13.1 Introduction

Hydropeaking alters hydraulic conditions as well as wetted areas and is disadvantageous for fish species with specific hydraulic preferences, in particular for spawning grounds and fish stages with restricted mobility such as juvenile fish and larvae (Moreira et al. 2019). Habitat models describe the environmental conditions for fish and use the requirements of fish related to these conditions to calculate habitat suitability. Moreover, habitat modelling is also an appropriate tool to quantify the impacts of hydropeaking since these impacts can be interpreted as a decrease of habitat suitability. In contrast to the standard habitat parameters as water depth, flow velocity and granulometry, the temporal change of habitat conditions and the speed of this temporal change are highly relevant for hydropeaking analyses. The habitat model system CASiMiR has been extended by a hydropeaking module to take account of these impacts (Schneider and Kopecki 2016). More precisely the following features have been integrated:

  • downramping rates, stranding risk and spatial integration

  • upramping rates, drift risk and spatial integration

  • redd stability

  • habitat shift and habitat persistence

These model features serve on one hand as tools for the detailed hydraulic-based assessment of hydropeaking scenarios and mitigation measures. On the other hand, they provide input parameters for a matrix-based assessment of hydropeaking impact strength (Boavida et al. 2020).

13.2 Testcase GKI (Inn, Austria)

The upper Inn River in Tyrol is affected by hydropeaking from the hydropower scheme Pardella-Martina in Switzerland (Meier 1991). The diversion hydropower plant GKI (Gemeinschaftskraftwerk Inn) has been designed to mitigate hydropeaking along the stretch between the villages Ovella and Prutz (Herdina 2018). This Testcase enabled the comparison between the impacted and mitigated situation. The newly developed CASiMiR hydropeaking module, covering different risks for fish habitats arising from the rapid flow changes has been applied in the Testcase GKI to quantify and assess the ecological impacts of different hydropeaking events. Analyses were carried out in three morphologically different river stretches: the homogeneous channel-like reach Kajetansbrücke, the more heterogeneous reach Mariastein 1 with alternating gravel bars and the most heterogeneous reach Mariastein 2 with characteristics of a braided river (see Fig. 13.1).

Fig. 13.1
figure 1

a Location of the hydropower plant GKI; Aerial pictures of the selected investigation stretches: b Kajetansbrücke, c Mariastein 1 and d Mariastein 2

For all three reaches unsteady 2D hydrodynamic models have been set up with HYDRO_AS-2D. These are the basis for the CASiMiR analysis and deliver detailed information for some of the before mentioned effect factors such as water level change rates and wetted areas, but also further parameters such as maximum flow velocities and sediment movement.

13.3 Identifying Representative Hydropeaking Events

For the peak impact assessment, sensitive periods for specific life stages of grayling were considered. March to April covers the grayling spawning and brown trout larvae season, whereas May to June is the emergence and early larvae period of grayling. October and November again describe the spawning season of brown trout. The flow time series in Fig. 13.2 reflects the temporal discharge variation at the gauging station Kajetansbrücke in the investigation stretch. The increase in the discharge for the months of May and July is caused mainly from snowmelt in the Alps. Summer months do not only present higher flows, but also higher variability of discharge.

Fig. 13.2
figure 2

Hydrological regime at Kajetansbrücke a 1999 hydrograph, b representative week hydrograph for a winter half-year

Main outcomes of the analysis are that the May/June period shows characteristics that are different to March/April and Oct/Nov. The base flow is not as constant as in the other periods, high fluctuations between 50 and 150 m3/s are detected, small events with a low baseflow are not present (see Fig. 13.2a). In consequence, the mentioned three periods were analysed statistically to derive representative events for the CASiMiR analysis. Not necessarily an average event or the most extreme event is suitable to describe the impact but rather events with high amplitudes and high change rates that occur regularly. Similar as in the COSH tool (Sauterleute and Charmasson 2014), after statistical evaluation, percentiles can be used to define this kind of events. The identification of peak events was made using a Peak Detection Model applying several steps. These steps and their relevance are listed in Table 13.1.

Table 13.1 Steps of hydropeaking event detection

Using the Peak detection model, events with different start- and end-flows and different up- and downramping rates could be identified as typical for certain periods of the year. Based on this analysis, for the current (impacted) situation, five different peaking events were detected and used for further calculations with CASiMiR:

  • Event 1: Big event [5 ➔ 94 m3/s] (Mar/April and Oct/Nov)

  • Event 2: Small event [5 ➔ 49 m3/s] (Mar/April and Oct/Nov)

  • Event 3: Big event with low base flow [20 ➔ 97 m3/s] (May/June)

  • Event 4: Small event with low base flow [16 ➔ 46 m3/s] (May/June)

  • Event 5: Big event with high base flow [68 ➔ 130 m3/s] (May/June)

For these events, we analysed:

  1. (a)

    Stranding risk

    The stranding risk for juveniles of grayling and brown trout was calculated using a fuzzy rule-based approach. Combinations of “low” water depth and “high” water level change rate lead to “high” stranding risk. “High” water depth does not imply any risk, independent of water level change rate. Thresholds for “high”, “medium” and “low” risk were defined with values lower 12 cm/h being uncritical, between 12 and 30 cm/h being critical and higher than 30 cm/h being very critical (Schmutz et al. 2013). Results for reach Kajetansbrücke and reach Mariastein 1 and event #3 are shown in Fig. 13.3 for grayling, in terms of a Risk Index RI between 0 (low) and 1 (high).

    Fig. 13.3
    figure 3

    Maximum stranding risk for juvenile grayling for event #3 conditions: a Kajetansbrücke, b Mariastein 1″. Weighted stranding risk area for juvenile grayling: c Kajetansbrücke, d Mariastein 1

    To gain quantitative information for the risk in the whole reaches the model elements with increased risk were integrated in terms of a Weighted Risk Area (WRA) equivalent to the Weighted Usable Area WUA used in Habitat suitability investigations (Bovee and Cochnauer 1977).

    $$WRA=\sum_{i=1}^{n}{A}_{i}\cdot {RI}_{i}$$

    Ai = area of model element I, RIi = Risk Index of model element i.

    The WRA for both reaches are shown for all 5 representative peaking events in Fig. 13.3. The WRA for juvenile grayling during event #3 is about 1.7 times higher in reach Mariastein 1 than in reach Kajetansbrücke. However, this is nearly the relation of the wetted area in both reaches, so the proportion of risk area is about the same. In reach Kajetansbrücke the risk index RI is higher in some locations, but there are also extended areas with low risk available that are usable for larvae.

  1. (b)

    Drift risk

    In CASiMiR drift risk is calculated by a fuzzy rule-based approach that considers combinations of “low” water depth and “high” flow velocity as critical. Results in Fig. 13.4 indicate that high-risk areas for juvenile grayling are larger in reach Kajetansbrücke and that low-risk areas appear in reach Mariastein 1 in much higher quantities. The WRA depends on the events. For event #3 the WRA is about the same in both reaches, but since reach Kajetansbrücke is only about 0.6 times as big as reach Mariastein 1 the risk potential in the first one is higher. However, for the other events WRA in the second reach is partly higher than in the first reach.

    Fig. 13.4
    figure 4

    Maximum drift risk for juvenile grayling for event #3 conditions: a Kajetansbrücke, b Mariastein 1. Weighted drift risk area for juvenile grayling: c Kajetansbrücke, d Mariastein 1

  1. (c)

    Spawning habitats and persistence

    Spawning areas in river reaches with hydropeaking are affected by two factors. First, they can fall dry during low flow and second, eggs can be damaged by sediment movement during high flows. Figure 13.5 shows the suitability of theoretically available persistent spawning grounds during the whole peaking event #3 (minimum suitability for all time steps) together with the areas where suitable spawning substratum (size approx. 16–32 mm) gets in to motion during increasing flow (dashed areas). Our analyses revealed that (a) almost no areas are available in the analysed stretches that have a suitability higher than 0.2 (persistent spawning areas) and (b) large areas would be affected by movement of suitable spawning substratum, if present (currently not the case).

    Fig. 13.5
    figure 5

    Persistent spawning areas for grayling during event #3: a Kajetansbrücke, b Mariastein 1

  1. (d)

    Habitat shift

    Finally, the spatial shift of suitable habitats for juvenile fish is a hazard for grayling and other fish. Due to their reduced mobility, juveniles up to a certain age cannot overcome larger distances when their habitats are moving with the water edge. This risk can be assessed by visualizing the location of good juvenile habitats for base flow and peak flow and overlaying them.

    Figure 13.6 shows the location of good juvenile habitat for both flow situations during event #3. In some locations (marked with green circles) the distance between good habitats for base flow and peak flow is comparatively small. These are the habitat shifts, which larvae can presumably follow when experiencing a peak event.

    Fig. 13.6
    figure 6

    Displacement (shift) of suitable habitats for juvenile grayling between based and peak flow for Event #3 conditions: a Kajetansbrücke, b Mariastein 1

13.4 Conclusions and Outlook

Currently the Upper Inn River is affected by hydropeaking. Along the stretch between Ovella and Prutz we have exemplified the application of the CASiMiR hydropeaking module. This module allows to analyse habitat availability as well as the risk of stranding and drift (of fish larvae) during different hydropeaking events. It enables the comparison of different planning scenarios and can serve as an assessment and planning tool.

The GKI is Europe`s first large hydropeaking diversion plant. Combining a buffer reservoir and a diversion stretch, it is possible to divert the hydropeaking further downstream to a larger catchment (Moreira et al. 2020), as foreseen by the water management framework plan Western Tyrol (Reindl et al. 2017).

The concept of the GKI will reduce hydropeaking in the Upper Inn between Ovella and Prutz, as the rapid habitat shifts are omitted (Herdina 2018; Moreira et al. 2020). The application of the CASiMiR hydropeaking module to the current situation and the situation after the implementation of GKI has confirmed that with the operation of the new HPP the hydropeaking impacts on fish will be significantly reduced and that the diversion of large hydropeaking events is an effective mitigation measure.