This
Shiny dashboard
enables the
Spatio-temporal analysis of historic and projected flood formation in the Rhine River basin
Summary
The genesis of riverine floods in large river basins often is complex.
Streamflow originating from precipitation and snowmelt and flood waves from different tributaries can superimpose and cause high water levels threatening cities and communities residing along the river banks.
In the framework of this study, we provide insights into the genesis of historic and projected streamflow peaks in the Rhine River basin.
Investigations base on hydrological simulations using the mesoscale Hydrolgical Model
mHM
forced with historic meteorological data based on observations and an ensemble of climate projections.
A detailed description of the model simulations analysed in this study can be found in
Rottler et al. 2021.
For each meteorological forcing, we determine the ten highest runoff peaks at four locations along the Rhine River (Cologne, Kaub, Worms and Speyer) in the fifty year time window 1951-2000 for the historical simulations and 2050–2099 for the climate projection scenarios.
The spatio-temporal analyis of the streamflow peak genesis includes the assessment and visualisation of antecedent liquid precipitation, snow cover changes, total runoff generated, quantile extent and the above-average runoff from the main sub-basins of the Rhine River (High Rhine, Neckar, Main and Moselle) up to ten days before a streamflow peak.
Overview figures
In the following, we provide a selection of figures produced using the open source graphing library
Plotly,
which provide an overview on streamflow peak characterstics. Figure content can be adjusted by clicking on the legend elements and zoom in/out options used to investigate regions of the plot in detail.
Detailed information on the plot content show up when the mouse moves over the plot.
To view the spatio-temporal analysis of the peak flow formation, switch to tab 'Peak genesis' and select your peak of interest.
Annual mean temperatures of model forcings
Mean annual temperatures are averaged over the model domain, i.e. the Rhine River basin until gauge Lobith, for historic simulations and GCM-RCP combinations.
Streamflow peak timing and magnitude
Timing and magnitude are two key characteristics of a streamflow peak.
Moving upstream the Rhine River from Cologne via Kaub and Worms to Speyer, more and more streamflow peaks are recorded outside winter.
Furhtemore, it seems that the potential for streamflow extremes differs depending on the GCM data used.
Streamflow peak magnitude and flood extent
In large river basins, very high water level only are possible if a sufficient part of the basin is generating runoff.
This seems also valid for the Rhine River basin at the selected gauges. Results indicate that the magnitude of streamflow peaks correlates with the extent of the event.
In the framework of this study, the extent of an event is estimated as the fraction of upstream grid cells that generated runoff above their long-term 99 % quantile (determined based on simulations for the time frame 1951-2000 using EOBS-based meteorological forcing data) at least on one day during a ten day peak formation period.
Streamflow peak timing and total snowmelt High Rhine
In the Rhine River, rainfall- and snowmelt driven runoff overlap. To assess the importance of snowmelt from the High Rhine basin with regard to streamflow peaks along the Rhine River, we investigate the cumulative snowmelt in the High Rhine basin upstream of gauge Basel up to ten days before the recorded streamflow peak.
Timeline
The timeline on top of the panel shows the date of the selected streamflow peak (date top right) and the ten days prior to the event. Depending on the selected day during peak formation, dates are highlighted.
Panel a
Simulated and observed Rhine River streamflow at the four locations Speyer (before the confluence of the Neckar River), Worms (after the confluence of the Neckar River and before the confluence of the Main River), Kaub (after the confluence of the Main River and before the confluence of the Moselle River) and Cologne (after the confluence with all major tributaries). Streamflow is displayed as the fraction of the long-term median average for a given day simulated for the time frame 1951-2000 using EOBS-based meteorological forcing.
Panel b
Simulated streamflow (observations of Basel for High Rhine) of the High Rhine, Neckar, Main and Moselle River. The High Rhine catchment is until the city/gauge of Basel. The simulated streamflow of the tributaries Neckar, Main and Moselle is taken just before their confluence with the Rhine River and does not reflect an actual excisting river gauge. Streamflow is displayed as the fraction of the long-term median average for a given day simulated for the time frame 1951-2000 using EOBS-based meteorological forcing.
Panel c
Cumulative above-average runoff, i.e. runoff above the long-term mean simulated for the time frame 1951-2000 using EOBS-based meteorological forcing data. The importance of rainfall and snowmmelt is estimated based on the amount liquid rainfall and snowmelt in the sub-basins during peak formation
Panel d
Quantile extent estimated based on the fraction of upstream grid cells that generated runoff above their long-term 99 % quantile (estimated based on simulations for the time frame 1951-2000 using EOBS-based meteorological forcing data) at least on one day during the ten day peak genesis. The quantile extent is calculated and displayed for areas upstream of gauges Cologne, Kaub, Worms and Speyer.
Panel e
Cumulative liquid precipitation up to ten days before the selected streamflow peak. Liquid and solid precipitation are differentiated based on the temperature threshold determined during model calibration.
Panel f
Cumulative snow cover changes (snow accumulation and snowmelt) per cell up to ten days before the selected streamflow peak.
Panel g
Cumulative discharge generated per cell up to ten days before the selected streamflow peak.
Panel h
Cumulative routed discharge per cell up to ten days before the selected streamflow peak.
Authors
This analysis was conducted by
Erwin Rottler
(member of the working group Hydrology & Climatology, University of Potsdam) in the framework of investigations focusing on future changes Rhine River flood seasonality.
The analysis was supported by
Axel Bronstert
(head of the working group Hydrology & Climatology, University of Potsdam) and
Gerd Bürger
(senior scientist within the working group Hydrology & Climatology, University of Potsdam).
We thank our colleagues from the Department of Computational Hydrosystems, Helmholtz Centre for Environmental Research (UFZ), particularly
Oldrich Rakovec,
and
Luis Samaniego,
for their support during model setup and data analysis.
Feedback
Should you have any comments, questions or suggestions, please do not hesitate to write us an email:
rottler(at)uni-potsdam.de
Code
The source code of this Shiny dashboard is available at:
github.com/ERottler/rhine-flood-genesis
Acknowledgements
We acknowledge the datasets generated in the
EDgE
proof-of-concept project performed under a contract for the
Copernicus Climate Change Service.
ECMWF
implements this service and the
Copernicus Atmosphere Monitoring Service
on behalf of the European Commission. We acknowledge EDgE colleagues
Rohini Komar
and
Stephan Thober
for establishing the mHM model setup and performing the downscaling of the CMIP5 data sets, respectively.
We acknowledge the E-OBS dataset from the EU FP6 project
ENSEMBLES
and the data providers in the
ECA&D
project.
We acknowledge the
ISI-MIP
project for providing the bias corrected CMIP5 climate model data. The
Copernicus Land Monitoring Service,
implemented by the European Environmental Agency, provided the European Digital Elevation Model
EU-DEM
, version 1.1. We also acknowledge the
HOKLIM
project
by the German Ministry for Education and Research (grant number 01LS1611A). We also thank various other organisations and projects for providing data used in this study, including
JRC,
ESA,
NASA,
USGS,
GRDC,
BGR,
UNESCO,
ISRIC,
and
EEA.