Plant invasions in riparian areas of the Middle Danube Basin in Serbia

Riparian areas experience strong invasion pressures worldwide and represent important points of spread for invasive alien plants (IAPs) in the European mainland. The Danube Basin is a well-known point of high plant invasion levels. Given that the middle part of the Danube Basin is critically understudied and the general lack of data for Serbia, the study aimed to provide an insight into the spatial patterns of plant invasions in the riparian areas of Serbia (Middle Danube Basin area). A total of 250 field sites, distributed along 39 rivers (nine catchment areas) and six canal sections, were studied during a four-year period (2013–2016) for the presence and abundance of IAPs. At the landscape scale, we studied distribution patterns of IAPs, differences in invasion levels in different catchment areas and between rivers and canals. At the local scale, we investigated how the proximity to roads/railway lines, housing areas, different land-use types (primarily agriculture), and dominant vegetation on site related to invasion patterns. Of the 26 studied IAPs, those with a well-known weedy behavior, long history of cultivation and strong affinity for riparian areas prevailed in the study area. Riparian zones of the Danube catchment exhibited the highest invasion levels in terms of IAPs richness and abundance, followed by the catchment areas of the Timok, Sava and Zapadna Morava rivers. Surprisingly, the Danube-Tisa-Danube canal network had the lowest invasion level. At the local scale, agriculture in proximity of the field site and dominant vegetation on site were observed as significant predictors of the invasion level. On the other hand, proximity to roads/railway lines and housing areas was not related to the invasion level. Finally, our study provides the first systematic overview of IAPs’ distribution data for riparian areas of the Middle Danube Basin in Serbia, which could provide a basis for long-term monitoring of IAPs and development of future management plans.

Examples of Floodplain Restoration Evaluation Studies in the Danube River Basin

<p>Countries located in the Danube River Basin (DRB) are in danger of being affected by major catastrophic floods along the Danube and its tributaries. Floodplain restoration measures are among win-win nature-based solutions (NBS) for flood risk reduction but practitioners see their limitations in comparison to technical measures, when looking at their effectiveness and profitability. Within the framework of the EU Interreg Danube Floodplain project, this presentation shows the benefits of floodplain restoration in terms of monetized ecosystem services (ES). Our work focused on multiple ES groups for four study areas in the Danube catchment, located in Czech Republic, Romania, Serbia, and Slovenia. This was done with the help of stakeholder engagement, hydrodynamic models results, and the Toolkit for Ecosystem Service Site-Based Assessment (TESSA). Moreover, the approach was complemented with alternative methodologies (e.g. surveys on social media). Results show positive annual combined benefits of floodplain restoration measures, suggesting the helpfulness of evaluating these NBS through ES assessment. The work done will help increasing the knowledge on floodplain and their ES, and on how to rapidly evaluate them. Moreover, it will bring decision-makers further evidence in favor of floodplain restoration measures to be implemented for a general benefit of the communities.</p>

Large-scale river network modeling using Graph Neural Networks

<p>In the recent past, several studies have demonstrated the ability of deep learning (DL) models, especially based on Long Short-Term Memory (LSTM) networks, for rainfall-runoff modeling. However, almost all of these studies were limited to (multiple) individual catchments or small river networks, consisting of only a few connected catchments. </p><p>In this study, we investigate large-scale, spatially distributed rainfall-runoff modeling using DL models. Our setup consists of two independent model components: One model for the runoff-generation process and one for the routing. The former is an LSTM-based model that predicts the discharge contribution of each sub-catchment in a river network. The latter is a Graph Neural Network (GNN) that routes the water along the river network network in hierarchical order. The first part is set up to simulate unimpaired runoff for every sub-catchment. Then, the GNN routes the water through the river network, incorporating human influences such as river regulations through hydropower plants. The main focus is to investigate different model architectures for the GNN that are able to learn the routing task, as well as potentially accounting for human influence. We consider models based on 1D-convolution, attention modules, as well as state-aware time series models.</p><p>The decoupled approach with individual models for sub-catchment discharge prediction and routing has several benefits: a) We have an intermediate output of per-basin discharge contributions that we can inspect. b) We can leverage observed streamflow when available. That is, we can optionally substitute the discharge simulations of the first model with observed discharge, to make use of as much observed information as possible. c) We can train the model very efficiently. d) We can simulate any intermediate node in the river network, without requiring discharge observations.</p><p>For the experiments, we use a new large-sample dataset called LamaH (<strong>La</strong>rge-sa<strong>m</strong>ple D<strong>a</strong>ta for <strong>H</strong>ydrology in Central Europe) that covers all of Austria and the foreign upstream areas of the Danube. We consider the entire Danube catchment upstream of Bratislava, a highly diverse region, including large parts of the Alps, that covers a total area of more than 130000km2. Within that area, LamaH contains hourly and daily discharge observations for more than 600 gauge stations. Thus, we investigate DL-based routing models not only for daily discharge, but also for hourly discharge.</p><p>Our first results are promising, both daily and hourly discharge simulation. For example, the fully DL-based distributed models capture the dynamics as well as the timing of the devastating 2002 Danube flood. Building upon our work on learning universal, regional, and local hydrological behaviors with machine learning, we try to make the GNN-based routing as universal as possible, striving towards a globally applicable, spatially distributed, fully learned hydrological model.</p>

Integrated Valuation of Nature-Based Solutions Using TESSA: Three Floodplain Restoration Studies in the Danube Catchment

Floodplain restoration measures are among the most well-known nature-based solutions for flood risk reduction but practitioners see their limitations in comparison to technical measures when considering both their effectiveness and profitability. The aim of this study is to show the co-benefits (besides flood risk reduction) of floodplain restoration and handle them in terms of monetized ecosystem services (ES). Our work focused on six ES groups for three study areas in the Danube catchment along the Krka, Morava, and Danube rivers. ES mapping through stakeholder engagement is also considered. We applied the methodologies suggested in the Toolkit for Ecosystem Service Site-Based Assessment (TESSA) complemented with alternative methodologies (e.g., questionnaires on social media). Results show annual combined benefits of floodplain restoration in a range from 237,000 USD2019 at Krka to 3.1 million USD2019 at Morava, suggesting the utility of ES assessment. The combination of stakeholder workshops and the TESSA guidelines, as well as the newly developed methods, were all central tools to provide decision-makers with arguments to use nature-based solutions for an integrated and holistic riparian land use management.

Current and future flood risk assessment in the Danube region

<p>Severe hydro-meteorological hazards have been increasing during recent decades and, as a consequence of global change, more frequent and intense events are expected in the future. Climate informed planning of adaptation actions needs both consistent and reliable information about future risks and associated uncertainties, and appropriate tools to support comprehensive risk assessment and management. <br>The Future Danube Model (FDM) is a multi-hazard and risk model suite for the Danube region which provides climate information related to perils such as heavy precipitation, heatwaves, floods and droughts under recent and future climate conditions. FDM has a modular structure with exchangeable components for climate input, hydrology, inundation, risk, adaptation and visualisation. FDM is implemented within the open-source OASIS Loss Modelling Framework, which defines a standard for estimating ground-up loss and financial damage of disaster events or event scenarios. <br>The OASIS lmf implementation of the FDM is showcased for the current and future fluvial flood risk assessment in the Danube catchment. We generate stochastic inundation event sets for current and future climate in the Danube region using the output of several EURO-CORDEX models as climate input. One event set represents 10,000 years of daily climate data for a given climate model, period and representative concentration pathway. With this input, we conduct long term continuous simulations of flood processes using a coupled semi-distributed hydrological and a 1.5D hydraulic model for fluvial floods. Flood losses to residential building are estimated using a probabilistic multi-variable vulnerability model. Effects of adaptation actions are exemplified by scenarios of private precaution. Changes in risk are illustrated with exceedance probability curves for different event sets representing current and future climate on different spatial aggregation levels which are of interest for adaptation planning.</p>

Enhancing River-Sea System Understanding by providing insights into headwaters– the Upper Danube Austria Supersite of DANUBIUS-RI

<p>Austria has a share in three international river basins (Danube, Elbe, Rhine), but by far the most of its territory (> 96%) drains into the Danube. This Austrian territory accounts for 10% of the total area of the Danube River Basin and belongs entirely to the Upper Danube Basins, which extends from the source of the Danube in Germany to Bratislava at Austria’s eastern border to Slovakia. Austria contributes approx. 25% (ca. 50 km³/a ) to the total yearly discharge of the Danube into the Black Sea (ca. 200 km³/a).</p><p>Human activities have severely altered the Upper Danube catchment, impacting both the main stem and the main pre-alpine tributaries. Due to the Upper Danube’s considerable natural gradient and mountainous character, this part of the Danube is extensively used for hydropower production. Ten large (> 10 MW) hydropower plants are situated along the Austrian Danube (out of a total of 41), and only two Danube stretches can still be characterized as free-flowing (Wachau, Nationalpark Donau-Auen).  Besides energy generation, other human activities such as agriculture, shipping, industrialisation, urbanisation and tourism, have been and still are changing the process and system dynamics of the Upper Danube.  Climate change is additionally affecting this already heavily impacted River System.</p><p>The Upper Danube Austria and its pre-alpine network of tributaries is therefore an ideal case study region to investigate the multiple effects of human activities on riverine systems and was chosen as a “supersite” within Danubius-RI, the “International Centre for Advanced Studies on River-Sea Systems”. Danubius-RI is being developed as distributed Research Infrastructures with the goal to support interdisciplinary and integrated research on river-sea systems. DANUBIUS-RI aims to enable and support research addressing the conflicts between society’s demands, environmental change and environmental protection for river -sea systems worldwide and brings together research on freshwaters and the interface to marine waters, drawing on existing research excellence across Europe.</p><p>The supersite “Upper Danube Austria and its pre-alpine network of tributaries” covers the freshwater spectrum within the river-sea continuum, ranging from alpine and pre-alpine headwater streams along major Danube tributaries to the Danube River, including adjacent floodplains in the Upper Danube catchment. The research focus lies on the interactive effects of climate change, land use pressures, and hydromorphological alterations on the biodiversity, ecological functions, and the ecosystem service provision of streams and rivers in the Upper Danube basin and their role within the catchment.</p><p>The Supersite “Upper Danube Austria and its pre-alpine network of tributaries” joins forces of eight Austrian research institutions and is led by WasserCluster Lunz and the Institute for Hydrobiology and Aquatic Ecosystem Management (IHG) at the University of Natural Resources and Life Sciences, Vienna (BOKU). Research on sustainable management and restoration of riverine landscapes (WFD, FD, HD, Biodiversity  Strategy) in the Upper Danube Catchment is an important contribution to a healthy River-Sea System of the Danube River Basin as a whole.</p>