacBroom et al., 2017. The presented methods also offer the possibility of an even smaller-scale estimation of the damage potential based on the recording of channel dimensions. This approach could be refined in the future through the complementary use of remote-sensing data (see Bjerklie, 2007). The construction types themselves can also make an additional contribution to resistance due to their typical flow-relevant surface properties. For each construction type, the surface roughness is a key influencing factor for flow conditions, shear stress, and flow velocities (micro-and macro-roughness, see Hurson & Biron, 2019).
Here the classification of surface roughness could provide more details on the erosion and abrasion resistance of differently shaped streams, as the processes determining flow resistance are dependent on the roughness scale (Carey, Stone, Norman, & Shilton, 2015;Sabrowski, 2008). Furthermore, the material bonding within each structure is a critical factor in the development of damage mechanisms. To assess impact resistance to entrained sediments, debris, and alluvium, it is necessary to consider the type of material composite of the construction types. Accordingly, assumptions could be made on impact resistance in terms of how strong the impulse must be to dislodge individual elements from the structure (see McBride et al., 2007;Suaznabar et al., 2017). In this context, the duration of certain flow stresses is also relevant. Again, more research is needed on the resistance of structures as a function of their design as well as on the applied flow forces and durations.
As an indication of potentially stressed areas of watercourses, the macroscale classification of watercourse types according to the presented HyVAC method can be usefully applied since certain design features can be derived from construction types and adjacent land use.
Thus, in the example of the Müglitz, particularly robust construction types with large armoring stones are to be preferred due to the vulnerability of this type of watercourse, whereas in the example of the Řasnice, it is clear that excessively massive armoring is unnecessary, even if robust construction types are needed along certain sections.
This example shows the application-oriented character of our methodology, whereby, on the basis of the presented scales, suitable levels of action and planning are addressed in each case. Furthermore, the interdisciplinary approach adopted here is intended to once again highlight the potential of existing databases for synergetic research.

ACKNOWLEDGEMENTS
The article was written as part of the EU-funded research project on Saxon-Czech flood risk management (STRIMA II, funding program "Ahoj sousede. Hallo Nachbar". IINTERREG VA/2014-2020). We gratefully acknowledge the support of the Saxon State Office for Environment, Agriculture and Geology (LfULG) and Agentura regionálního rozvoje, spol. s r.o. (ARR) for the provision of data and good cooperation.

DATA AVAILABILITY STATEMENT
Excel calculations that support the findings of this study presented in the charts are available from the corresponding author upon reasonable request.